TABLE OF CONTENTS

PREFACE

Ecosystem management of arthropods, what does it entail?………………..….... 3

Literature search

How many arthropods are there?……………………………………….... 4

Synopsis of literature search submitted……………………………. ….... 5

The nature of the arthropod information in the literature…………..…..... 6

PART ONE: Ecosystem process sustainability (top-down approach)

Soil food web structure……………………………………………………..…....10

Critical environmental determinants of soil arthropod density and diversity…...16

Critical ecosystem processes provided by soil arthropods…………….………...18

Information gaps

Top-down approach for monitoring …………………….…….....22

Highlighting biogeographically significant higher taxa;

analysis of species distribution patterns…..…………..….23

Why monitor with arthropods?………………………………......24

The most pressing needs………………………………………....27

Why focus on arthropods?…………………………………….....28

Monitoring priorities for focal ecosystem priorities………....…..30

PART TWO: Individual arthropod species and endemism (bottom up approach)

Introduction…………………………………………………………………...….33

Why is Klamath/Siskiyou region unique?…………………………………….....34

Individual potentially sensitive species: are there landscape patterns of

co-occurrence?…………………………………...…………………..…..40

Information gaps

Notion and utility of indicator species………………………………...…47

Long-term validation monitoring……………..………...……………......49

Landscape diversity and habitat management…..…………………….....54

Highest priority information gaps (bottom-up)….…………………...….56

Summary……………………………………………………………………..…..62

PART THREE: Specific prioritized research recommendations

Soil nutrient recycling………………………………………..…………………..63

Trophic pyramids; arthropods as vertebrate food………………..………………64

Pollination interrelationships of low-elevation forest species….………….…….67

Characterization of arthropod communities……………………………………...69

Summary…………………………………………………………………….…...72

FOOTNOTES………………………………………………………………………..…..73

LITERATURE CITED…………………………………………………………………..74

APPENDIX ONE: Glossary of terms…………………………………………………....84

APPENDIX TWO: What the original arthropod advisory panel to FEMAT said……....85

APPENDIX THREE: Bee (Apoidea) species of conservation concern (BSHT)……..…92

3A. Tentative listing of Vancouveran bioregion endemic BSHTs……………..102

3B. Tentative listing of BSHT bees further south……………………………...107

APPENDIX FOUR: Biogeographically significant higher taxa (Coleoptera,

Hymenoptera, Lepidoptera (butterflies), Diptera)….…………..……....109

APPENDIX FIVE: Example of BSHT presentation…………………………….…..…124

APPENDIX SIX: Tentative listing of Collembola of the PNW………..separate document

APPENDIX SEVEN: Tentative listing of Diptera of PNW…………....separate document

APPENDIX EIGHT: Tentative listing of Oribatida of PNW…………..separate document

APPENDIX NINE: Tentative listing of Carabidae of PNW…………...separate document

APPENDIX TEN: Tentative listing of spiders of the PNW……….…...separate document

APPENDIX ELEVEN: Tentative listing of Diplopoda of PNW……....separate document

APPENDIX TWELVE: Tentative listing of Cerambycidae of PNW….separate document

APPENDIX THIRTEEN: Tentative listing od Scolytidae of PNW…...separate document

APPENDIX FOURTEEN: Tentative listing of Apoidea of PNW….….separate document

FIGURES AND TABLES:

Figure 1. Structure of the invertebrates in a typical soil foodweb………………………12

Table 1. The usual components of the soil foodweb in nearly all ecosystems……...13-14

Figure 2. Harpaphe haydeniana……………………………...……………………...…19

Figure 3. Zootermopsis angusticollis……………………………………………….…..23

Figure 4. Amaurobiidae distribution of species………………………………….….….35

Figure 5. The distribution patterns of arthropods inhabiting the PNW……………...…36

Figure 6. Grylloblatta campodeiformis……………………………………………...…37

Figure 7. Agulla……………………………………………………………………...…37

Figure 8. Bequaertomyia jonesi………………………………………………….…..…38

Figure 9. Deuterophlebia inyoensis………………………………………………….....38

Figure 10. Tanypteryx hageni………………………………………………………..….39

Figure 11. Eulonchus tristis…………………………………………………………..…39

Figure 12. Snow scorpionfly, Boreus………………………………………………..….60

Figure 13. Glacier fly, Chionea………………………………………………………....60

Table 2. Comparison of standardized insect traps……………………………………....71

PREFACE

Ecosystem management and arthropods, what does it entail?

We were asked to review the available information and provide recommendations which would allow the USDA Forest Service and USDI Bureau of Land Management to protect the ecological functions of forest soil and litter arthropods and coarse woody debris chewers. We were also asked to focus on information gaps that could affect future policy and to comment on any potential problems relevant to current policy, particularly as it relates to the soil, litter, and coarse woody debris components of ecosystem processes. We will confine our remarks to those issues judged to be of the highest priority and broadest applicability only.

“Ecosystem management” is an approach based on maintaining large-scale biological processes (FEMAT 1993). In theory, efficiently maintained processes within the realm of naturally occurring variability should insure a healthy forest that is maximally buffered against perturbations that would adversely affect both short-term resource production and long-term forest sustainability. Ecosystem processes (such as primary production, nutrient cycling, pollination, succession, predation) are carried out by multitudinous arrays of individual species; arthropods play important roles in each of these processes. In a complex circular argument (supported by most ecologists, though with little experimental evidence), stable ecosystems permit the evolution of multiple species with similar but not identical functions—this “redundancy” of species then feeds-back positively on the system stability as a whole by permitting more efficient substitution and dampening the variability in ecosystem processes when the ecosystem is stressed. Hence, “forest health” can be interpreted to mean the preservation at relatively natural population densities of all the species within a forest ecosystem. Since it would be difficult to monitor all the arthropods within a forest ecosystem, priority must be placed on establishing the identity of species which participate in different forest processes and which measures of diversity are appropriate to indicate both the natural levels of variability and warnings when the system is being severely stressed.

There are probably in excess of 10,000 species of arthropods within the region encompassed by the Northern Spotted Owl (west of the Cascade Mountain crest from the Canadian border to and inclusive of northern California). Most of these arthropods are intimately associated with the soil/litter for at least part of their life cycles. No comprehensive listing exists of these species, nor is there any systematic compilation of the relevant literature. This diffuse state of information forms the setting of everything else to be stated below. In order to stimulate a useful collation of this material, the Forest Service requested that we assemble pertinent literature on these soil/litter-dwelling groups, especially with reference to basic ecological process studies and inventory techniques. To that end we were provided a list of 1262 references produced by a search of an archival database; the search keywords were an abbreviated list of the families of litter/soil dwelling arthropods known to occur in the Andrews Experimental Forest (Parsons and others 1991). This list was examined, several hundred references selected and annotated (either with official abstracts or a personal evaluation). Since the original list provided was severely limited, we initiated a preliminary literature search based on all the genera of arthropods known to occur in the Pacific Northwest (since the large majority of them are soil-associated) based on the relevant library databases. This process, to be subsequently accompanied by production of a faunal list of all the Pacific Northwest (PNW) species, will take several more years to complete. For the purposes of this contract though, we have searched the databases (tens of thousands of additional references) on major taxonomic groups: spiders (Araneae), turtle-mites (Oribatida), carabid ground-beetles (Carabidae), longhorned-beetles (Cerambycidae), bark-beetles (Scolytidae), flies (Diptera), springtails (Collembola), bees (Apoidea). Additionally, we searched the database for inventory techniques and soil ecological terms in order to compile an annotated bibliography of the most pertinent references (most recent continually updated version subsequent to this one may be found at http://www.osu.orst.edu/dept/entomology/moldenka). The current bibliography is over-representative of such topics as bark-beetle ecology and ground-beetle toxicology studies. All of the references have been given additional keywords to facilitate use by subsequent workers (Appendix 15-16).

In order to prioritize topics or locations of special concern, a subset of the resident species (“indicator species”) are chosen to permit comparative analysis of multiple issues. Since there is currently no complete listing of resident arthropod species for this region, how can one evaluate the possibility of using a restricted set of designated “indicator species” for inventory studies? How can one determine hot-spots of biodiversity for special conservation attention? Is it possible to prioritize a short-list of sensitive species of special concern? The answers will necessarily contain a significant amount of archival bias; better-known species provide more extensive ability to interpret the results. Even after our faunal listing and literature search is completed in several years, the impediments will basically remain the same since the ecological knowledge base for species varies significantly (including no knowledge for many taxa). The appended professional judgments are based on communications with a great number of entomologists and ecologists.

Literature Search

How many arthropods are there? Which ones are most worthy of attention?

What are the landscape level patterns of arthropod richness and rarity?

We have conducted an extensive search in the literature for information pertinent to these multi-faceted issues in the Northwest. We started with a list of 1262 references, provided by the Forest Service and originally generated by the HJ Andrews arthropod inventory list of families of arthropods (Parsons and others 1991). We searched all the usual web databases (e.g., Zoo Record, World Cat, Biosis, Agricola (FirstSearch and Webspirs), Biol & Ag Index, CAB Abstracts) for listings involving these suggested taxa. These searches generated tens of thousands of references; we reduced the number of especially pertinent references to several hundred, obtained abstracts for any in this category that were not provided on the web, and categorized them into groupings that would make them far easier to use. Since it was apparent that the original list did not include many families of arthropods known to be associated with the soil or coarse woody debris, and since the most useful sorts of references were frequently found by doing generic-level searches instead of family-level searches, we initiated a multi-year attempt to broaden this search of the literature to all the genera of certain groups of largely soil-associated arthropods known to live in the PNW. This search is on-going and any information resulting from it will be available to any management agency that requests it (see http://www.osu.orst.edu/dept/entomology/moldenka/). All the genera of Oribatida, Scolytidae, Cerambycidae, Diptera, Apoidea, Collembola and Araneae inhabiting the PNW have already been searched. A similar endeavor focusing on understory-dwelling herbivorous arthropods was compiled by Pacific Analytics (2001) and is available on the US Forest Service web page (http://www.fs.fed.us/r6/nr/fid/

pubsweb/litsurvey/index.html).

First drafts of species lists of the Apoidea, Araneae, Collembola, Oribatida, Carabidae, Cerambycidae, Diplopoda, Scolytidae and Diptera of the Pacific Northwest are in Appendices 6-14. The majority of species in each of these groups are soil- or CWD-associated).

Synopsis of literature search submitted

The dataset (SoilArth) was compiled by searching the literature databases for 1) genera of arthropods inhabiting the Pacific Northwest, 2) for various sampling methodologies, and 3) for other relevant terms: soil arthropods, fire, forest underburning, clearcutting, compaction (soil compaction, forest soil compaction), Pacific Northwest forest, woody (woody soil, woody forest, woody forest soil, coarse woody debris, woody debris removal), litter (soil litter, forest soil litter, litter removal), herbicides (forest herbicides), canopy removal, non-target effects, fertilizers, pesticides, forest management, endemism, trapping methods of various kinds (pitfall, etc.), soil biology, soil ecology, forest environment, forest ecology, forest ecosystem management, biodiversity, Klamath, Siskiyou, and any other permutations or phrases that came up as categories while doing the searches. For example, if the word “fire” was used in a specific way in the databases, e.g., as “fire ecology”, then that phrase was entered and researched as well. Thus one word could lead to a number of offshoots, each of which was followed as long as deemed relevant.

The dataset was entered in the software program, ProCite, and contains 2964 references separated into two groups, those with abstracts and those without. Descriptive keywords have been added to each reference for quick identification. Specific fields have been selected on the ProCite workform (under the View menu > Configure Record List) to show “author”, “title”, “date”, and “keyword”. Other pertinent information is also present for each reference, such as source, notes, abstract, etc. ProCite gives one the option of opening a larger “preview pane” for each entry that can be pre-selected (View menu > Configure Record List) to show some of these fields, such as abstracts. The source of the reference (e.g., journal title) automatically appears in the preview pane.

You may use the bibliography in the following manner. If you have ProCite5 and are interested in finding references about Dendroctonus bark-beetles, simply click on the tab below the spreadsheet marked “Search”. In ProCite4 “Search” would be on the menu above the spreadsheet. Type Dendroctonus in the “search for” window, making sure that “all records” shows in the “look in” window. If using more than one search word, “and”, “or”, “not” or some other operator must be included between each word. (Look under the “Operators” menu above this window for options). For example, in our dataset, we generated 788 records when the word Dendroctonus was typed into the search window; Scolytidae generated 462 records; and barkbeetle 1. “Bark and beetles” yielded 219, “bark and beetles or Dendroctonus” yielded 902, “bark and beetles or Dendroctonus or Scolytid” yielded 916, and “bark and beetles or Dendroctonus or Scolytid or Scolytidae” yielded 1086 records, etc.

You may also use the keyword column in which each reference is labeled by subject. By tapping the grey “Keywords” column header, the list will automatically sort itself alphabetically, and if you tap it again, it will sort itself in the opposite order (true of all the column headers). If the keyword list is not showing on the screen, go to the “View” menu > Configure Record List, and put a checkmark by “keywords”. This column will then appear on screen. If your interests pertain to a broader subject, such as fire ecology, just insert “fire and ecology” into the search window, or look under fire ecology in the keywords list.

The nature of the arthropod information in the literature

To understand patterns of abundance, species diversity and endemism of soil-associated arthropods on a landscape basis, we focused on 3 groups of soil and coarse woody debris (CWD)-associated species: spiders (excluding the poorly-known Linyphiidae), Carabidae, and Apoidea. (The Apoidea are all soil- or CWD-associates for 50 weeks out of the year; during their 2-week adult activity period they are seldom further than a few inches away from the soil.) We confined our analysis to these three higher taxa since they are: (1) each highly diverse, (2) each generally well-known taxonomically and (3) each represents a group the senior author is personally quite familiar with from previous research over the past four decades. We felt that we were able to both assemble and interpret the information on these three groups more adequately than any other arthropod groups (Moldenke 1976b, 1979c). The Oribatida and Collembola are so poorly scientifically documented as to render in depth analysis of beta and gamma diversity futile. The specific numbers relevant to distribution patterns, rarity and endangerment we report are based on the most recent group we analyzed (the Apoidea), but the general patterns apply equally well to all three groups. Both spider and carabid distributional analyses suffer from imprecise habitat information, and are (at best) overlapping map patterns for each species, each with poor resolution. Hence, separation of the alpha and beta components of diversity is problematic. The gamma diversity component of the elevated biodiversity of the Klamath/Siskiyou region is apparent for both of these groups, as is the general North/South gradient of biodiversity. For both the Carabidae and the Amaurobiidae (spiders) the Klamath/Siskiyou region is known worldwide as a center of biodiversity (Leach 1972, David Kavanaugh, unpublished data on Nebria [13]). None of the diversity patterns we have documented for these three groups come as a surprise to us or differ from patterns assumed by the FEMAT panel of experts, but we felt it necessary to quantify these general distributional patterns due to the gravity of the questions being asked.

To use the entomological literature, it is necessary to understand the fundamental nature of the taxonomic database. Taxonomy usually proceeds in three distinct phases. The first phase is akin to “Wow, this looks different! Let’s call it a new species, since no one has collected in this region before.” This leads to a common pattern of many different scientific names for the same widespread species. The second phase is analysis of specimens from throughout the entire extent of the species distribution; the emphasis is on basic similarities rather than minor differences. Such “taxonomic revisions”, as they are called, lead to long lists of synonyms and a de-emphasis on local differentiation. The third phase is an in-depth analysis of the landscape patterns of variability, separating out the forms that are geographically (or otherwise) isolated from those that are maintained within breeding populations. Though most plants and vertebrates have passed through all three phases of taxonomic development, few insects have. Perhaps butterflies (Hinchliff 1994) and tiger-beetles (Willis 1968; Pearson and Cassola 1992) are the only examples in the Pacific Northwest to have entered phase #3. Other arthropod groups are split between phase #1 and phase #2, with most groups still in phase #1.

The third phase is critical for ensuring scientific credibility; variants of a species are appropriately recognized for their potential contribution to long-term evolutionary potential of the overall species. Long-term evolutionary potential, in this case, means divergence of new forms or variants in unique environments or in peripheral locations. For example, Gavrilets (1999) found that “Rapid speciation is most likely for populations that are subdivided into a large number of small subpopulations.” Thus, to maintain the evolutionary potential of a species that displays local variants, one might strive to formally identify those variants (i.e., the third phase mentioned above) and then subsequently conserve them.

The taxonomic status of bees (Apoidea) provides a good example of the status of arthropod taxonomy. As a whole, the group is relatively well-known in the Pacific Northwest. Nearly all of the genera have been revised during the past 40 years, and identification keys (for use by the trained specialist) are generally available. A list of all the species known to occur, or suspected of possibly occurring, in the Pacific Northwest is appended (Appendix 14). However, even within this well-known group as a whole, the second, third and fourth most speciose genera have not yet been revised. The second largest (Nomada) and the third largest (Dialictus) are comprised of more species without names than with names. The fourth largest (Osmia) undoubtedly contains so large a number of synonyms as to be intractable. And yet, I state without chance of refutation, that bees are a well-studied group relative to the total range of arthropods. The most speciose genus in the Pacific Northwest is Andrena; the database records 174 species in the region. During the phase #2 revision of just these northwestern species, 297 names were placed in synonymy (nearly twice as many as there are valid names!). At the same time, 31 of them were described as new to science. This is the state of the arthropod database.

One of the major procedural difficulties that faces arthropod ecology is the process of identification. Even when recent monographic treatments of a group are available, it is difficult to impossible for a novice to use the keys due to either specialized vocabulary or a lack of comparative identified museum material. In an attempt to relieve this procedural problem and allow easier identification of field-collected specimens, we have been working on a user-friendly computer key “COMTESA – COMputer-assisted Taxomony and Ecology of Soil Arthropods” (www.ent.orst.edu/comtesa). The goal is to provide illustrated keys in a database matrix format which allows rapid identification of specimens and subsequent access to literature citations and ecological data. The process of programming has been largely completed, and portions can be viewed as a test case. Instructions in Appendix 17 will allow the reader to visualize how the identification and information-retrieval functions will operate. The entire process is still under construction, but instructions relevant to identifying a bumblebee specimen are included in Appendix 17.

PART ONE: Ecosystem process sustainability vis-à-vis arthropods (top-down approach)

Although there are many components within the structure of any ecosystem, ecologists generally focus first on the structure of the energy flow through the trophic pyramid as the prime emphasis in their descriptive models. As such, ecologists are interested in the quantity of energy/biomass that comprises the various subcomponents of a trophic pyramid as well as the regulatory processes that determine the energy flow through the entire pyramid (Schowalter 2000).

In PNW conifer forests it is well established (1) that the majority of primary production is sent to the roots both for root growth and (primarily) for mycorrhizal functioning; and (2) that the majority of the photosynthetic tissue produced on an annual basis ends up in the litter (and is not consumed by typical herbivores). This signifies that the soil, litter and coarse woody debris portion of the conifer forest ecosystem is a very large, if not the most significant component (especially if standing dead woody material is included). Though most of this biomass or nutrient potential is encompassed within dead plant matter, bacteria, fungi and protozoa – arthropods do comprise a significant portion. The major significance of arthropods in this environment is neither their standing crop biomass nor their respiratory rates, but their role as regulators of microbial growth, nutrient decomposition and plant growth. Though some of these effects are direct (i.e., fungivory, or major source of prey items for secondary consumers), most are indirect (i.e., shredding plant material) and contribute to the process of buffering soil “health” and building the soil matrix itself. Hence, the biological processes taking place in the soil are as significant to a forest as those taking place in the canopy.

The biomass and nutrients encompassed in arthropods is passed on to the next higher level of the trophic pyramid as well. The majority of species of vertebrates are insectivorous, or primarily insectivorous during periods when their young are growing the fastest. Examination of these energy transfers demonstrates how integrated the forest ecosystem is. Since both the soil arthropods and vertebrate predators may be highly mobile, there can be very significant transfer between ecosystem components. For instance, between 26-53% of the diet of terrestrial birds and aquatic fishes can be made up of arthropods coming from the opposing environment during the course of the year (Nakano and Murakami 2001).

It is known that the soil organic matter, litter, coarse woody debris and their component arthropod faunas vary enormously through the region inhabited by the Northern Spotted Owl. Though this has not been adequately quantitatively documented for arthropods, it is universally assumed to be true since fundamental changes have been documented in the plant community structure and quantified as well in terms of soil nutrient capacity and depth of organic layers. It is also known that the arthropod fauna of concern varies significantly through successional time at any one locality, since the vertical structure of the plant community as well as the herbivore-resistence of the leaves changes greatly. Through the seasonal limitations of low temperature and evapotranspiration, total soil metabolic activity is strongly bimodal – which ultimately controls most arthropod activity as well. Hence, the soil arthropod communities no doubt vary tremendously across both spatial and temporal aspects.

The large organic component of the soil and its widespread heterogeneity produces a high level of species richness in the soil, litter and coarse woody debris faunas. As a general rule, the “soil” is the most diverse component of any ecosystem, especially in so far as arthropods are concerned. This diversity of arthropods can be functionally subdivided into a rather large number of relatively separate functional guilds, each of which is characterized by species belonging to rather distinct taxonomic groups (see Moldenke and others 1999 for analysis of PNW conifer soil fauna).

Soil food web structure (see Lewandowski 1999)

The basic structure of all natural soil food webs requires a minimum of four components; (1) the living plant root, which is incapable of enzyme production, but obtains nutrients for the plant passively through water absorption; (2) the soil bacteria and fungi, which are in most circumstances fuel (carbohydrate)‑limited, but are capable of a bewildering diversity of degradative enzymatic decomposition processes; (3) the fauna which graze upon the microbial biomass, thereby releasing, as part of their waste products, water‑soluble (and hence "useful") forms of plant nutrients; and (4) the predator fauna which keep the population of microbial grazers below levels which might limit the microbes in their pivotal role as the "biological sponge" which prevents loss of nutrients from an ecosystem.

The cycle of growth and decomposition is driven by photosynthesis. Plants send more than 50% of the carbon fixed by photosynthesis directly to the roots to be excreted in order to facilitate microbial growth. This fuel is shared with both specific symbiotic fungi (mycorrhizae) and the general microbial population of the rhizosphere (the volume of soil within 1 to 2 mm of a living root tip). Bacteria are important as local foci of intense metabolic activity on a scale of several millionths of a pinch of soil. Fungi are capable of slower metabolic activity, but because their basic body form, similar to a railroad system, allows the simultaneous exploitation of widely spaced, yet different, required nutrients, activity can be maintained continually. Both bacteria and fungi are relatively incapable of movement; they require insects and other arthropods to bring them to sites of new resources or to physically mix the resources currently available. Arthropods that graze upon the fungi are seldom able to consume them entirely, resulting in stimulation of rapid regrowth (like pasture plants subjected to grazing by cattle). The result of all of this microbial growth is the immobilization of nutrients into living biomass. The long-term health of any ecosystem is intimately dependent upon minimizing the rate of nutrient loss: the greater the biomass of soil microbes, the slower the rate of nutrient loss. The progressive decrease in soil organic matter (both living and dead) in agricultural soils in North America is a major factor leading to chemical pollution in waterways and aquifers in the United States.

The grazing by soil fauna leads to momentary availability of soluble nutrients for two reasons. (1) The ratio of carbon to nitrogen in the tissues of bacteria and fungi is less than that of animals. As a consequence, an organism feeding on a bacterium or fungus must excrete any nitrogen in excess of its own requirements (or become poisoned by it). (2) As an animal locomotes in search of food, it burns (respires) carbohydrates for energy. As CO₂ leaves the body, other nutrients (nitrogen, etc.) must leave it as well in order to maintain the ratio within living cells. Some of the nutrient release by the fauna is immediately recaptured by microbes in their vicinity, while a significant portion is released close enough to a root to be swept into the plant by the inward rush of water caused by transpiration from the leaves. Since most faunal activity is within the rhizosphere, faunal activity does result in nutrient uptake (but generally only when the stomata are open). A generalized foodweb of the diversity of soil‑ and litter‑associated invertebrates is represented in Figure 1. The arthropod groups comprising each of these categories are presented in Table 1.

Soil is physically and chemically heterogeneous at the scale of microbes and arthropods. The movement of arthropods aerates the soil and transports microbial inocula to regions of nutrient availability. Microbial diversity, though fundamental, is difficult to

measure. Arthropod diversity is far easier to monitor and is likely to reflect microbial diversity. Soil degradation is associated with a decrease in soil fauna diversity (Coleman and Crossley 1996; Benckiser 1997). Some forms of degradation are more easily measured structurally or physically, for instance the depth of litter or the resistance to a penetrometer. But complex soil chemical changes may best be monitored through arthropod biodiversity. Any change in invertebrate diversity should be interpreted not just in isolation, but as an integral part of the basic functioning of the soil foodweb, which supports all plant life.

There are indistinct boundaries on the definition of soil-associated arthropods; the herbivorous fauna on the grass, forb and sometimes shrubby vegetation is a major component of the food resource of epigeic (soil-surface-dwelling) macropredators. Though many invertebrate predators (i.e., Calosoma sycophanta) routinely ascend tall trees to hunt, they represent exceptions and are not considered in this discussion. Far more important, are all the species of truly arboreal herbivores, which descend to the ground to pupate. Such species not only provide food resources to soil‑dwelling predators, but they represent important additional linkages between potential canopy health and prescribed burning protocols.

Generally, the biomass of organisms high in the trophic pyramid is relatively insensitive to shifts in species composition of the lowest trophic levels and is theoretically sensitive only to the total primary production of the ecosystem. However, the species comprising the higher levels of the trophic pyramid are very sensitive to changes in management policies. To understand whether land management practices will affect community structure, it is necessary to differentiate the producer biomass into separate components (i.e., within column I) and to quantify how efficiently this biomass is transformed into the different groups of primary consumers (i.e., column II). The majority of the biomass and species richness of any arthropod community is contained within the primary consumers. Any (and perhaps, every) major change in management policy will alter the percentages of arthropod biomass between boxes of the primary consumers, and consequently within the secondary consumers as well. Such changes are

Table 1

not necessarily "bad", but it is important to recognize such a change (and quantify it) when it occurs. The primary consumers are the engines that drive these 20 separate processes within an ecosystem. In general, increases either in the biomass of resources or their spatial heterogeneity will favor increased species richness within each separate process.

The actual dynamics of the soil foodweb in any Pacific Northwest community is far more complex than the one outlined in Figure 1. The foodweb structures among differing community types (forest, desert, grassland) are likely to differ significantly, both in terms of the relative biomass included within the boxes and the rates of energy flow between them. No foodweb analysis of invertebrates has ever been attempted in the Pacific Northwest on other than a very generalized basis. Figure 1 represents: (1) a structure into which most communities can be accommodated, and (2) a structure that is adapted for invertebrate bio-indicator use.

The farther to the right one proceeds in Figure 1, the larger the individual organisms become and the easier they are to collect and monitor. However, the further to the right one proceeds, the lower the species richness (and the more generalized the behavior and, therefore, the less useful it is at indicating basic changes in energy functioning within the ecosystem).

There are more than 10,000 species of arthropods (though many of them do not even have scientific names yet) within the Pacific Northwest. Though the specific identity of the members of any one box in Figure 1 will vary between community types and between sites, at the family‑level of taxonomic resolution there is considerable consistency. We present the taxa in Table 1, as a checklist for comparing different sites. This list is certainly not all‑inclusive, but it attempts to encompass the great majority of arthropod taxa found associated with the soils of the Pacific Northwest. It is not yet possible to enumerate all the species of arthropods at any one site, but it is certainly logistically practical to monitor several of the component groups when most of the species are collected in a particular trapping system. For instance, epigeic macropredators, macroherbivores and macrodetritivores can be collected in pitfall traps and compared between treatments (Niemela and others 1993). The insect canopy fauna of different age stands can be compared to see if there are qualitative as well as quantitative differences (Schowalter 1989, 1995). The trapped specimens can be weighed to get an estimate of biomass, and analysis at the resolution of morphospecies of the component groups can easily indicate significant changes in species richness. These taxa are listed in the table, not as an attempt to facilitate a one‑species‑at‑a‑time approach to conservation/monitoring, but rather as an educational device. Most ecosystem managers are unaware of the functional roles played by invertebrates, and even if they are aware of the roles, they have never learned who does what. Managing our natural resources for the long‑term will require ensuring the persistence of all the functional links in the community.

Critical environmental determinants of soil arthropod density and diversity

The overall level of soil arthropod biomass and diversity is regulated by a number of factors under natural conditions. The relative importance of these factors vary in different environments within the Pacific Northwest, and consequently affect the soil arthropod fauna substantially.

Grazing—In general, any photosynthetic structures that end up in the stomach of a vertebrate grazer are subtracted from the resource base for invertebrates. A reduction in heavy grazing by mammals will also increase habitat structure and refugia.

Drought—Precipitation is generally highly unpredictable for certain critical periods in the year. Total annual precipitation coupled with periodicity determines gross primary productivity in large part. Primary productivity is the most important determinant of secondary productivity.

Habitat Refugia—Every species requires refugia of some sort: the more numerous and diverse the potential refugia, the more individuals and species will prosper. In all habitat types a premium must exist on physical means that promote vertical circadian migration patterns. In forested systems, consolidated clay soils are frequent, which limits vertical movement to those few taxa that have the actual ability to burrow, or to those fauna that can utilize the burrows created by other species. Very few soil invertebrates are physiologically capable of true burrowing; most of them, such as anecic earthworms and geophilomorph centipedes are absent from drier soils. Cicadas, millipedes, gryllotalpids and stenopelmatids may exercise 'ecosystem engineer' keystone roles in these environments. In unconsolidated soils, true burrowing is not required and many/most taxa may be able to bury themselves sufficiently to prevent desiccation.

Surface compaction/cementation—Any change leading to a decreased ability of the fauna to pass through the boundary layer or a decreased infiltration of precipitation is generally detrimental. In many environments a cyanobacterial crust forms on the surface of the soil. The relationship that this crust has on the total community is unclear; one would expect it to facilitate faunal diversity due both to its nutrient biomass and ability to absorb water.

Litter depth—In forest environments the litter layer is apparently positively correlated with increased arthropod density and diversity (Madson 1997). Litter serves as both a food source and in a three‑dimensional habitat capacity. In scrublands a definitive litter layer is generally absent, but individual detrital pieces are the center of soil invertebrate activity, especially after they are anchored in place by fungal hyphae and are no longer subject to wind displacement. Early in the morning vertical migration of the microfauna creates an evanescent hot spot of activity within the litter (Wallwork 1970).

· Erosion—In forest environments erosion is promoted principally by fire. However, on steep slopes within semi‑arid forests, the passage of mammals (cows, deer, humans) can also seriously disrupt the fragile litter layer. Under circumstances in which nutrients are limiting, the continual flux of recyclable litter and nutrient‑rich topsoil downslope is likely to pose a serious stress to upland vegetation. In the scrublands, erosion (both wind and surface rainwater) is accelerated by any disturbance of the surface layer; the most usual form of disturbance is by mammalian passage.

Condensation/seasonality—For much of the year many surface soil microenvironments are severely limited by water; the only reliable source of water is dew formation. Any physical or chemical attributes of the soil surface that promote condensation should be encouraged.

Fire, the ultimate consumer of resources—For dominant “climax” organisms, most effects of fire must be detrimental, at least initially. Historically there were probably a number of beneficial effects upon the subdominant and early successional taxa, as there clearly are in grassland/savanna biomes worldwide. The multitudinous roles of fire in Pacific Northwest forest and scrub communities have never been examined with respect to their effects on upon non‑pestiferous invertebrates. Any features of a prescribed fire management practice that increases heterogeneity and promotes islands of unburned soil surface will minimize the detrimental effects of fire.

Water‑holding capacity—In most of these environments anything that favors infiltration and deep water‑holding capacity is paramount. Most of these variables are long‑term soil genesis properties that are liable to remain the same under different management scenarios. Though coarse woody debris acts as a sponge anywhere in the soil profile, it probably has its greatest hydrating potential below the surface layers. Stumps, with their vascular bundles oriented to facilitate upwards water flow, may not be as useful in retaining water below‑ground as an equivalent amount of coarse woody debris oriented parallel to the surface.

True deep‑burrowing—Vertebrate burrows within the mineral soil are probably an important resource (both occupied and unoccupied burrows) in all ecosystems as they are in the well‑publicized Florida sand pinelands (Myers 1985). There must be numerous inquiline and accidental “guests,” that require such burrows for persistence in the community. The percentage of the fauna dependent upon vertebrate burrows probably increases with aridity. Classic studies by Ehrlich and others have shown the population density of the Eulphydryas butterfly to be largely controlled by the populations of gophers, whose burrowing activity alters host‑plant phenology, permitting caterpillars to reach a threshold body weight before aestivating. As first noted by Darwin, community-wide effects of mammal burrowing are also documented in many temperate communities in which bumblebees are the major pollinators; bumblebees require abandoned burrowing-mammal nest chambers in which to establish their colonies in the spring.

Deep soil compaction (through overgrazing or mechanized vehicle use)—This is probably the most severe threat in any environment to soil arthropods, since as far as we know it is biologically irreversible. It creates an abiotic environment and potentially permanently immobilizes resources within it.

Critical ecosystem functions provided by soil-associated arthropods

From the plant's perspective, nutrient recycling must make available the limiting chemicals incorporated within the forest floor litter and provide a supply of required chemicals from the inorganic geologic substrate. Because plant roots are incapable of secreting enzymes to carry out these chemical transformations (Curl and Truelove 1986), plants are ultimately and completely dependent upon exoenzyme production by bacteria and fungi in the soil, and to a lesser degree upon the enzymes secreted within the digestive systems of soil-inhabiting fauna (from protozoa to vertebrates). Some abiotic chemical degradation occurs in the most extreme environments (deserts, bare rock surfaces), but these are exceptions to the general case of dependence upon microbial enzymatic degradation. Microbial degradation of resources within the soil matrix leads primarily to growth of microbes themselves, because the potential absorptive surface area of microbes greatly exceeds that of roots. Thus, microbes can generally compete more efficiently for soluble nutrients. Microbial exoenzymes benefit plants only if some of the enzyme catalysis occurs directly next to a root tip or if mycorrhizal fungi directly share resources with the plant. Bacteria and fungi serve as a biological sponge in the soil, "immobilizing" or "pooling" nutrients from dead organic residues and the inorganic substrate by incorporating them into their own living biomass (Powers and others 1999).

Soil arthropods (and other invertebrates) facilitate microbial growth and chemical degradation by:

Shredding the dead plant parts in forest litter and burrowing into coarse woody debris. This increases surface area for microbial colonization.

Moving either the resources or the microbial inocula in a continual homogenization of the soil.

Improving the water‑holding and oxygen‑penetrating capabilities of the soil through the geometry and chemistry of their fecal pellets, which are the O and A layers of the soil itself.

Enhancing the expression of the full range of chemical potentials of soil microbes by facilitating the succession of taxa with different exoenzymic capabilities.

Liberating into the soil solution (also known as “mobilizing” or “mineralizing”) the nutrients pooled within the microbial biomass of the rhizosphere by grazing upon the bacteria and fungi currently active around or within the microbial sheathing of physiologically active roots (Webb 1977; Seastedt 1984; Visser 1985; Setala and Huhta 1991; Moldenke and others 1994).

Shredding and Burrowing

The half‑life of different kinds of coarse woody debris on the forest floor is largely a function of the abundance and diversity of wood‑boring insects (largely ambrosia, long‑horned or metallic woodboring beetles) (Schowalter and others 1988). Wood-boring insects carry microbial inocula and provide increased access for fungal attack (Ingham and Moldenke 1995). Often these beetles require the presence of actively growing fungi to concentrate limiting nutrients within the wood (Crowson 1981; Martin 1987).

Nutrients in dead leaves or needles are largely unavailable to most microbes. A bacterium in the leaf litter is analogous to a person in a pantry without a can opener. Eventually a person can beat a can open, but it is much more effective to have a can-opener. The arthropod shredder is the bacterium's “can-opener”. Bacteria and fungi will eventually use all of a dead leaf, but they are far more efficient if the leaf is shredded first. Shredders (i.e., millipedes [Figure 2], earthworms, sowbugs) crush vast quantities of plant cells from which they extract only the most readily available nutrients‑‑the rest enter the normal soil recycling chain as the shredders defecate the crushed fragments (Hopkin and Read 1992). Passage through numerous shredding devices (the mouthparts of even much smaller organisms, i.e., penknife mites) is required before all the resources are finally available for complete enzymatic digestion (Dawod and FitzPatrick 1993). This process of continually refined shredding takes time, which accounts for the persistence of humus layers in the soil.

Figure 2. Harpaphe haydeniana is the most important and widespread macro-shredder in the Pacific Northwest conifer forests. Studies indicate that it shreds 33-50% of all leaves (both coniferous and deciduous) that become part of the forest litter (Baumeister 2002 and unpub.; Carcamo and others 2000). (illustration from Audubon Field Guide 1998)

Transportation and Homogenization

People take mobility for granted until they become physically disabled. There may be 500 million bacteria in a teaspoon of forest soil, but each one is largely incapable of movement, even though some bacteria have flagella that permit limited movement. Each bacterium needs diverse nutrients for growth and reproduction. From a bacterium's point of view, a resource that is several microns away is infinitely far away, because some other competitor is likely to be closer to the resource. There are several solutions: find a method of travel to cross the distance (hitchhike); have someone bring the resource to you; or hope that all your competitors between you and the resource already have enough of the resource and will not steal it before it diffuses back to you. Soil fauna fill the first two roles and provide a better chance of success for the bacterium.

Bacterial and fungal inocula can be carried either on the outer body surfaces of invertebrates or in their intestinal tracts (Visser 1985; Anderson 1988). In general, the number of spores carried phoretically is directly proportional to the surface area of an organism. The viability of ingested inocula is normally proportional to the time it requires to pass through the length of the gut (Anderson 1975). Well‑fed individual invertebrates have high percentages of viable inocula in their feces, whereas poorly fed individuals produce feces with minimal viable inocula. The best‑known transporters of both organic substrates and microbial inocula are anecic earthworms (species that feed largely on the soil surface, defecate deep in the soil, and construct extensive tunnels in the rooting zone) (Lavelle 1988).

Building the Fabric of the Soil

Kubiena (1938) at the University of Iowa was the first American to carefully document the contribution of different invertebrates to the microstructure of the soil. Pawluk (1985) at the University of Alberta documented the invertebrate biogenic characteristics of all natural Canadian soils. Thin‑section photography in Pacific Northwest forests (Boyle and Moldenke unpublished) has revealed that the humus, A‑layer, and much of the litter and B‑layers are mostly invertebrate feces. Living between (and sometimes within) these feces are the roots, microbes, protists, and invertebrates, together with scattered undigested dead plant and animal material (Bal 1982; Rusek 1985). The content of the feces can be largely organic (litter consumers), inorganic (deep soil endogeic earthworms), or heterogeneous mixes (mobile millipede coprovores). Only in soils with very low organic content (deserts) or frequent disturbance (annual row crop agriculture) do invertebrate feces fail to be the predominant structural elements (Pawluk 1985).

The size, shape, and percentage of organic content in the feces control many physical and chemical properties of different soil types (Martin and Marinssen 1993). A matrix of large fecal pellets promotes macropore formation, facilitating aeration and infiltration. Dense mixes of small feces promote wettability and refugia from predators. The majority of soil‑dwelling arthropod species probably feeds principally upon the fungi growing on the surface of fecal pellets (Anderson 1975). As they graze on the hyphae they often physically abrade the pellets themselves, exposing additional nutrients to microbial attack.

Each time an annellid worm or arthropod ingests solid food it secretes copious carbohydrate lubricants (mucus). These carbohydrates do act as energy sources for intestinal microbes, but in most instances the mucus ultimately surrounds and embeds the fecal material as well (Edwards and Lofty 1972; Lee 1985). When the content of the fecal material is largely inorganic, as it is for most deep‑dwelling earthworms, this input creates a branching "drilosphere" throughout the soil of additional resources and subsequent microbial growth (Bouchet 1975; Barois and others 1993).

The feces of earthworms are continually reingested to exploit the nutrients filtered from the soil solution by microbial populations. Thin‑section soil micrographs unambiguously document that invertebrate feces dominate forest soil structure, but it is the largely unstudied rate of re‑consumption of these feces that drives the turnover and availability of the embedded nutrients.

Microbial Succession

There may be as many as 40,000 kinds of microbes in a single teaspoon of soil (Tiedje 1994). These different microbes exhibit a wide variety of chemical capabilities (Hawksworth 1991; Fox 1994; Palleroni 1994). The vast majority of microbes is inactive at any given time; only the subset of species capable of using the actual specific chemical composition of resources currently available is metabolically active (Lavelle 1994). As such microbial species grow, and consequently change the chemistry of the remaining resources, the fauna consume them (presumably after most of the microbes have reproduced), permitting other microbial species to succeed them.

Although the rate of microbial grazing by soil fauna in different soil types remains largely unquantified, such grazing rates may be very high (Visser 1985; Coleman and Crossley 1996). More than 200,000 individual fungivorous arthropods can inhabit every square meter of conifer forest soil on average (volumetric measurements of soil organism density are confusing, since abundance drops logarithmically with depth within the top 10 cm; Petersen and Luxton 1982; Moldenke 1990). The most frequently occurring fungivorous arthropods are oribatid mites, endeostigmatid mites, and springtails. The fungivorous springtail, Onychiurus, is the prime determinant of the fungal community composition of a conifer soil in Scotland, (Newell 1984). By preferentially consuming a fast‑growing species, it keeps fungal diversity high; when Onychiurus is removed, a single fungal species predominates, representing 90 percent of the fungal biomass.

Microbial Grazing and Consumption of Microbes

When nitrogen, either as N0₃ or NH₄ions, is released as an enzymatic byproduct into the soil solution, how does it enter a root? Since the carbon:nitrogen ratio of bacteria and fungi is very low compared to that of the plant material and organic matter, soil microogranisms act as accumulators (“sinks”) of nitrogen. Bacteria growing within the rhizospheric high-carbon mucigel have ample supplies of labile carbon. They require nitrogen and other soluble nutrients that are being drawn toward the root as water enters the root's vascular system.

A large fraction, perhaps most, of the nitrogen in growing plants enters a root as the result of fauna consuming the microbes in the rhizosphere (Moore and others 1988; de Ruiter and others 1993). As rhizosphere‑grazers (the protozoa, rhabditid nematodes, bacteria‑ and fungus‑feeding mites, and springtails) scrape bacteria and fungi off the root surface, consume them, and defecate (Kuikman and others 1990). Some of the nitrogenous byproducts in these feces can penetrate the disturbed microbial sheaths of the roots. As long as the populations of microbial grazers do not become too dense, they stimulate mycorrhizal growth and exoenzymic activity (Finlay 1985).

Two studies substantiate this interpretation. First, a culture of fungi, bacteria, protozoa, and nematodes from native grassland soil was incubated in a microcosm, and the rate of nitrogen mineralization was estimated by intermittent drainage of soil water (Hunt and others 1987). When biocides were added to kill the protozoa and nematodes grazing on microbes, mineralization rate decreased by more than 82 percent.

Second, normal soil profiles of English oak forests containing natural populations of soil microbes and protozoa were reconstituted in the laboratory (Anderson and Ineson 1982; Anderson and others 1985). Nitrogen mineralization was monitored in the leachate, and a base rate for microbes in the absence of any plant roots or arthropods was established. Addition of a growing oak seedling did not affect the rate. The oak root cannot directly cause mineralization, and it can act only as a passive sponge. Addition of an arthropod shredder greatly enhanced nitrogen mineralization. The increased rate of nutrient mineralization was directly correlated with the biomass of shredders added, relatively independent of specific taxon. Surprisingly, synchronous addition of both oak seedling and arthropod shredders resulted in no detectable difference from the basal rate of nitrogen leaching with only microbes present. The “missing” nitrogen that should have leached in their combined presence manifested itself in an increased rate of seedling growth (more than four-fold higher than in the absence of arthropods).

INFORMATION GAPS:

A wide variety of forestry practices are known to have both direct and indirect effects on the soil arthropod fauna, particularly the two most discussed practices of cool under-burning and thinning. Therefore appropriate assay protocols must be developed to quantify both the immediate effect and the time-course of recovery on the soil arthropod fauna. There are appropriate foci for concern and analysis at both the individual species level as well as the total functional group level.

Top-down approach for monitoring:

From the aspect of ecosystem functioning there are two types of species of importance: (1) keystone species (one species whose removal can cause a disproportionate alteration of a critical biological process); and (2) biogeographically unique endemic higher taxa (the equivalent of a platypus or a kangaroo) that occur nowhere else in the world and represent unique genetic solutions to ecological challenges faced by the fauna through evolution. The importance of a keystone species is obvious; however, experience throughout the world has demonstrated that the identity of such keystone species is generally only determined after they have been lost.

Biogeographically unique endemic higher taxa (BSHT; defined in more detail on page 38) are important on a scale of global biodiversity, but also of special significance in our forests since they are often either very abundant (e.g., Pterostichus (Hyperphes) predaceous ground-beetles) or significant ecosystem regulators (e.g., Eulonchus pollinating flies).

Highlighting Biogeographically Significant Higher Taxa: analysis of

species distribution patterns in the literature

A species may be a part of a widespread group of species, either as a local endemic species within a speciose group or as a widespread species found in many biogeographic realms in North America or the World. In contrast, other species are representative of monotypic genera or monotypic families that are known from no other region of the world. This second group of endemic taxa (= unique) can be monotypic or locally speciose. Analysis of the overall range of a higher taxon (subgenus, genus, family…) can determine the uniqueness of the arthropod fauna of the Pacific Northwest. Individual species within this unique category that live nowhere else on earth (biogeographically significant higher taxa (BSHT)) are our endemic arthropod equivalents to kangaroos and coeloeanths. On a global scale, such unique taxa assume a far greater significance than higher taxa distributed in many additional regions of the world.

Within the 150,000 mi² region of the Northern Spotted Owl it is clear that there are a finite number of Vancouveran Bioregion (the region of cool mesic forests from Monterey Bay to Anchorage, AK) endemics. For example, examination of the tentative list of unique bee taxa (=BSHT) within the range of the Northern Spotted Owl reveals a list of approximately 50 taxa; this number includes both (1) paleoendemic relicts as well as (2) neoendemic radiations (Appendix 3). Considering that the broadly defined Vancouveran Bioregion has been isolated for a rather short time geologically-speaking, there are actually a good deal more unique endemic bee taxa than one might expect. However, as would be expected, the great majority of resident bee higher taxa are widely distributed.

A list of all the Vancouveran uniquely endemic (=BSHT) taxa (not necessarily rare taxa) discovered thus far in this introductory wider search of all soil and CWD-related groups is appended (Appendix 3a). In this broader study of 5 groups (Oribatida, Hymenoptera, Coleoptera, Diptera, and Collembola) it is clear that there are relatively few Vancouveran-endemic higher taxa as a general rule. Only about 200 taxa are considered potential candidate BSHT from the thousands of higher taxa within the three most diverse insect orders (Appendix 4). The groups containing the most Vancouveran

Figure 3. Zootermopsis angusticollis, the dampwood termite, represents one of the most primitive termites in the world. It is found exclusively within the Sierra/Cascade Mountain axis of North America (illustration from Audubon Field Guide 1998).

BSHT are the Andrenidae (solitary bees: 50% of the total Hymenoptera listed), Carabidae (ground-beetles: 25% of the total Coleoptera listed), and Staphylinidae (rove-beetles: also about 25% of the total Coleoptera). Many of these BSHT are not rare within the Pacific Northwest. The long and unique association of these taxa within the region has permitted many of them to become very successful. Some of them are so successful that they are even potential pest species (i.e., Zootermopsis, damp-wood termite, Figure 3). This within-region abundance does not render them any less unique on a broad scale, however. These BSHT have a far greater biodiversity value globally than any other Pacific Northwest species. It is fortuitous, from the management point of view, that so many of them are abundant and not presently threatened.

Action recommended:

1. Carefully work through all orders of arthropods and confer with taxonomic specialists to document lists of BSHT. This preliminary step will require about two more years of our effort.

2. Transform every member of the resultant list into a state of political “acceptability”; these are, after all, our truly unique species on a global scale. Assemble the list into a web-based book (with pictures of everyone) of all our unique Vancouveran Bioregion taxa (see Appendix 3). This task is needed to focus the attention of the entire scientific community on the phenomenon of regional uniqueness. If there are keystone species in PNW forest ecology, they are likely to emerge from this special group (BSHT) because of their long evolutionary history within the region.

Predicted Outcome: Many of the species within this group will be abundant enough not to merit concern from a sensitive or endangered species point of view. The exceptions are liable to be taxa strongly associated with old-growth, some of which were previously enumerated by Lattin and Moldenke (1992; Opler and Lattin 2001). Since there are few, if any, formal mandates for management to attend to BSHT, despite Interagency Committee on Biodiversity Ecosystem Management Plan’s (ICBEMP) own inclusion of “long-term evolutionary potential” as one facet of ecosystem integrity, it is critical to develop healthy working relationships between management and the academic and conservation communities.

Why monitor at all? Why use arthropods?

How do you define the appropriate indicator/focal taxa?

The prime emphasis on “arthropod indicator analysis” belongs at the level of the total fauna. In order to contrast different types of forestry implementation practices, the appropriate level of resolution is “Which practice minimizes the detrimental effects on the whole ‘soil’ arthropod community?” The critical descriptors are therefore: (1) total density of arthropods, (2) total biomass of arthropods and (3) total species richness of arthropods (each broken down into broad functional guilds). The required resolution of species richness is at the morphospecies level (i.e., predaceous beetle speciesA, B, C,…; parasitic wasp speciesA, B, C…; etc.). Forestry practices (such as clear-cutting, underburning and fuel removal) may initially decrease total density of soil-dwelling arthropods by 75-90% and decrease species richness by more than 50% (Moldenke and others unpublished data). If there are biologically significant differences between different practices they will certainly show up at this coarse-scale of resolution; it would be better to monitor a wider variety of forestry practices at this morphospecies-level than it would be to quantify far fewer instances at the taxonomic expert-level of resolution. Though such immediate management effects are usually transitory, decreasing the amplitude of reduction in biomass and richness would be preferred as a general rule (Niemela and others 1993).

We have been addressing the question of appropriate management practices involving arthropods with numerous scientists. This discussion was the result of a poster we specifically prepared on the subject for presentation (August, 2000) at the International Congress of Entomology in Brazil; the theme of this meeting was “Entomologists preserving biodiversity in the new millennium.” This poster is currently on display in the USFS Portland laboratory. Of course, nearly everyone had his/her own unique perspectives on these topics. Moldenke would summarize their frank and provocative responses as follows:

A) Originally, with the exception of only a few people, the response was: “Why bother to use arthropods as indicator species? It can never be cost-effective, even if anyone knew enough to do such management. Let people concentrate on vertebrates and rare plants as imperfect surrogates for the entire ecosystem including the insects.” (But of course they always made exceptions for their own particular pet bug of choice.)

B) However, once into the subject, the conversation always turned to a different theme: “since arthropods directly or indirectly run nearly all the ecosystem processes in all terrestrial ecosystems, you can’t manage any ecosystem effectively unless you know the important players and what they do.” These discussions focused on process (not species lists), that is, how to determine who the important species are, and how to monitor them. The consensus conclusion was that abundant species are the most likely candidates for “important” species, but that theoretically one couldn’t rule out the possibility that a keystone species (that was simply not abundant/apparent under usual circumstances) could be critically important as well. However, rare species with critical roles in process control cannot normally be procedurally detected until the ecosystem has already been radically altered. No one was familiar with any coordinated attempt to analyze the arthropod species important to forest ecosystem processes anywhere in the world. The general conclusion was to proceed in the normal method of science, namely, to assemble the information on the ecology of each species incrementally, and slowly build a more comprehensive knowledge base of the system and how it works. Therefore, in summary, there is no miracle cure. So, the consensus was that arthropods are important, but true system knowledge will only come with better understanding of their individual natural histories.

C) These conversations would always finally conclude with the idea that the arthropods important to ecosystem processes are the abundant species, and it is far easier to monitor abundant species efficiently and cost-effectively than it is to monitor rare species.

Still, all of the specialists consulted had the innate reaction that we should also be talking about rare species when discussing biodiversity. But, rarity did not usually become the true focus of the conversations; however, uniqueness did. Speaking with international scientists helped remove us from our Northwest-centric point of view. Individual rare species, especially if closely related to abundant and widespread species elsewhere, did not comprise anyone’s top priority for conservation concern — however, biogeographically unique higher taxa certainly did. There was universal agreement that it was necessary to determine which higher taxa (subgeneric-level or above) were endemic to the Pacific Northwest. If Noah had only an economy-sized arc, these would be the taxa he would focus on. Who are the kangaroos and coelocanths of the Northwest arthropod world? Though biogeographically unique higher taxa are not necessarily rare or “sensitive”, if these regionally endemic groups were rare as well, then they surely should have priority over any other concerns. Though the vast majority of these higher taxa unique to the Pacific Northwest may be, in fact, relatively abundant (as the current literature databasing has shown) – the consensus was that a comprehensive listing of all these BSHT endemic taxa was an absolute first step. (For instance: there is 1 genus of oribatid turtle-mite; 47 taxa of ground beetles; and 41 groups of bees endemic to the Vancouveran Bioregion; etc.; see Appendix 3).

As Northwesterners, we seem to discuss at length the issues of old-growth and salmon decline; the scientists Moldenke spoke with generally felt that it would be better for entomologists to prioritize other issues with completely independent effects. Since the politics of spotted owls has changed the trajectory of forest stand management already and for the foreseeable future, the basic remaining issues of old-growth management deal with forest patch size and establishing a defined group(s) of invertebrates that are representative of old-growth conditions and can be monitored to assay the “ecosystem health of old-growth”. These are issues for which we are currently designing protocols (see studies by Bruce Marcot unpublished, Gary Miller and Moldenke unpublished). Since it was not within the purview of this contract to examine aquatic arthropods, the issue of whether management for salmon is compatible with management for arthropod taxa remains to be examined. Nearly all of the mitigation efforts designed for salmon would have beneficial effects on riparian ground-beetle diversity. But ancillary wetland habitats not considered by fish biologists (i.e., bogs, marshes, etc.) perhaps sustain even higher species richness and endemicity of terrestrial arthropods (e.g., the majority of species of ground-beetles [Carabidae] are associated with microenvironments that are basically “riparian”) and need appropriate attention as well (Lindroth 1961-69; Thiele 1977; Stork 1990; John Richardson [1], in preparation).

So what are the aspects of arthropod biodiversity that the international entomological consultants determined to be the most critical?

(1) Make biodiversity useful, not confrontational. Focus on positive actions already being taken. For instance, in habitat types already extensively impacted by humans (e.g., the Willamette Valley), or in ones rapidly being lost (e.g., ocean coast), use the enormous richness of arthropods to show how important even small relict stands are in conserving overall biodiversity. A small reserve comprising several acres of threatened habitat may indirectly provide habitat for hundreds of native arthropod species. Make managers aware that in relict stands, environmental heterogeneity is of paramount value, and that the entire stand should not be managed for the benefit of one single charismatic vertebrate or flower. Use the ecological interactions governed by arthropods as a high-profile aid in promoting the better-known diversity of the flora (i.e., specialist bee pollinators of sensitive plants). Focus on the interdependence of ecosystem foodwebs.

(2) Though focussing on individual sensitive species is necessary, educate the public about the broader issue of habitat conservation and focus efforts on particular species that do not immediately threaten the general public (and land managers) economically. For instance, our studies have shown that there are liable to be many truly rare and endemic alpine species of arthropods (see Appendix 3). Unless ski developments impact certain critical microenvironments, preservation strategies for those species will not threaten anyone economically. Whether rare and endemic arthropod species live in out-of-the-way places or not, it is important to stimulate the gathering of better and more extensive information on all infrequent taxa (arthropods and non-arthropods alike) as well as to educate the public on the inter-relatedness of all species within an ecosystem. The goal of generating useful new information is far more likely to be achieved when the atmosphere is non-confrontational. Far more rare/sensitive species will be preserved through general education and non-confrontational monitoring than will ever be preserved through legal confrontation involving only arthropods.

What are the most pressing short-term and long-term information needs?

The task at hand is to prioritize the potential information needs. The most important of these gaps in knowledge can be grouped into two general topics: a) short-term direct effects of management practices upon the arthropod fauna; and b) the nature of monitoring and longer-term composite effects of multiple management practices.

The first category of concern acknowledges that all anthropogenic activities within the forest, ranging from comparatively benign hiking trails to intensive timber removal, leave detectable imprints upon the structure of the forest. Not only are these imprints expressed in terms of physical changes, but more significantly they are expressed in terms of altering the composition (the relative abundances of the species) of the forest ecosystem. Many such changes are physically localized, of limited duration and affect only a small percentage of the species present. However, changes that do not initially seem to directly affect dominant plants or vertebrates may be overlooked even though their effect on the arthropod community may be significant. (Example: a permanent hiking trail through a meadow often provides the compacted bare soil nesting-site requirement that allows dozens of species of native bees to colonize, promoting reproduction in numerous insect-pollinated species.)

Other concerns relative to arthropods involve much more obvious effects to the ecosystem. Management practices such as timber removal or prescribed burning (and the opposite, fire suppression) have system-wide effects on all the resident taxa, probably without exception. Both the amount of change in species abundance, diversity, and ecosystem interactions, and the time course of the change vary with the individual taxonomic groups of species. Such changes are neither necessarily “good” nor “bad” on the ecosystem-scale; they merely represent different assemblages of resident species. One stage of forest succession is no better than another in terms of ecosystem processes; value can only be assigned relative to a single species and whether its populations are facilitated or diminished. Old-growth may be “good” for the spotted owl, but it is “bad” for numerous other species which would prosper under other conditions.

Likewise the total number of species present in a community is not always a proper gauge of its “health”. Within a generally west-facing watershed, on a localized plot basis, the west-facing slopes will generally be characterized by the highest species richness (of either plants or insects, they are often correlated); the north-facing and south-facing slopes within that watershed will support far fewer species on a localized plot basis. No one of the three slope-faces is any “healthier” than the other, even though the total number of resident species may vary widely between them. This commonly observed pattern occurs because many of the species inhabiting the north-facing and south-facing slopes within such a watershed are not found on the opposite slope-face; however, nearly all the species found on the west-facing slope are found on the other slope-faces as well (for the Klamath/Siskiyou Region see Wilson 1988). Though all three slope-faces are equally “healthy” (in spite of different species compositions), a management practice that alters, for instance, the soil organic matter content of the north- or west-facing slope—which results in a species composition characteristic of a south-facing slope within that watershed—could legitimately be considered decreasing the “health” of that community. This would be so, especially, if the effect is long-lasting.

Disturbance (in many forms) is a natural process; without disturbance the landscape would be far less heterogeneous and species richness would drop precipitously. Though it is both convenient and conventional to ascribe any changes to the status quo as “detrimental”, as ecologists we must avoid this pitfall. The principal priorities are to document whatever changes take place in the community of organisms and to determine how long the changes to individual species persist. This is not a simple task since the status quo is itself a moving target. (No site is immune from either long-term succession or annual climatic cycles.) Nevertheless it is an important responsibility.

So why focus on arthropods? Plant response is generally far easier to monitor than the response of arthropods. Whenever silviculturists detect an unexpected or a detrimental plant response to a particular management procedure, a potential concern is raised. Further study is elicited and the procedure is effectively modified. However, a problem arises in the natural history of forest plants. All the potential indicator tree species are long-lived, and as such, they are not particularly sensitive to change. Almost all arthropods are characterized by shorter life cycles and, of course, have been used as sensitive indicators of change under many ecological circumstances. Changes to the bacteria and fungi would have the most fundamental effects on the ecosystem, but they are extremely hard to detect and quantify and often arthropods have been documented to serve as acceptable surrogates. There is widespread agreement within the community of ecologists that arthropods can and should be used when appropriate to quantify change and estimate the time course of “recovery” of fundamental ecological soil processes in particular. Most types of anthropogenic change can be measured effectively by direct measures or indirectly by using indicator species of plants or vertebrates. Due to the unique temporal and spatial scales involved with soils, arthropods may be the most efficient indicator taxon (Benckiser 1997).

The questions become, then, which particular management practices are appropriate to assess with arthropods, and which practices have the highest priority? Any practice with an effect on either nutrient cycling or soil water-holding capacity will have the most profound effects in any ecosystem. Any scientific review of the multiple functional roles of soil- and litter-dwelling arthropods will highlight numerous causative determinants elicited by arthropods in all the world’s forested ecosystems (Fitter and others 1985; Shaw and others 1991; Lavelle 1994; Moldenke and others 1994, 1999; Coleman and Crossly 1996; Moldenke 1999).

Two general management practices will have the most significant and farthest-reaching effects in PNW forestry: timber removal and prescribed fire. Regardless of the logical and pertinent arguments that compare both processes to natural disturbance events, it is recognized that many aspects of both are not similar to natural disturbances. In our opinion, and the opinion of every member of the FEMAT advisory panel, all of the innumerable variations on timbering, thinning or prescribed fire practices will significantly affect both the community of soil arthropods and the soil processes in which they participate. The principal reason to document the effects of contrasting management practices that could be applied for the same purpose (i.e., thinning) is to compare the alternatives with respect to: 1) the amount of the biological effects, and 2) the length of the recovery times. These two elements should then factor into the overall considerations of economics, practicality, and timing that determine under which specific circumstances a particular variant of a management practice would be used. Instances of delayed conifer regeneration after cutting in stands throughout the region are being analyzed by Harmon [14] and others; it is likely that these instances are tied to altered soil biological properties.

In practice, silvicultural decisions are already made largely on the basis of soil properties; silviculturists are justly concerned about long-term site productivity. However, assessments of potential damage to the biological properties of the soil ecosystem are usually qualitative and subjective; it is likely that limited use of indicator soil arthropods would permit far better calibration of the decision-making process. The soil is an incredibly diverse environment; its diversity is central to its many basic functions. A management priority is to simplify any process; thus, it can be modeled and predicted more easily. However, the soil ecology literature is replete with examples – don’t oversimplify. Monitor the biological responses to aid in decision-making. In agriculture, though soil fertility is often correlated with physical and chemical factors, the processes involved in crop growth are actually caused by biological aspects of soil ecology (Benckiser 1997). While arthropods, per se, are seldom direct causes of plant growth rates, their regulatory roles (and hence their indicator value) on soil microbes are indispensable (Anderson and others 1985; Finlay 1985; Fitter and others 1985).

“We can’t afford to do something new and expensive.” Leaving aside the obverse argument that holds that you can’t afford not to do it – since there is no immediate way to come to a solution. What disturbs ecologists most about many forestry practices is that there is an almost complete lack of data on the relative effects on basic ecosystemic processes. Though a resultant change in the soil community assemblage of organisms is merely a change, and is theoretically neither “good” nor “bad”; there are some types of changes in the soil that are justifiably nearly universally considered “bad”, i.e., erosion, compaction, loss of water-holding capacity, nutrient volatilization, etc. Though most of these soil properties can be measured directly, what is more important is how long it takes for the system to recover. Under these circumstances it is far more effective to measure biological properties. It is always crucial to bear in mind that arthropod monitoring is appropriate when the process concerned cannot be monitored more effectively with another taxon (i.e., plants). In practice, this situation often applies to soil processes (where the fungi and bacteria are intractable) or under temporal constraints where longer-lived species will not function adequately.

There are 2 different types of recovery: short-term quantitative and long-term qualitative (the second is discussed in Part II). Of paramount immediate importance is the short-term response in overall biomass and species richness. For instance, it is more critical to discover that fall-burning decreases the biomass, abundance and total richness of total soil microarthropods far more than spring-burning (Christine Niwa [2] and others, unpublished data) than it is to learn just which species are affected. If there were changes of different orders of magnitude in the entire biological community, then these normally would take precedence over the responses of individual species. Any technician can count total numbers, weigh total biomass and estimate species richness at the resolution of morphospecies. Information on total morphospecies richness and biomass is the level of information that is urgently needed in most circumstances for comparing alternative management practices. Thus, the only practical method for testing many different variants of a management practice, or for testing the effects of the same practice under different forest conditions, is the “quick and dirty” method that quantifies short-term community level phenonema. This will suffice to answer most comparative analyses to a reasonable level of confidence. Although there are few examples within the Pacific Northwest, the literature analysis has documented numerous examples of this approach worldwide, especially in agricultural management (Paoletti and others 1993; Benckiser 1997).

Monitoring priorities for focal ecosystem processes

Three ecosystem processes involving soil-dwelling arthropods each emphasizes the fundamental significance of soil arthropods in maintaining the productivity of ecosystems: 1) as both indicators and regulators of microbial nutrient transfers, 2) as trophic transfer for vertebrate species, and 3) as the provider of trophic sustainability through pollination. All three processes are integral to all ecosystems and are therefore of special importance relative to management concerns; they cannot be ranked in importance. There are of course many other linkages, but our consultants and we have chosen to emphasize these three.

1) As outlined below, a long-term management implementation of spring prescribed-burning will fundamentally alter the pattern of nutrient cycling. The USFS Fire Suppression Program has already fundamentally altered the natural biphasic system; spring-burning will alter it again but not back to the original scenario, as far as soil ecologists can predict (Bruce Caldwell [3], personal communication). The elimination of fire from the understory over the past half-century and consequent increase in litter and humus has resulted in population levels of soil-dwelling arthropods that presumably exceed previous levels in all but the most mesic older forests. Any reinstitution of fire will greatly decrease composite densities and indirectly decrease local biodiversity as well. Depending on desired outcomes, this severe reduction would be both unavoidable and, in fact, desirable if the goal is to mimic pre-settlement conditions.

However, switching the pulse of fire-released nutrient availability from the dry summer fire season to the general growth period in springtime will have numerous unforeseen effects. The very high levels of active fungal biomass in the soil in springtime will probably buffer the soluble nutrient loss from most forested sites sufficiently to avoid radical competitive changes between arthropod species. An increased ability to adsorb the soluble nutrients during the spring when the microbes are metabolically active would decrease the loss of nutrients to ground water and indirectly increase nutrient availability during the whole growing season by facilitating population densities of fungivorous soil arthropods. However, introduced weedy species are liable to increase dramatically because of higher available nitrogen concentrations, altering the fundamental composition of both the plant and arthropod communities. Forests with open understories (ponderosa or Jeffrey pines) and chaparral/scrub communities are likely to change fundamentally in both soil productivity and arthropod species composition. The most important factor limiting our understanding of such developing changes is the lack of current descriptions of the native arthropod community composition.

2) Likewise, the riparian network today may be very unlike the early historic condition. Elimination of beaver greatly changed the flow of water across the landscape, grossly altered the species composition of streamside plants, and doubtlessly fundamentally changed the abundances of both truly aquatic invertebrates and riparian-terrestrial species. For most of the range of the Northern Spotted Owl, we would maintain that we have a very imprecise concept of what these historic riparian plant and arthropod communities were like. For much of the landscape, low-order streams which today are narrow and free-flowing were once dozens of meters wide, and largely still. Though beaver are becoming more prominent, they will probably never regain their original landscape effect. We need to document what happens to biodiversity and energy flow over the long-term as beavers regain control as “ecosystem engineers”over a watershed.

In the short-term we need to carefully consider the often-opposing needs of both salmon and terrestrial birds/bats for the food resource provided by aquatic insects. The actual functionality of riparian buffer widths at maintaining diversity and population densities of both aquatic and terrestrial-riparian arthropods needs to be assessed in both the mesic forests of the North and the more xeric forests of the South (see research of Rykken and Moldenke, unpublished [12]). The linkages between aquatic invertebrates and vertebrate species should be of interest to managers. For example, salmon feed and depend on aquatic macroinvertebrates. Furthermore, according to the Oregon-Washington Species-Habitat Project database (O’Neil and others 2001), 196 species of amphibians, reptiles, birds, and mammals use aquatic macroinvertebrates as a food resource. For many of these species, this is a critical resource.

3) Pollination services are taken for granted. As ecologists, we often assume that whatever the circumstances, plants are being efficiently pollinated. However, many environments are pollinator-limited—even within environments richly characterized by pollinators, a large percentage of species lose out in competition to species with big showy flowers (Moldenke1975, 1976a, 1979b, c). Pollinator services are thus of concern in all environments. Within the Vancouveran Bioregion the different community types have rarely been characterized with respect to their pollinator abundances and behaviors, unlike those in California (see Moldenke 1975, 1976a, 1979a). Marcot and others (1998) estimated that, in the interior Columbia River Basin, invertebrates pollinate most of the rare or potentially rare vascular plants (66 percent), and about half of these plants (33%) are pollinated by solitary bees (Rathcke and Jules 1993; Bond 1994; Haynes and others 1996; Spira 2001). In this region, insects as a whole play vital roles in reproduction of rare flowering plants, whose viability depends on the presence of their invertebrate pollinators and dispersal mutualists.

Sunlight is critical for most pollinating insects because their flight metabolism requires this exothermic energy source. In mature forest stands within the range of the Northern Spotted Owl, cross-pollination (except for wind-pollination of the dominant tree species) is essentially inoperative (Moldenke 1976a, 1979b, and unpublished data) unless there is a gap allowing sunlight in to the forest floor. In the absence of sunny gaps, the perennial plant species (both shrubs and herbs) either clone vegetatively or those that have chasmogamous flowers self-pollinate (e.g., the mycotrophs) for decades at a time. In true old-growth forests with canopy gaps, sun-flecks provide for the only extremely limited pollination in the understory. Outcrossing opportunities for most flowering plants are restricted to the earliest stages of succession when direct sunlight is available and bees and other pollinator groups are very abundant. Some communities, like the oak savanna, have been so altered by overgrazing and fire exclusion that nearly all the native arthropod pollinators have been eliminated (William Stephen [4], Moldenke unpublished data). Ambitious plans to use fire to reestablish native savanna will succeed only if the pollinators become reestablished as well. It is quite possible that refuges for bee species exist on Forest Service land in southwest Oregon; only a careful inventory will tell. Pollination is a process which regulates genetic variation (and hence ecological amplitude and long-term adaptability) of plant populations. As such, it is as important to dominant understory plant species as it is to rare ones. Nearly all the dominant understory plant species are potentially insect-pollinated (Moldenke 1976a, 1979a).

PART II: Individual arthropod species and endemism (bottom-up approach)

Introduction:

Historically, the federal land-holding agencies have used the individual species approach to examine the nature of forest sustainability. Theoretically it is based on the presumption that, over time, all individual component forest species become to some extent coevolved to produce the forest processes observed today. Therefore, loss of any individual species is in some sense a loss to the integrity of the forest ecosystem. Though the original team of entomological consultants during the FEMAT process utilized the top-down approach almost exclusively (USDA 1988, 1994), they recognized that the natural focus of many interested groups of citizens was at the level of preserving rare species (Lattin and Moldenke 1992; see Appendix 2). Primarily from the rare individual endemic species point of view, the panel designated the Klamath/Siskiyou region to be of special concern within the region inhabited by the Northern Spotted Owl.

Gamma-diversity (the co-occurrence of species characteristic of different biotic realms) is highest in southern Oregon and northern California where the Vancouveran mesic forest, Californian Mediterranean scrub and the Great Basin cold steppe faunas are thoroughly intermixed. Though beta-diversity (co-occurrence of species from separate habitat types – i.e., wet and dry slopes) is very high throughout the entire range of the Northern Spotted Owl due to the mountainous terrain, it is especially high in the moisture-limited regions of southern Oregon and northern California, where the moisture gradient between North-facing and South-facing slopes is a prime determinant of the distribution of nearly every arthropod species. Alpha-diversity (the level of species richness within a single habitat type) generally increases on a global scale from North to South; again highlighting the Klamath/Siskiyou region within the area encompassed by the Northern Spotted Owl.

Within any defined region, the prime regulator of total species diversity is generally beta-diversity – whether defined at the habitat or microhabitat level (Gavrilets 1999). The known edaphic diversity of soil types within the Klamath/Siskiyou Mountain region accelerates the rate of local beta-diversity (Whittaker 1961; Daniel and others 1995). Within a forest ecosystem a much undervalued aspect of beta-diversity is successional stage; perhaps as much as 50% of the resident species of arthropods associated with the Forest Biome inhabit only the earliest 20 years of forest succession (McIver and others 1990, 1992; Moldenke and John Lattin unpublished data). Similar beta-diversity values based on forest succession are well-documented for birds (O’Neil and others 2001). Perhaps an additional 50% of the remaining arthropods associated with the Forest Biome are found exclusively within island-like habitats included within the forest mosaic (mountain meadows, bogs, avalanche scars, etc.). This feature of landscape diversity has never been directly quantified in the literature (although see Murphy and Wilcox 1986); our faunal analyses of carabid beetles and bees suggest these percentages apply to the total arthropod fauna. This interpretation was widely shared by the panel of experts convened during the FEMAT process. Surprisingly, the bottomline is that the majority of the Forest Biome insect fauna (soil-inhabiting or otherwise) is probably not associated with merchantable forest stands. The portion that is associated with merchantable forest stands needs one level of monitoring protocols, that portion not associated with timber-producing lands needs a different level of monitoring. (This last statement must be taken as a professional judgement, in need of subsequent analysis.)

Analysis of the 3 groups of “soil” arthropods best documented in the literature and personally best known to us (e.g., ground-beetles, spiders and bees) reveals that approximately 10% of the fauna would qualify as “sensitive” (conservatively defined most broadly to include all species that were either very limited in distribution or broadly distributed but very uncommon locally). This percentage is in line with the other taxonomic groups already dealt with in Survey & Manage (USDA 1994, 2000); the great majority of these species cited as “sensitive” are probably only apparently so, due to inadequate field surveying to date. Significantly, a majority of these sensitive species show correlated distribution patterns; there are hotspots of sensitive arthropod species (Rabinowicz 1981; Prendergast and others 1993; Izco 1998). The region with the most sensitive arthropod species is the region of greatest habitat conversion, the low-elevation oak woodland and prairie. Two other regions with a high concentration of sensitive species are both geographically extremely narrow interrupted diverse environments: the maritime terrestrial and the above-timberline.

Why is the Klamath/Siskiyou region unique? The legacy of the Klamath/Siskiyou mountain orogeny.

The combined interests of principally the World Wildlife Fund (conservation priority on biodiversity) and the US Forest Service/Bureau of Land Management (land management agencies under the Northwest Forest Plan) have centered upon the challenges of appropriate management for biodiversity while still providing for multiple resource values in the Klamath/Siskiyou Mountain region of northern California and southern Oregon. The biogeographic significance of this region has been recognized and well-documented for flowering plants and amphibians for several decades (e.g., Whittaker 1960, 1961; Stebbins & Major 1965; Wolfe 1969, 1978; Axelrod 1976; Raphael 1988; Welsh 1990; Daniel and others 1995; Bury and Pearl 1999; DellaSala and others 1999; Lang 1999)—for arthropods much less is known (see amaurobiid spiders Figure 4, Leech 1972), and what is known is not usefully centralized as mentioned previously.

During the early Miocene, global climate was significantly warmer than at present, and subtropical forests covered most of what comprised North America. The later Miocene was a time of global cooling and considerable mountain building in western North America. The ancient mountain region of the Klamath/Siskiyou was left as an island of mesic temperate vegetation, separated by vast expanses of xeric climate to the east and south, and cooler forest types on the low young mountains to the north. Many relict evolutionary lines were isolated in this area, and either swept away subsequently from all the rest of North America (like Sequoia), or found currently in amphi-continental distributions (like plethodontid lungless-salamanders). Additionally, parent-material chemistry is extremely diverse in the Siskiyou Range due to its unique geologic history. Subsequent climate changes (principally glaciation) and mountain building (principally the Cascades and the Sierra Nevada) have provided for the evolutionary radiation of many noteworthy taxa (e.g., Nebria see Kavanaugh unpublished data [13]). Perhaps the best known example is Leach’s (1972) studies on the amaurobiid spiders. Amaurobiid spiders are a very primitive group that is represented in North America by 83 species; 42 of these represent neoendemic radiation of a number of relict genera restricted to the Klamath region, while the entire rest of the continent is inhabited by 38 extremely widespread species (Figure 4).

Figure 4. Distribution of the species of Amaurobiidae (Araneae) in North America north of Mexico. More than half of the total diversity occurs only within the Klamath region of Oregon and California. The species occurring in the rest of North America are characterized by extremely wide distributions.

Figure 5. The distribution patterns of arthropods inhabiting the Pacific Northwest.

As a result of the recolonization of wide areas of the Pacific Northwest after glaciation, most of the arthropod taxa encountered in the Pacific Northwest have extensive distribution patterns (Figure 5). There are 5 principal distribution ranges of Pacific Northwest arthropods:

A: all montane western USA and SW Canada (30% of total arthropod

fauna)

B: transcontinental sub-boreal Canada, northern USA and forested

western United States (20%);

C. Cascade Range and Sierra Nevada (20%)

D. xeric western USA and northern Mexico (15%)

However, there are a significant number of arthropod taxa restricted to just the Vancouveran Region (10% of total species), with paleohistoric ties to the Klamath/Siskiyou ancient mountain orogeny (see Appendix 3). It is probably a safe generalization to say that subsequent to glaciation much of the Alaska, British Columbia, Washington and northern Oregon region was recolonized largely without species radiations, but in the region of southern Oregon and northern California, which were basically unscoured by glaciers, adaptive radiation was often rampant.

Even if all groups of arthropods were thoroughly collected and monographed, there would still be a variety of issues about appropriate management of arthropod biodiversity. The principal approach we are taking in this report is to restrict our prime focus to examples of biogeographically significant higher taxa (BSHT). Once identified, the ecological roles of these taxa then have to be established and tied to the larger issues of biodiversity, perhaps as focal species. In delineating biogeographically significant taxa of the Klamath/Siskiyou region we are interested in determining which groups are found

Figure 6 (left). Grylloblatta campodeiformis lives primarily along meltwater underneath glaciers and permanent snowfields. It feeds upon thawing insects that fall into the frigid water. (illustration from Audubon Field Guide 1998).

Figure 7 (right). Agulla is often placed into an order all by itself (=Raphidioptera). They are found only in the Mediterranean Region and the Pacific Coast of North America. Though the newly hatched young feed on pollen, all the other stages are ferocious predators (illustration from Audubon Field Guide1998).

nowhere else on earth (or in very disparate places)? Living within the Vancouveran Bioregion and nowhere else in the world, does not imply that these species are necessarily rare within our region.

The classic examples of BSHT in this region are the palaeoendemics: Tricholepidion (Thysanura), Grylloblatta (Grylloblattodea) (Figure 6), Agulla (Raphidioptera) (Figure 7), Bequaertomyia (Figure 8), Deuterophlebia (Figure 9), Tanypteryx (Figure 10), Eulonchus (Figure 11) and Cryptocercus (Blattaria). However, additionally, what is necessary, is extensive listing of all arthropod taxa tied to the unique Klamath/Siskiyou paleogeography. So we are looking for examples from library research of the following types:

a) genera (or other higher taxa) found exclusively in the Vancouveran region and nowhere else, either ancient relicts or possibly recently locally adapted taxa. These taxa can be:

(1) rare and localized (like some cave-dwelling examples, the millipede

Aprosphylosoma), or

(2) common and even pestilential (like the damp-wood termite, Zootermopsis (Figure 3)) – just so long as they are uniquely historically associated with the Vancouveran Region.

Figure 8 (left). Bequaertomyia jonesi is a very unique fly without any close relatives. Most recently it has been placed into the family Pelecorhynchidae. It is large and showy, but extremely rare (illustration from McAlpine, 1987).

Figure 9 (right). Deuterophlebia inyoensis is a representative of a small genus of endemic cold-water flies that resemble nothing else in the world. Notice the extremely reduced head, extremely elongate antennae and fan-shaped wings; if anything, the larvae are even more unique than the adults (illustration from McAlpine 1987).

Figure 10 (left). Tanypteryx hageni is a high-altitude dragonfly whose nymphs are largely terrestrial. It’s only close relative inhabits the Himalaya Mountains (illustration from Audubon Field Guide 1998).

Figure 11 (right). Eulonchus tristis is a keystone pollinator of several dozen genera of plants on the forest floor. It is also the parasite of the commonest and largest invertebrate predator on the forest floor, the trap-door spider Antrodiaetus (photo by Moldenke and Ver Linden).

b) genera (or other higher taxa) with wider disjunct distributions, presumably the result of:

(1) mesic temperate forest disjunction by Miocene drying (like the

Atypoides (Antrodiaetidae) trap-door spiders),

(2) recent immigration of Palaearctic taxa across Beringia to the PNW

exclusively (like Tanypteryx dragonflies)

(3) sub-tropical forest relicts due to subsequent climate change (like

pitcher-plants and pipevine swallowtail butterflies), or

c) genera (or other higher taxa) with perhaps wider distributions but with centers of species diversification within the Klamath/Siskiyou (like the ground-beetles Hyperphes).

Some groups of arthropods have numerous examples of biogeographically unique Klamath/Siskiyou taxa, others much fewer, and other major groups none at all. For instance, within the Coleoptera there are numerous examples of many Klamath/Siskiyou associations within subgroups of the Carabidae, Staphylinidae and Curculionidae – but practically none within the equally hyperdiverse Chrysomelidae (only the relict Timarcha and the localized Pseudoluperus). Examples of biogeographically unique higher taxa listed in the Appendix 4, fall into two distinct groups. For instance, the examples of the Staphylinidae (Omalinae) are associated principally with the more northerly mesic forest types of Arcto-Tertiary origin (Moore and Legner 1975, 1979; Newton and others 2000), whereas in the Cerambycidae and Apoidea many are associated with the more southerly Sierran forests of Madro-Tertiary origin (Linsley 1961-63). It is too early to generalize about most other groups.

Within our preliminary list of biogeographically significant higher taxa (BSHT) from the Carabidae (see Appendix 4) 41 of the 50 cited genera/subgenera contain at least one species that is widespread and abundant (and hypothetically not in imminent danger of extirpation), and only 9 genera/subgenera that are definitely of potential concern for conservation. Nearly all of the genera that fall into this latter category (true of many of the other arthropod groups we studied as well), can be directly associated either with ongoing legally mandated monitoring in riparian zones or associated with old-growth (Lattin and Moldenke 1992; Lattin and Opler 2001). If the number of “sensitive” groups of biogeographically significant higher taxa is actually quite limited (true at least in this particular case, we don’t know if one can extrapolate to all groups), and if these groups can easily and logically become a part of some ongoing monitoring research, then they pose less of a financial burden to already mandated management procedures. For instance, monitoring could be done simultaneously for both ground-beetles and mollusks relative to testing the efficacy of riparian buffer characteristics.

Individual potentially sensitive species: are there landscape patterns of co-occurrence?

Of special significance, however, is the possibility that certain groups of sensitive arthropods are characteristic of particular microhabitats completely ignored in the system of set-asides based on more charismatic taxa (like owls and salmon). This may well be the case for certain groups (e.g., Andrena – solitary bees) that prefer habitats with minimal forest canopy embedded within the general forest matrix. To the extent that such habitats are characterized by entire sets of multiple species with limited distributions, once the relevant species are recognized the entire habitat type may become the appropriate focus for conservation.

Species with limited abundances or distributions which are not correlated with a particular set of microenvironments would still presumably comprise a list too numerous to manage for individually. The logical practical solution, suggested by most of our colleagues at the international meeting, would be to short-list only the BSHT species that are rare. Potentially rare species that belong to otherwise widespread genera should be associated with specific habitat types and jointly managed at the habitat-level of resolution.

The prime responsibility for management of rare arthropod taxa lies not in the actual survey protocols, but in the preliminary acquisition of pertinent information either previously published or in the knowledge of taxonomic specialists. This requires a multi-step process:

1) defining the lists of both: (A) biogeographically significant higher taxa (BSHT) encompassing all arthropod groups (i.e., Appendix 3 & 4); and (B) individual potentially rare/sensitive species;

2) assembling pertinent data on geographic distribution, abundance and habitat preference. Based on the already assembled listing, only a short list of biogeographically significant higher taxa will be found to be sensitive as well (either rare or locally endemic);

3) filling in the information gaps for both the lists of sensitive biogeographically significant higher taxa and the list of potentially rare/sensitive species through an organized field-collection inventory program analogous to the process that generated the Franklin & Dyrness (1973) plant association database. (For instance, habitat preferences are already known for nearly all the bees and ground-beetles that would be on such a list. There is likely to be a sizable percentage of other arthropod taxa that would be imprecisely known.)

The point is, that without an authoritative baseline database on the abundance and distribution (especially of individual biogeographically significant higher taxa), the issue of which arthropod taxa ultimately deserve special attention cannot be addressed. Such a database could be refined and continually updated in the future, and the resolution even ultimately extended to the species-level. We hope to provide the conduit for assembling such a long-term database on the web pages of the OSU Department of Entomology (www.osu.orst.edu/dept/entomology/moldenka/; with individual entries associated clearly with specific collaborators). There is no question that the biodiversity of this region is quite special in any number of aspects. The time has come to actually document the hierarchy of special features. Seldom is there such an opportunity for conservationists, museum scientists and management to actually work together in an integrated and positive fashion.

How the USFS/BLM/USGS/etc. might use this information depends upon continual negotiation among the various interest groups. However, whatever action will be taken, should be dependent upon the best database possible.

Action necessary: Preparation of a species list of all arthropod species in the PNW with annotation relative to range, abundance and habitat choice (when known). Subsequently analyze all species into component groups based on a concern for rarity or threat. Only if this introductory analysis is done in a comprehensive fashion for the entire Vancouveran region will the individual interested organizations (both public and private) be enabled to prioritize their efforts. The archival information for the 6 unrelated groups we have examined in detail (bees, ground-beetles, Collembola, spiders, long-horn beetles, scolytid beetles) does begin to permit an enumeration of potentially sensitive taxa and their categorization into habitat types as well. This process has worked best for the bees, with which Moldenke is most familiar. Equivalent analysis would depend far more heavily on additional subsequent consultation with specialists for the poorly-documented Collembola, for instance. Many apparently rare/sensitive species within the Vancouveran bioregion are correlated with certain habitat types.

1) MARITIME: The list of endemic maritime terrestrial arthropods of all groups will be very limited. Most endemic species will turn out to be widespread along the Pacific Coast from Canada to Baja California. Because of this widespread characteristic of maritime terrestrial endemics, there will be a very short list of both rare and localized taxa (presumably ones requiring specialized microhabitats deserving of immediate mitigation attention). Collaboration with Dr. George Poinar (emeritus professor, OSU), an eminent zoologist with extensive interests in coastal arthropods, would be a logical prerequisite to produce a comprehensive illustrated guide to the natural history and ecology of all the endemic maritime shore arthropods of the Pacific Northwest. Maritime terrestrial habitats are the most anthropogenically-impacted habitats in the world; even formerly widespread abundant taxa are now often imperiled. Even in the “unspoiled” PNW, many of the most dominant maritime taxa are now introduced species.

There are 10 species of terrestrial maritime-associated bees. The in-depth bee distribution analysis (Appendix 3) reveals that 100% of the maritime-associated bees are widely distributed throughout the North/South extent of the coastal habitat. Though many may be of localized importance for the efficient pollination of the flora in these communities, no one species would be selected solely for special conservation concern. In the long-term, it is the entire fauna of this community type that is in serious jeopardy; the appropriate level of resolution for mitigation is the community type (e.g., coastal bluff, salt marsh), not the individual species (though one needs to have a full enumeration to monitor how efficiently the community mitigation is being applied). For instance, management for the endangered silver-spotted fritillary (Speyeria zerene hippolyta) along the coastal bluffs should take into account the habitat requirements of the pollinator fauna of the host-plant (Viola) as well.

2) ALPINE: The list of endemic alpine species of all arthropods will be relatively extensive. Perhaps a majority of them will be categorized as rare to very rare from literature databasing. Although perceived rarity may usually be due to an under-represented collection bias in difficult-to-access locations, very careful attention needs to be given to distinguishing their microhabitat associations. The alpine zone consists of many very different microenvironments/floras (Sawyer and Thornburgh 1977; Nancy Diaz [5] unpublished data); some of these environments are far less frequent and more readily impacted by current human activity than others. However, there will probably be very little of this necessary microhabitat information available in the literature. Consultation with specialists and significant amounts of new field observation will be necessary to determine which species, locations, and environments might need the highest category of attention.

There are 59 species of exclusively alpine bees in the PNW. Appendix 3 demonstrates that 76% of them (limited to the species restricted to the alpine zone) are categorized as infrequent, rare or very rare. This same pattern is found in the spider and beetle analyses, though with not quite as high a percentage of rare species. This difference in percent rarity is presumably due do the heliophilic requirements of bees and the truly insular nature of their habitats, whereas forest-dwelling alpine/subalpine beetle and spider endemics are somewhat less fragmented into separate populations. All of these habitat types are subject to high levels of natural disturbance (i.e., avalanche, landslide, rodent-tilling, climatic variance), and population levels of many of these less-frequent species are liable to undergo drastic localized population density shifts quite naturally. Frequency of occurrence in expected habitat type, rather than an average abundance, would thus be a more useful measure for many alpine species. (For instance, in alpine pollination studies conducted in 2001 on Mt. Hood, Mt. Ashland, and Dutchman’s Peak (Moldenke unpublished data), the commonest bees observed overall were each restricted to only a single meadow, even though their specialized host plants were abundant and widely distributed throughout each of the 3 peaks.)

3) SAVANNA/PRAIRIE: Species endemic to the oak savannas and prairies of the Puget Trough, or once presumably widespread and abundant in the Puget Trough and now nearly extirpated, will be the category of arthropods which will ultimately need to receive the most attention and subsequent active management.

Regardless of the speed and thoroughness possible for the databasing of species in this Puget Trough category, the unavoidable outcome is that certain groups of arthropods will be better known, or more readily identifiable, than most others. Analysis of the presence of these groups should be the best indication of the possible presence of the less tractable groups. Ultimately using these “focal arthropod species” could probably be the method most widely employed in the monitoring of watersheds throughout the region relative to agriculture, human population and industrial impacts.

A positive feature of the current listing of oak-savanna bees, for example, is the potential for entomologists to work directly with botanists in re-establishing this ecosystem type. Many of the plant genera that most ecologists would like to see re-established or strengthened have relied in the past on specialist-pollinator bee species. Many of these pollinator species visit only one species (or several closely related ones) in an area, and thus become extremely efficient pollinators of those plants when they exist in sparse populations; generalist-visiting bees are relatively poor pollinators under these conditions since their calculation of flight distance versus floral benefit renders them extremely inefficient within scattered populations. The rare bee, in this scenario, can be recognized by the public as a potential critical ecosystem player, not just another Latin name on a list. Some plant species of Lomatium, Sanicula, Calochortus, Eschscholzia, Limnanthes, Ribes, Nemophila and Symphoricarpos (for example) are incapable of self-pollination and their reestablishment in any oak-savanna may well hinge on the presence of efficient pollinators. Such tight obligate relations – reciprocal or not – would argue for needing to better understand, and preserve, the full system of plants and pollinators (Rathcke and Jules 1993; Spira 2001). Understanding the pieces – the pollinators, the plants – can lead to a more efficient approach to conserving entire communities. Reserves, such as the Finley Wildlife Refuge at Corvallis, have the potential to maintain populations of a large number of Puget Trough species. The recent discovery of Diadasia nigrifrons, a specialist-pollinator of Sidalcea at Finley, is probably a major factor in the persistence of the endangered Sidalcea at that locality (Steven Gisler [6] unpublished data).

4) LOWLAND HIGH-ORDER RIPARIAN/ASH-SWAMP FOREST: The lowland riparian ecosystem is not well-studied from an arthropod perspective. Though the truly aquatic species of invertebrates are relatively well-known, few of them will appear on any “sensitive” short list since most are both relatively widespread and frequent. The same is not necessarily true for the terrestrial riparian-edge species. It is widely known that in most of the Puget Trough the native species in this category have been replaced by introduced species, and this pattern of continued displacement will accelerate in the future. The major cause behind the diminution of the native fauna is widespread habitat conversion – vernal pools filled in; oxbows eliminated; marshes changed to pastures; gently sloping banks levied, introduction of reed canary grass, etc. The two most likely cases of total species extinction in the Pacific Northwest – the Columbia River Tiger Beetle (Cicindela columbica) and the Willamette Basin Willow Bee (Perdita salicis euxantha) – involve the mining or submergence of unstable sandy riverbanks with an open canopy. This enumeration of species associated exclusively with the lowland riparian habitat is likely to be very difficult (since most of the environment is already altered) and very controversial (since such a large percentage is in private ownership).

Obviously there will likely be little management support for developing a list of sensitive arthropod species, with the complexity such a list could invoke. The most appropriate use of such valid sensitive species would be as a means of more efficiently identifying the specific environments, locations, and ecological communities in which they may still co-occur. Within the U.S. Forest Service lands, many (but not all) such environments have already been recognized in various ways – buffered riparian areas are one major example. Then, it would be appropriate to study the degree to which existing regulations suffice to provide for such rare arthropod assemblages (e.g., how well does the Aquatic Conservation Strategy serve to provide for locations, environments, and microhabitat components for such rare arthropod species & assemblages?) Management will probably be far more interested in efficient, “simpler” habitat management approaches than promoting only species-specific monitoring techniques, even if the former is ultimately based on an understanding of individual species.

In the long-term, this listing of potential sensitive species will be of critical importance as scientists begin to appropriately monitor riparian restoration and more sustainable agricultural practices. Most of the native species likely to be useful in this analysis are probably now resident only in the lower-altitude riparian regions of federally managed lands. We have not had the time to research this potential list of taxa yet; the longer the list, the more useful they will be in subsequent analyses in regions adjacent to federal lands. It is logical for the federal government to take the lead in this collaborative process (along with, say, the Nature Conservancy), since the threatened taxa probably reside today only on public land.

5) HABITAT ISLANDS: Habitat islands (i.e., meadows, balds, etc.) embedded within the forest mosaic account for a sizable percentage of the total arthropod diversity of “forested lands” in the Northwest (beta-diversity in the parlance of ecologists). It is difficult for us to estimate at this time what percentage of this endemic obligate habitat-island arthropod fauna belongs on a short list of most sensitive taxa. For the bee analysis, since all the mid-elevation species are heliophilic, clearly all are associated with habitat islands and hence extremely limited in abundance; however, only a very small percentage of these mountain meadow species are not widely geographically distributed. Thus, a management policy designed to preserve the integrity of the system of habitat islands should suffice to ensure the health of nearly all the associated individual species (ultimately requires validation monitoring subsequent to a listing of the habitat types and resident arthropod species). Analysis of the spider and beetle faunas implies a similar conclusion (e.g., few localized endemics but a large number of habitat-limited species), though the habitat-association of the taxa are less precisely stipulated in the published literature.

There is, of course, already an established list of arthropods associated with caves in the Pacific Northwest (Benedict and McEvoy 1995). Agonum belleri, a carabid ground-beetle, is a well-known inhabitant of sphagnum bogs; it is a very unique species, considered by many specialists to be in its own unique subgenus. This species is probably representative of a small cadre of bog-specialist species. The most likely possibility of unique localized endemicity within habitat islands within the entire Vancouveran Bioregion lies in the serpentine/limestone islands in the Klamath/Siskiyou region. Though extensively investigated botanically (with many local endemics documented) it has not been thoroughly examined entomologically. The examples of amaurobiid spiders (Leach 1972) and flightless Melanoplus grasshoppers (reviewed in Strohecker and others 1968) have raised the possibility that localized endemicity in serpentine habitats is characteristic of many insect groups with limited mobility. Any such species of both limited range and restricted local habitat associations would qualify for a short-list of truly sensitive species.

6) EDGE OF RANGE: A thorough listing of all arthropod species at the edge of their geographic range would be very extensive. Approximately 10% of the bee species within the range of the Northern Spotted Owl in the Pacific Northwest would qualify in this category (based on bee analysis in Appendix 3). Though in vogue with many conservation groups, such artificial political boundaries seldom are considered significant by ecologists. Where range termini do assume unquestionable theoretical significance is where there is a true mixing of different extensive faunal and floral elements; such as the mixing of Rocky Mountain and Cascade distributions in the Wallowa Mountains (“suture zones”) and the mixing of the Great Basin, California Mediterranean and Vancouveran faunas in the high steppes, savannas and serpentine outcrops of the Siskiyou Mountains. In these instances it is generally assumed that peripheral forms sometimes undergo character or behavioral divergence from their parent species, as a prelude to parapatric subspeciation or speciation (Moldenke 1979d). Preliminary databasing of distributional information would allow for identification of such populations of special interest (Dobson and others 1995).

Areas with high species richness may not coincide with areas of high endemism (Prendergast and others 1993; Kerr 1997). The major factor in determining regional species richness is habitat heterogeneity (beta-diversity). Heightened habitat heterogeneity is usually correlated with island-like habitat inclusions within a more widespread basic habitat. The primary goal of ecosystem management of biodiversity, therefore, should be to preserve the pattern of habitat-island inclusions. Since beta-diversity is highest within the Klamath/Siskiyou region this management concern must be adopted and maintained there. The resolution of the most utilitarian initial management plan must be at the habitat/association level (Hansen and others 1991; Brown 1991, 1997).

High levels of alpha-diversity (within-habitat diversity) for arthropod groups generally increase with decreasing latitude within the Pacific Northwest, as they do for plants and most other animal groups. Though the reasons are complex, an increase in alpha-diversity generally indicates an increase in locally rare species as well (Bell 2001). Therefore, within the range of the Northern Spotted Owl, the Klamath/Siskiyou region is likely to support the most rare species. Endemism does not always correlate with scarce species, however, since both geologic habitat heterogeneity and paleohistorical refugia are enhanced in the Klamath/Siskiyou region, in the case of the Klamath/Siskiyou endemism is correlated as well. If management goals are to preserve both areas of elevated species diversity (hot spots) and elevated endemism, then a system of quantitative field surveys must be established to document the existence of such areas. In North America, the Klamath/Siskiyou region is characterized by the highest levels of narrowly endemic arthropod taxa, exceeded only by peninsular Florida and peninsular Baja California (Moldenke, unpub analysis).

The truly rare arthropod species with limited distributions will be an extensive category (or set of categories) relative to other kinds of animals and plants. Based on the bee, spider and ground-beetle databasing, approximately 10% of the fauna would qualify (depending on one’s definition of “rare”). Ten percent of 10,000-20,000 total arthropod species in the Pacific Northwest would be an unworkable list of 1,000-2,000 species.

Listed rare species with very limited distributions will be more frequent in poorly studied taxonomic groups. It will be important to distinguish “relatively” well-documented examples from those likely to be artifacts of infrequent collection and lack of revisionary study. Along with the list of biogeographically significant Vancouveran higher taxa, this list of well-documented rare species should have the most ultimate importance. Many of these species, probably a large majority, are rare (and presumably have been rare for millennia) for reasons unrelated to management activities but determined by naturally limited habitat associations. Rareness/habitat-limitation renders these species especially susceptible to even minor changes in their environment.

However, as a starting point, what is needed is a cataloguing, analysis, and understanding of the locations, macroenvironments, and microhabitats that these rare species are associated with. Limited quantitative fieldwork designed to determine the habitat association of Pacific Northwest arthropods (equivalent to Franklin & Dyrness 1973) is the easiest way to proceed both cheaply and rapidly (see Part Three). The advent of the very real possibility of computer-image identification processes in the very near future has rendered this hhuge job both time affordable and for the most part independent of taxonomic enotomologist specialists (Moldenke, personal communication).

Another aspect of rareness and endemicity should be mentioned here. There are liable to be many undescribed arthropod taxa within the range of the northern spotted owl. How many rare, endemic species still await discovery? Can we delineate additional endemic “habitat-types” that need to be intensively sampled to document the presence of undescribed species of limited distribution (i.e., serpentine communities in the Klamath/Siskiyou)? Most ecologists would agree that a species-by-species approach to conservation is much less effective than protecting habitats. Most ecosystem management decisions both can and must be based on habitat-level resolution. Preserve the habitat diversity and you preserve the most important aspects of arthropod species richness. Preserve both (1) the spectrum of dominant plant communities and (2) the total alpha-diversity of plants within a region, and presumably (an untested assumption generally shared by the entire FEMAT panel of experts) you likely have preserved most arthropod functional diversity as well. Since endemicity of plant taxa is high in serpentine habitats oin the Klamath/Siskiyou, that is where arthropod endemicity is likely to be highest as well. Generally, species-level resolution for arthropods would be required only in instances of the analysis of these special habitats for as yet undescribed species. Large-scale direct monitoring for individual rare species of arthropods is impractical, we believe. It is often difficult to monitor for arthropods without a sampling procedure that kills the subject—clearly counter-productive when dealing with sensitive species. Long-term validation monitoring to discover whether the species composition of communities is changing would also require species-level identifications; however, use of abundant focal species (see below) would greatly facilitate analysis.

Information gaps:

The notion and utility of arthropod indicator species (focal species)

The theme of the International Congress of Entomology in Foz do Iguasu, Brazil (August 20-26, 2000), was “Entomologists preserving biodiversity in the new millennium.” As such, there was a great deal of discussion (formal and informal) about the issues of inventorying arthropod biodiversity and the proper implementation of the concept of indicator taxa.

The discussion of inventory revolved around the usual issue of the limiting problem of available taxonomists to do the appropriate identifications, which requires in most instances the descriptions of heretofore unnamed taxa. Nearly all of this discussion was focused on “third world” tropical biodiversity issues (Brown 1997). To be quite frank, the bandwagon mentality here was simply to generate lengthy species lists – a highly questionable political strategy at the very least, one that might work in the tropics but would not be politically acceptable in North America, where augmenting lists of sensitive species is placing the fundamental Endangered Species Act in jeopardy of repeal. Scientists at the British Museum of Natural History (London) seem to be the most organized driving force behind this type of approach; they have published (or are publishing) several books to try to establish programmatic approaches to systematized inventory methodology (vide: Paul Eggleton and Peter Hammond) in order to compare disparate localities. The work out of the British Museum often uses a functional ecological approach to the categorization of species as well.

Indicator or focal taxa have been utilized in a number of different ways throughout the world. As such, precise definition of indicator status is critical to useful implementation. In the present context, there are two basic types of “indicator” applications (Stork 1990; Ruzicka and Bohac 1993; Faith and Walker 1996; Thiele 1997):

A) using the presence of particular rare species to indicate the supposed possible co-occurrence of additional unrelated rare species that are not inventoried directly (used most frequently for delineation of potential nature reserves);

B) using the local species richness of one taxon to represent the local species richness of total taxa (used to understand the pattern of biodiversity across the landscape, and usually with respect to quantifying the biotic response to a management practice).

The first type of “indication” is currently particularly argumentative. In the British Isles (the best known biological region of the world) analysis has shown that truly rare species (localized endemics that are not habitat-limited) do not have overlapping distributions (Prendergast and others 1993). The facts are not in dispute, but discussion centers around how these data are to be applied in other parts of the world. No one would argue that a number of rare species would co-occur on the top of Mt. Kilimanjaro considering the whole of East Africa; this is an instance of a limiting envirotopographical habitat type. However, within a widespread environmental type (or closely similar types) is there co-occurrence of rare species? This was the question that John Lawton’s (keynote plenary speaker) analysis addressed. Is his analysis applicable to the Pacific Northwest, in particular to the Klamath/Siskiyou? The answer is probably both “Yes” and “No.” One has to examine the causal effects of what makes species rare to understand this issue (Rabinowitcz 1981).

May (1975) showed that the distribution of species abundances within a community could be adequately modeled by assuming a stochastic process, in the absence of natural selection (i.e., competition, etc.). Ever since, the relationship between distribution and abundance of individual species coupled with habitat heterogeneity has been particularly contentious. Current fundamental concepts in “neutral macroecology” (Bell 2001) predict that community composition will change (even without any selection taking place) as the heterogeneity of disturbance changes. Any management scenario adopted will change current landscape disturbance (fire, timber removal) patterns and consequently the arthropod community structure in all habitats covered by the Northwest Forest Plan is bound to change.

The British Isles (100% of their surface area) have been heavily impacted by Homo sapiens for a very long time; as such, the occurrence of rare species is a product of both the “original” stochastic distribution of antediluvian rare species and the effect of the pattern of landscape fragmentation over the last 1000 years in particular. Island biogeographic theory predicts that each species will become locally extinct or will colonize a habitat fragment independently of every other species. In the case of rare species in the British Isles the theory and the data agree nicely. However, in the Pacific Northwest, heavy impact by humans has been confined to the lower elevations and has not affected most forest localities for more than two silvicultural rotations. Clearly, at lower elevations there should be co-occurrence of rare species in the remnant islands of oak forest (or other habitat types), simply because the envirotopographic habitat has been nearly entirely converted to non-indigenous species and there are very few occurrences of native species at all. In sites with relict native vegetation, the oak-forest-adapted species richness should correlate with increasing size of the habitat island and decreasing inter-island distance. In this instance, indicator species usage (A) would be valid. It is clear that 99% of land set-asides will continue to be based on the distribution and abundance of charismatic vertebrate and plant species, but it is also clear that vertebrate diversity patterns are not always coupled with invertebrate diversity (Prendergast and others 1993). Proper ecosystem management calls for arthropod indicator analysis to determine if there are certain regions of invertebrate concern not included in previous set-asides for vertebrate diversity (i.e., owls and salmon).

Indicator analysis also depends upon the “generalizability” of the information. Everyone agrees on two things: 1) the greater the potential species richness of the indicator taxon the better (i.e., use a combination of different taxonomic groups representing a wide range of ecological functions); 2) nearly all indicator analyses will, in general, continue to use only the individual scientist’s favorite taxon, regardless of ecological generalizability. These aspects of indicator analysis affect the Klamath/Siskiyou biodiversity patterns in two ways: 1) monitoring must employ techniques to collect a wide diversity of indicator/focal taxa; 2) data for analysis must utilize whatever subset of taxa that is possible to get authoritatively identified. The techniques currently in use by Rappaport [7] and Niwa [2] (unpublished data) seem to fulfill these criteria well. There is probably no disagreement within the scientific community that arthropods do in fact function as excellent indictors of diverse ecological functions, and that they fill this role better than most other taxa since their life cycles are generally shorter, and they are usually more speciose.

Long-term validation monitoring: direct effects become diffuse through time

Validation monitoring basically is designed to test whether a goal (with its associated implementation practices) has produced the desired result. Forestry is a long-term business; trees require a long time to mature. As the trees are growing, there are literally thousands of other species responding to the pattern of tree growth within any forest stand. The response patterns of understory plants and vertebrates are comparatively well-known and predictable, but the responses of other groups (including arthropods) are only sketchily known.

Taxonomic groups which are often chosen for practical reasons to serve as indicators (i.e., Carabidae, Formicidae, Trichoptera) span a wide range of habitat requirements and generally do function as sensitive surrogates. Within these often-utilized broad categories of species, attention should be centered on the genera that have evolved locally and have been associated for the longest time with this local biota (i.e, Carabidae: Promecognathus; Formicidae: Amblyopone; Trichoptera: Yphria). Though it will take at least another year to complete, we are assembling a list of all the biogeographically significant higher taxa of arthropods endemic to the Pacific Northwest (“BSHT”) and intimately associated with the Klamath/Siskiyou region. Incidental to their central utility as focal species, entomologists from around the world that Moldenke has consulted agree that these endemic Vancouveran taxa are the ones of central conservation concern (regardless of whether they are currently common or sensitive). Since these endemic taxa are and have been, in fact, the central ecological players in these ecosystems, it should come as no surprise that most of them are dominant within their own functional guilds. Alternation of the global climate between wet and dry phases during paleohistory has produced both mesic-adapted and xeric-adapted categories of common and regionally-restricted BSHT arthropods.

Validation monitoring must involve whole communities of species, not just individual species like spotted owls. Long-term effects are by definition rather permanent; therefore long-term changes that may affect whole assemblages of species (dozens or hundreds) are far more important than short-term reactions to a single disturbance event. So far, local forest entomologists/ecologists have been able to discourage the development of lists of “official” sensitive arthropod species. This is because ecosystem management is actually based on the assumption that if basic ecosystem processes are protected (unaltered) and disturbances mimic those that have occurred historically, then theoretically the populations of all species (even the arthropods which are seldom monitored directly) should function normally. These assumptions are based on the notion that often individual species management equates to habitat management in practice. Silviculturists can manipulate certain aspects of forest structure (i.e., amount of CWD), which in turn allows the appropriate community of organisms to exploit that resource. “Management,” in this sense, is a self-delusional term; the system is altered and one assumes (untested for arthropods) that subsequent competitive niche-partitioning will occur in repeatable ways and that the entire former community will become established.

An interesting approach towards habitat management has been presented by O’Neil and others (in Johnson & O’Neil, 2001, Wildlife-Habitat relationships in Oregon and Washington). This approach basically views a habitat as an assemblage of physical structures that modify the climate and produce the microhabitats necessary for cohabiting species to survive and thrive. Hence, providing the physical substrate can “guarantee” the health of the associated species (plural). There is a great deal of practical value in this approach. For instance, there is no doubt whatever that many species of pith-nesting bees are limited by the density of plants like Sambucus and Rubus in the subalpine meadows of southwestern Oregon (Moldenke, unpublished data). Small decreases in the density of these two low-frequency species will have enormous indirect effects on many species of flowering plants. However, the real world is usually too complex to micromanage efficiently. Different species of twig-nesters require different diameter pith; different species of twig-nesters also visit different species of flowering plants. Access to the pith is further controlled by the amount of large vertebrate grazers which clip the stems (or in the previous case by the abundance of Desmocerus longhorn-beetle Sambucus-borers). In other words, micromanaging by specific substrates has limited promise for the arthropod point of view since it can become just as onerous as managing for each species individually.

Another caveat to this generally useful approach lies in a practical reductio ad absurdum that is rapidly exceeded in trying to codify habitat requirements; perhaps such requirements can be encompassed for certain vertebrates, but for invertebrates it is often unworkable in a practical sense. This is because so much of the habitat is defined in terms of actual species identities and not just structures; for instance, the nesting habitat for certain Anthidium bees requires not only the necessary host-flowers, exposed compacted soil and absence of a canopy, but also the presence of any of several species of plants with wooly hairs on their leaves which are gathered for nest insulation. Even when the resource is large and “obvious”, the precise specifications of that resource are often easy to overlook. One emphasis within the Forest Service for the past decade (and detailed by O’Neil and others 2001) is to leave large logs on the surface of the forest floor. But, a log is not a log is not a log (Schowalter and others 1988). Many of the ecosystem processes in which these logs play prominent roles are actually dependent upon the presence of damp-wood termites, not the log itself. Habitat refuges for small vertebrates and arthropods, rapid recycling of nutrients and access to water-storing capacity during the dry summer is all largely dependent upon the current or past occupancy by termites. Termites prefer (very strongly so) to colonize Abies logs; Abies has been all but eliminated from many forest stands in the Northwest, hence the logs on the forest floor are seldom Abies logs. Coarse woody debris guidelines may insure that logs are there, but the desired functions may often not be.

Long-term validation inventory, by necessity, will involve species-level resolution. The reason to employ long-term validation monitoring lies in testing the assumptions that the landscape pattern of communities of diverse organisms can largely “take care of themselves” and that many decades in the future they will be representative of the desired condition. The desired conditions, for the purpose of the current argument, are both ecosystem processes and the persistence of an integral community of organisms (the species list approach). Since there are literally tens of thousands of invertebrate species in the Pacific Northwest, no one can assay for all of them. Rather, indicator/focal taxa will have to be established to serve as surrogates for the total diversity. If well-chosen focal taxa indicate significant divergence from predicted values, then there is cause for general concern for both altered processes and loss of rare species. If monitoring reveals the predicted values then it is relatively safe to assume that the habitats are not fundamentally altered and that the rest of the taxa (the scarcer ones) are equally unaltered. (There have been no tests of this widely held practical assumption for arthropods to date). There is nothing wrong about a list of potentially rare/sensitive arthropod taxa being developed (and in fact it should be), but, in practice, such species lists have historically shifted the emphasis off ecosystem management and on to inventorying for an unbalanced list of proprietary species-by-species interests. This inappropriate emphasis must be redirected to one that determines the ecological requirements of the listed species, overlapping sets of habitat requirements for multiple species and then a determination of whether current set-asides or management policies are already sufficient. This is not a trivial task, and it will require the cooperation and volunteerism of the environmental community.

For instance, certain plants frequently cited as associated with old-growth (Pityopus californicus, Allotropa virgata, Eburophyton austinae, and Monotropa uniflora – and the obligately-associated rare moth Hepialis californica) all share basic ecological characters. They are all achlorophyllaceous mycotrophs, with the same growth form and reproductive characteristics. The scientific question most relevant to increasing their scarce populations is how new clones become established. It is not unreasonable to postulate that actual reproduction may take place unassociated with current old-growth conditions, and that the current policy of LSR set asides may not facilitate population increases at all. Learning how to establish new populations will do far more good for all the species together than preserving remaining sub-vital individual clones. Since Hepialis is an example of a BSHT and liable to attract considerable attention in the future, the question of plant population management is of direct concern to entomologists. Though Hepialis seems to be obligately associated with old-growth conditions, the actual population success of its host may be dependent on factors not associated with old-growth at all.

Any sort of listing of rare/sensitive arthropod species raises issues of practicality and prioritization. As our literature analyses have demonstrated on several disparate taxa of soil-dwelling arthropods, approximately 10% of the total arthropod taxa of the Pacific Northwest could be considered worthy of a tentative sensitive designation (the most inclusive definition, a far cry from legally endangered status). Ten percent translates into literally thousands of species. Though less background information exists on rare species, it is clear that the vast majority of these rare species can be grouped into habitat-related categories. Habitat management for one, should (in general) function as habitat management for the majority within the same grouping. For instance, on lower elevation public lands proper reintroduction of fire and reestablishment of a Quercus garryana woodland would benefit at least 300 (if they still exist) different species of native bee pollinators, which would in turn benefit populations of dozens of species of sensitive plants. Changing the forest structure alone would hardly suffice, since many of the plant species and arthropod species have been exterminated on a local basis already. This highlights the necessity of careful validation monitoring; only observation will tell which species make it back on their own and which will need some sort of assistance. One should realize that it is the interactive ecosystem processes (i.e., pollination) that are important – one can’t have successful plants without the pollinators, and vice versa.

In the opinion of all our colleagues, it is very likely that an overwhelming percentage of potentially sensitive arthropod taxa are associated with particular microhabitats on public lands that are not generally regarded as timber-producing. Limited isolated habitats logically are correlated with less frequent species, if the species are obligately associated with only that one habitat type. Some of these habitats (i.e., subalpine meadows, balds, bogs and serpentine soils) require little more than an enlightened hands-off approach since they are generally limited in extent and economic valuation; others (i.e., oak woodland, terrestrial-riparian) will require intensive management. Validation monitoring is just as important on non-timber-producing land as on timber-producing land under the mandate of ecosystem management.

Even in the absence of management-related disturbances, such species would be of conservation concern. Fortunately, if these habitat types are not likely to be disturbed, then the potential threat is relatively small. However, it is still in the realm of hypothesis that these habitat types are associated with overlapping distributions of the less frequent species. Community compositions of these sites must ultimately be established as standards. For arthropods these standards would represent the equivalent of Franklin & Dyrness (1973) vegetation assemblages. The numerous subtypes of these assemblages, though useful botanically, are not necessarily coordinated with arthropod assemblages (though the practical extent of correlation needs analysis). Within each geographic subregion of the Pacific Northwest it will ultimately be necessary to obtain descriptions of the span of arthropod assemblages (species lists, relative abundances, microhabitat correlates, functional roles, etc.). In practice, this means examples from the driest versus wettest terrestrial habitats; at low-, mid- and high-elevation; and from the different major successional stages. Though this is clearly a very large and complex task, it is the sine qua non for any validation monitoring that tests for the persistence of such sensitive species. A vast number of specimens would be collected and archived in this process, but it is important to note that only a subset of “focal taxa” would need to be worked up initially. If community function and preservation is a goal, then the different communities have to be described in order to permit monitoring.

In order to minimize the detrimental effects of sampling (without replacement), validation monitoring must focus upon abundant species. It might seem, at first, to be counter-intuitive to utilize abundant species to monitor community integrity – since one central interest is to determine whether rare species are in danger of extirpation. The statistical nature of rarity, however, means that robust population estimates for rare species are difficult to obtain – usually involving prohibitively impractical sample sizes. A far more useful measure of the likelihood of rare species persistence is an assay of the most significant and abundant species within the community matrix. If the basic community composition has altered (relatively easy to determine), then it is very likely that a majority of the rare species will have changed as well. It is a question of triage and of directing limited resources efficiently. If some management practices significantly impact even the most abundant species, then immediate attention must center on mitigation; if the dominant species are unaffected then only occasional monitoring of specific sensitive species would be required. Therefore, it should be clear that analysis should be based on the more abundant species; the subsequent question becomes how to choose which focal/indicator groups to analyze.

The arthropod fauna of the Pacific Northwest has had a complex paleohistory and is comprised of derivatives from a number of different sources. However, since the Klamath/Siskiyou Mountain region became climatically isolated in the Miocene Epoch from the rest of North America, a large number of taxa have evolved that are unique to this area. A great many of our current arthropod fauna are descendents of these unique ancestors. Alternation of the global climate between wet and dry phases during paleohistory has produced both mesic-adapted and xeric-adapted categories of common and regionally BSHT arthropods. Our own field implementation of potential monitoring techniques with citizen scientists within the Klamath/Siskiyou during 1999-2000 has demonstrated that biogeographically unique taxa (BSHT), supplemented by appropriate locally-adapted neoendemic species clusters and the occasional dominant species should suffice as focal species. This interpretation of useful focal/indicator species is unique, but our experience shows that it should work well. Success of monitoring protocols is ultimately dependent upon sufficiently precise resolution. Synchronous multiple assay techniques (see Part III, section 4) yielded approximately 100 focal taxa per site for such an analysis.

Landscape diversity and management at the habitat scale

The idea for managing on a heterogeneous landscape scale by maximizing diversity on a stand scale is neither new nor terribly profound. Managing for the diversity of birds and plants through silvicultural manipulation/planning (for example, as potential indicators for arthropod diversity and critical ecosystem processes) would probably suffice for the vast majority of entomological concerns (widely held opinion, not tested locally within the Pacific Northwest). This view holds because: (A) silvicultural aspects of tree species identity/density/growth characteristics/etc. can be manipulated relatively easily; and (B) the biological and physical structure of the dominant tree community has far-reaching effects on total community structure and the flow of ecological processes (see McNeil and others IN McNeil & Johnson, 2001). The principal additional concerns specific for arthropod management (in contrast to bird management) would be spatial in nature; habitat resources far too fine-grained for bird or mammal responses would occasionally be sufficient to support dozens or even hundreds of very patchily distributed arthropod species restricted to certain specialized microhabitats.

Most management plans for specific species actually involve management by habitat; there is seldom actual “species management” per se for species without an economic value. Of course there is “species management” for certain high-profile species – especially the State Fish & Wildlife agencies that directly manage “species” (populations, or organisms, actually) through bag limits, trapping, reintroductions, removals, etc.

Spotted Owl is managed by habitat. Species that are not managed by habitat are things like condors that are removed from their habitat to breed, etc. The long-term cost/benefit ratio and success rates of this sort of thing are questionable even for condors – at any rate they are irrelevant for arthropods. Therefore, in practical terms arthropod management is, and will be, by habitat management (New 1999). The range and specificity of habitat factors and correlates should be more extensive but, at the same time, more fine-grained than the ones currently in practice for vertebrates and plants.

“Adopting” such a habitat management plan is not the same thing as accomplishing the long-term goals. All such management practices will have to be monitored to see if the desired effects on the arthropod community are being achieved. In practice this requires 2 types of information:

1) the best possible “indicators” (because you can’t measure everything, Hughes 2000 and others) [indicators can be species or other things too, such as environmental factors; the concept of “indicator species” per se has not fared well in scientific literature or in management; thus, the recently revised National Forest Management Act guidelines refer to “focal species” and “reference landscapes,” and not indicator species];

2) descriptive knowledge of the final state (or dynamics, not necessarily a static state) you hope to achieve. In the example of arthropods, this means knowledge of the community structures naturally occurring under a range of ecological states.

Under most circumstances one doesn’t even need focal arthropod species to generalize about the health of the arthropod habitat, although some studies suggest this (e.g., Moldenke and Lattin 1990a, b; Nilsson and others, 1995; Pankhurst and others 1997). The full range of bird or herbaceous plant diversity would do just as well as the arthropods themselves in many/most instances (e.g., Morrison 1986; Chase and others 2000; Canterbury and others 2000) – and there already is a significantly more interpretable knowledge base available for the non-arthropods. But sooner or later one has to test one’s assumptions about insects with insects (in a modest number of carefully chosen examples). Theoretically the only discrepancy with utilizing non-arthropod indicators is the question of habitat-grain size, which is significantly smaller for most arthropods than it is for most vertebrate and plant life forms. Producing old-growth habitat for spotted owls is not necessarily the same as producing old-growth habitat for Taracus skunk-spiders. The spotted owl needs trunk cavities of certain dimensions to nest, and 3-dimensional placement of smaller branches that it can navigate within but that keep great-horned owls out. Taracus is concerned with litter depth relative to humidity and prey abundance during the dry season.

Any and all management practices will change the community composition of species. Each management practice will have its own characteristic recovery time. In general, any one community composition is no better or worse than another community composition. Early successional spiders are different from late succession spiders – neither one is better than the other, but both are presumably necessary in overall forest succession. “Better” is a policy and social value, not a scientific criterion. If, and only if, you know the species which comprise these two temporally alternating communities can you compare an experimentally monitored community of spiders to see if you are “approaching some desired condition.” If, for instance, the goal is to “reestablish” old-growth habitat you need to compare the inventory from the treated site with a known community profile for the appropriate old-growth condition. (At present there are no such reference standards).

Where “better” or “worse” conceivably come into play is when unavoidable “collateral damage” from practices involving compaction, xenobiotics or erosion occur and entire ecosystem functions are partially impaired. Under these circumstances, “better” or “worse” is quantified by a numerical change in some system flux rate. The practice most often under consideration is controlled fire application. Both the direct and indirect consequences of fire are huge. Though the changes in an arthropod community profile are not necessarily good or bad, monitoring may reveal that the total richness, density or biomass may have been profoundly affected, thus indicating a probable change in fundamental ecological processes.

For instance, projects by Niwa [2] and Rappaport [7] are designed to document the indirect effects of prescribed thinning and burning on the soil arthropod community. Niwa found that the soil arthropod community after spring burning was different from that after fall burning, and that both differed from the control unburned condition. All 3 arthropod associations were diverse and each was comprised of native species. However, total richness, density and biomass all were significantly more reduced by the fall-burn than the spring-burn. Though a principal concern is how long it will take for the contrasting treatments to return to the ecological functions represented by control conditions, their inventory of soil arthropods as focal species would initially indicate that spring burning is to be preferred.

Highest priority information gaps (from the bottom-up approach)

These information gaps relevant to arthropods are of several basic kinds.

1) background information on the natural history and distributions of resident species;

2) availability of user-friendly identification tools;

3) knowledge of ecological interactions governing the composition of communities (encompassing multiple functional guilds).

Information needs #1 and #2 are obvious and need no further elaboration. They are of particular concern relative to categorization of sensitive (broadly defined) arthropod species [see Glossary of Terms]. Whether sensitive arthropod species warrant the attention given to other types of sensitive taxa is an all-together different issue; first it is necessary to know which species are in fact imperiled. The co-occurrence of sensitive species follows two different patterns:

A. many sensitive species co-occur at a site when the habitat type is limited (often by anthropogenic habitat conversion);

B. within a widely distributed habitat type unrelated sensitive species do not co-occur (unless it is a specialist predator or parasite on a sensitive host species).

In the first pattern, habitat management for one sensitive species may be basically habitat management for many sensitive species. Though this is an oversimplification, it assumes practical importance if a large percentage of the sensitive taxa are threatened by habitat conversion. In fact, in our estimation the vast majority of sensitive species within the range included in this report are threatened principally by broad-scale habitat conversion. Such habitat conversion would include the succession of oak-woodland to mixed forest in the absence of fire, building campgrounds in natural forest openings and channelizing wetlands to facilitate drainage. In this restricted habitat sense “indicator” taxa may prove useful, since such species might be expected to directly reflect environmental conditions, not the population conditions of other species, which may be indirectly correlated through use of the same habitat.

There are many other reasons why a species may be rare. A species can be rare because it is a paleoendemic and exists in old refugial environments; or it is a recently diverged new form, often occurring allopatrically or parapatrically at the edge of its parent species’ range; or it’s recently made rare from anthropogenic reduction in habitat or resources; or an obligate symbiont has suffered declines; or other reasons. There could be many reasons why specific arthropod species could be rare, and no one generalization would explain all such species. Since such rare species generally do not co-occur within a habitat type, it may make their management far more difficult.

There are different levels of sensitivity for arthropod species, just as there are for all other taxa. The highest level of endangerment, where only one or two populations of a species are known to exist (regardless of whether or not its preferred habitat is limited), is known with precision for few arthropod species. Perhaps the silverfish, Tricholepidion gertschi, is the best example within the southwest Oregon region. There are doubtless many more, but rare insects are also usually the least well-known. Many species are yet to be described scientifically; only a thorough search of the literature, targeted field collection, and consultation with taxonomic specialists will reveal most of them. Though it is necessary to highlight such rare species, in practical terms it is far more important to focus attention on species originally widely distributed but now significantly impacted by habitat conversion. In the region of the Northern Spotted Owl, there are primarily three groups of these species:

1) oak-savanna and prairie species;

2) low elevation ash-swamp high-order riparian species;

3) maritime edge species.

The first two community types have already been impacted so severely that we really have no way to estimate how the original communities were structured and how many species have already been extirpated. These are also the ecosystems currently under the greatest pressure and liable to continue to be so in the future (both from direct encroachment and from the indirect effects of introduced species). Rivers and low-elevation fresh-water wetlands do receive a lot of attention from the clean-water interests and the fish lobby. However, there is little information available on native arthropods in these habitats; much of the current monitoring involves introduced species. The Forest Service and BLM are not directly involved in much of this management of plants and other wildlife since the majority of the land is in private ownership.

Oak savannas appear at first glance to still be holding their own; this is largely an artifact of the presence of old oak trees themselves. Nearly all regions with such remaining stands of Quercus garryana no longer possess the native herbaceous understory nor the native entomofauna; replacement by introduced species has been almost complete. Although it has not been documented, it is possible that there are significant remnants of this entomofauna still extant on BLM and USFS lands in southwestern Oregon and also the eastern slope of Mount Hood where underburning has been reintroduced. Isolated patches throughout western Oregon and Washington held by The Nature Conservancy probably only support traces of the original species composition.

The maritime environment of Washington, Oregon and northern California is under the reality of ever-increasing habitat conversion. Though significant portions of the coastline are inaccessible and hence remain “wild”, the rapid spread of numerous introduced terrestrial and aquatic species poses a very significant threat to what remains. The true maritime terrestrial arthropod fauna originally contained a limited number of endemic species (probably about several hundred). Significant progress in compiling the arthropod composition of the various types of maritime communities could be accomplished easily since the total faunal diversity is relatively low and the community types relatively easily accessed for quantitative inventory. Fortunately, as far as is known, many of the native truly terrestrial maritime species are distributed widely North/South along the coast and are not in imminent danger of regionwide extinction. However, in terms of ecosystem management, we shouldn’t even be thinking in terms of minimally viable populations, but of healthy robust communities. The Forest Service (especially through its stewardship of Oregon Dunes National Recreation Area) shoulders the responsibility for managing some of the most extensive portions of this environment.

Relative priorities of the major arthropod faunas: 1) most importance placed on the oak-savanna/prairies; 2) less on maritime systems and ash-swamp high-order riparian species; and 3) least on the majority of the land (even though “forestry practices” have so alarmed the scientific community).

Forested lands comprise a large percentage of the land within the region, but these conifer forests have been less subject to habitat-conversion. However, within the forested lands managed by the Forest Service there are 4 topics of concern from an arthropod conservation point of view and 4 from an ecosystem management point of view.

Conservation concerns:

alpine/subalpine habitats
endemicity within environmental islands within the forest mosaic (habitats like edaphic mountain meadow communities are easily recognized, but there are other microhabitats (“special habitats” in Forest Service terminology – see Dimling & McCain 1996) with specialized arthropod faunas which are more cryptic (e.g., seeps, waterfalls))
species closely associated with old-growth forests (there are a lot of species in old-growth, and not all are close associates; the concern would be for those species that maintain their highest population levels in old-growth and are infrequently found in younger stands.)
integrity of riparian habitats

Management concerns (see Part III):

fire management (both decades of fire exclusion and reintroduction of cool spring prescribed burning) and nutrient recycling processes
riparian management (as hot-spots of invertebrate food resources for vertebrates)
pollination ecology of understory plants
ecological integrity of old forests in formal reserves (e.g., LSRs) and in smaller and more isolated residual patches

Though different ecologists and entomologists would give slightly different conservation priorities to these 4 categories of conservation concern, the first two are in a league of their own. The species restricted in distribution to the first two categories either have populations that are suspected of being highly isolated, or have been documented as being so. As such, the population dynamics of any organism with isolated populations becomes pertinent. Since the alpine species have demonstrably isolated populations (and lots of interpopulation variability; see Stephen 1957, Pongprasert 2000) they would receive the edge for conservation priority over the cryptic microhabitat specialists. The other two categories of old-growth and riparian associates are already being given indirect consideration by the NW Forest Plan. Since the unique biodiversity of riparian zones far exceeds that of upslope old-growth on a per-acre basis, and since riparian management is far more hands-on, riparian species should have conservation priority over old-growth taxa.

1) Alpine/subalpine: The largest percentage of species with limited distributions occurring within the general National Forest System are clearly those restricted to the subalpine/alpine Hudsonian zones. Though the percentage of Hudsonian endemics varies greatly from one taxonomic group to another, it probably averages between 20-25% of the total fauna (based on our own results, see species lists in Appendices 6-14). As cited in the published literature, many of these alpine taxa appear to be either very infrequent to very rare. Some significant aspect of this perceived rarity is undoubtedly due to the inaccessible nature of the habitat to scientists. However, the bottomline is that many of these taxa may actually be rare, and because most are constrained to a very long, narrow band of suitable terrain, the potential of global climate change to seriously affect them is great. They are all subject to serious density-independent climate-related population controls, as well as to the range of normal competitive relations existing at more moderate elevations. Due to the effects of Cenozoic palaeohistory on this entomofauna, a large percentage of the species show localized distinct forms, further complicating the issue of biodiversity and species delineation (Stephen 1957; Pongprasert 2000)

Two examples of uncommon arthropods found in higher elevations are flightless snow-scorpionflies (genus Boreus with 13 western species) which are often found in moss on which they feed, and often on the snow in winter (Figure 12); and the primitive wingless glacier-fly Chionea (Figure 13), which occurs mostly on glaciers and icefields in which they feed on red algae, but has been recently discovered in upper elevations of the forested southern Washington Cascades (Marcot and Moldenke, unpub. data)

Figure 12. Snow scorpionfly, Boreus.

Figure 13. The primitive wingless glacier fly Chionea.

2) Endemicity within the forest mosaic: Though poorly documented locally, habitat heterogeneity and associated microhabitat specialization play a large part in forest entomofauna biodiversity (Furniss & Carolin 1977). Forest managers, quite logically, classify a landscape into potential tree species categories. Though these categorizations often include an understory element (see Franklin & Dyrness 1973), they probably do not have appropriate resolution to reveal the diversity of generalized arthropod communities within a watershed. These sites may be as extensive as several dozen acres or as small as a quarter of an acre and still maintain their ecological integrity. Microsites such as bogs, marshes, balds, meadows, swamps, and rock seeps/waterfalls fall into these categories. Though of little or no importance in timber production, they are characterized by literally hundreds of species of arthropods that occur nowhere else in the forest mosaic (based on studies in the Andrews Forest; Gary Parsons and Moldenke (unpublished data) and Heyborne (2000)) Since many of these microsites are of limited geographic extent, it is likely that, by the laws of island biogeography, similar types of sites will contain different species. The occurrence of such sites should be a part of any watershed analysis. Once they are characterized and mapped throughout the region (see Dimling & McCain 1996), a carefully chosen subset needs to be inventoried. The proportion of the forest diversity encompassed within these specific microsites is liable to be surprisingly large (perhaps 40-50% of total diversity; see McIver and others 1990, 1992); because of their spotty distributions many of these widely distributed but only apparently abundant species should be classified as cryptic-sensitives.

Many microsite-adapted species are liable to be characterized by broad distributions throughout the entire region, yet be poorly represented in reference collections. Their apparent scarcity is real because they are actually associated with extremely limited microsites rather than the generalized forest mosaic; as such, they represent a critical cryptic subset of the generally poorly documented entomofauna of the Pacific Northwest forests.

3) Old-growth-associated taxa have seen a great deal of attention in the past 10 years. Though arthropods are alluded to, there is surprisingly little documentation of the degree of old-growth association. As far as soil fauna are concerned, the information we know can be summarized as follows: A) the fauna of old-growth forests is very diverse B) the fauna of experimentally-paired merchantable-age forests is also very diverse; C) many of the species occurring in the merchantable-age forests also occur in the old-growth and vice-versa; D) the relative abundances of the species which occur in either forest type is usually distinctive. So, there may be relatively few species that are entirely restricted to old-growth, but old-growth arthropod assemblages certainly merit special conservation attention because many species reach their peak abundances under old-growth conditions. The general distinction in insect fauna between mature and old-growth forest stands seems to encompass canopy, terrestrial, and soil arthropods, that is, the entire arthropod community in toto (Schowalter 1989; Moldenke unpub. data; Progar & Schowalter, 2 unpub. mss). There are two issues of critical importance relative to old-growth: 1) How fast do mature forest stands reach old-growth conditions? (A large proportion of LSR-designated stands are certainly not at that age yet); 2) How large does an isolated stand of old-growth have to be in order to preserve forest-interior species and environmental characteristics? (see the research design of Marcot (http://www.SpiritOne.com/~brucem); Cadenaso and Chen and others 1993; Pickett 2001; Heliölä and others 2001). Studies in southwest Oregon by Madson (1997) have indicated that the most indicative species are medium-sized arthropods such as pselaphid and staphylinid beetles (ca. 3-5mm), rather than the largest arthropods, e.g. ground-beetles (collected by pitfall traps) or the smallest arthropods such as mites and springtails (collected by Berlese extraction).

4) Riparian zones are hotspots of biodiversity, especially for the terrestrial entomofauna (e.g., France 1997). Ongoing studies by Chan and Olson (Olson and others 2000a-c; Tappeiner and others 2000; Randio and others 2001) and others have documented that these environments are characterized by a remarkable degree of environmental heterogeneity. Presumably this microclimatic heterogeneity plus any additional habitat diversity created by the unpredictability of streamflow events (as well as an abundance of food resources) causes significantly enriched biodiversity in the streamside environment (Antvogel and Bonn 2001; Collinge and others 2001). Any management practices, especially those emerging from the issue of salmon restoration, that tend to decrease stream heterogeneity will pose serious threats to this diversity. The microhabitat associations of these arthropod species still need to be documented especially as it relates to stream width and lateral distribution of species into the upland forest floor (current unpublished research by Moldenke, Chan and Olson).

Summary for Part II:

The over-riding consideration in a bottom-up approach to forest arthropod community integrity is the enormous richness of native arthropod species. Preserving the level of landscape plant community diversity is the most important approach for prteserving arthropod diversity. Short-term validation monitoring would entail focal species analysis of the effects of management practices; procedural difficulties would be minimized by chosing BSHT and other very abundant species as indicators. Long-tern validationh monitoring would require description at the species-level (“fine scale”) of arthropod community structure (analogous to plant communities); this process has not yet been initiated. As far as a conservation emphasis on individual species is concerned, consensus places priority on BSHT. Since most BSHT are extremely abundant, therefore, the occasional rare BSHT should be afforded priority #1. Habitat associations need to be established for any additional rare species (non-BSHT), and conservation priorities centered on habitats with numerous co-occurring rare arthropod species. Individual rare non-BSHT arthropod species that do not occur in special environments should receive less priority; the largest group of these may be species “largely associated” with old-growth.

PART III: Specific prioritized research recommendations:

1) SOIL NUTRIENT RECYCLING AND ARTHROPODS

Processes in soils are both the most important and most complex drivers of ecosystem productivity. The productivity of forest ecosystems in southwest Oregon is limited both by water and nitrogen. The pre-settlement fire regime had complex interactions with both limiting factors. The difficulty in reestablishing conifers after harvesting in parts of the region has focused attention on natural soil processes. Despite this attention, there is little agreement within the silvicultural community as to the proper mitigation techniques.

Under most circumstances, Pacific Northwest forest floor nutrient recycling is driven by arthropods and fungi. It is a slow-release system that provides nutrients throughout the year. Since physical shredding of the litter is necessary to expose the nutrients to microbial exoenzyme attack, an appreciable litter layer builds up due to the feeding-deterrent chemicals present in most of the dead leaves. As the litter layer builds, the community of soil arthropods becomes ever more dense and diverse. Mesic forests west of the Cascade Crest, from which fire has been excluded, have very deep litter layers and consequently support the highest densities and diversities of soil arthropods anywhere in the world (Moldenke, 1999). Water-soluble nutrients are released in the fecal material every time the organic material is passed through the digestive system of an arthropod (usually a fungivore). Fungi actively grow throughout the entire year and so the activity of the fungivorous insects provides nutrients throughout the year. Each time organic material passes through the digestive system of an arthropod, additional nutrients are mineralized for the plant roots and additional mucus is secreted onto the fecal pellet that promotes saprophytic growth. As soon as the fecal pellet is covered with microbial growth it is fed upon by yet a different species of fungivore, and additional nutrient ends up in soluble form. Fecal pellets which start out very large (from a millipede, for example) are soon fractured into tiny pieces, whereupon they may be ingested by an earthworm and reformed into another large fecal pellet and the process of alternating fragmentation and microbial surface growth starts all over again. In what has been described as a “cascade process”, nutrient is released in small amounts, but continually (Moldenke and others 1994, 1999).

Fire in the forest understory changes all the usual processes of nutrient release. Fire consumes the organic matter on the forest floor and leaves the nutrients within the ash. (A very hot fire will actually vaporize a sizable percentage of the nutrients.) Ash is highly subject to wind erosional processes and nutrient capital is often transferred down to the riparian zone in significant quantities. Much of the nutrient remains though, and its fate is determined by the sequence of subsequent rains. Natural fires occur during the late summer and fall and are often followed by thunderstorms. The nitrogen in the ash is very soluble and if the first subsequent rainfall is significant (as is usually the case with a thunderstorm), then a large percentage of this immense quantity of soluble nutrient will be lost to groundwater since there is little remaining active fungus in the upper soil layers to absorb it. However, if the first wetting is just a gentle misting (the normal winter precipitation) then the nitrogen is picked up by soil bacteria, which undergo an immense population bloom. Within several days this bacterial bloom is followed by a predaceous protozoal bloom and subsequently by a predaceous rhabditid nematode bloom – and most of the nitrogen is once again released in water-soluble form in their excrement. If this series of population blooms occurs during the dry summer/fall, there would be insufficient fungus able to assimilate it, and again it would be largely lost to the ground water.

During the spring, however, prescribed burning is thought to engender a different set of rules (Bruce Caldwell [3], personal communication). The fungal biomass in the soil is active and it assimilates a great deal of the nitrogen directly. The soil bacteria are also active and they assimilate probably nearly all the remaining rain-caused pulse. The bacterial-protozoal-rhabditid succession which follows in turn liberates a huge pulse of nitrogen, but the already stimulated fungal biomass will efficiently absorb the majority. So, the natural fall-occurring fire ecology places a premium on already-limiting nitrogen, but spring burning will probably act as a short-term fertilizer on plant productivity. Frequent spring-burning is liable to have profound long-term effects on vegetative competition since plant species differ in their responses to additional nitrogen; little change is to be expected within the forest itself, but weeds would be expected to invade shrubby communities. Since many of the most sensitive plants and insects occur in the serpentine scrub communities, spring-burning may pose a very serious threat under these special conditions. Spring-burning does not mimic the natural fire cycle in many basic aspects, rather it may completely shift the soil nutrient transformations. Widespread adoption will likely have immense consequences on the arthropod soil fauna that we are unable to predict at the moment.

Water is also limiting in these southwestern Oregon forest environments. Fire effects on the water-holding capacity of the soil are complex and vary with soil type and heat intensity. Certainly most of the short-term effects of burning will be to decrease water-holding capacity by consuming both the insulating litter layer on the surface and much of the organic humus layer below. Cool prescribed spring-burning should minimize these adverse effects. In the short-term, prescribed-burning may kill many of the burrowing invertebrates that permit rain infiltration (i.e., ants, cicadas, cicindelids, bees, worms), but long-term it may change invertebrate succession and facilitate colonization by some of these same groups (currently there is no data). In southwest Oregon inadequate infiltration of thermal summer storms is quite possibly an undocumented process limiting production over much of the landscape. Particular groups of soil-dwelling arthropods have a disproportionate role in regulating this rate of infiltration.

This topic for analysis is cited first on the basis of overall concern. Soil studies are complex and must be interdisciplinary for maximum effect. We have not chosen any one specific aspect as most worthy of analysis.

2) FOOD FOR THE TROPHIC PYRAMID

How does the abundance/biomass of insects affect potential vertebrate population levels/territory size? A) foliage-gleaning birds; B) fly-catching birds/bats; C) terrestrial mammals, amphibians, and reptiles.

Vertebrates require large inputs of food energy to maintain activity and successful reproduction. Insects (and other arthropods) comprise a major portion of these resources for most vertebrate species. Even species which primarily consume plant matter by bulk are very dependent upon nutrients in the arthropods less frequently consumed. Insect abundance/availability is not uniformly distributed across the landscape. In particular, it peaks in availability near streams or within open shrubby communities or early stages of regrowth. Within these two habitats, insect abundance is heavily correlated with particular microenvironmental features. It is very likely that successful territory establishment and subsequent breeding of local birds and mammals are heavily dependent upon hot spots of arthropod food resources. We need to have a better idea of how these hot spots are distributed and whether they change seasonally.

Many basic forest management practices have profound effects on the distribution and effectiveness of these hot-spots of arthropod occurrence, particularly timber harvest, prescribed burning and fish management. Though the ecological connection between food resource levels and vertebrate success is an obvious one, it is difficult to establish exactly when resources are limiting – since a bird’s population, for instance, is responding to a multitude of factors simultaneously (i.e., predation, food-limitation, disease, nest-site availability, off-site mortality for migratory species, etc.). A first step, however, is to quantify the pattern of available food resources.

Different species of vertebrates acquire their food in fundamentally different portions of the environment.

A) Foliage-gleaning birds patrol the surfaces of leaves and stems primarily for large-bodied arthropod herbivores (most importantly, caterpillars), moderate-sized arthropod predators (primarily spiders), and whatever small micro-grazers and resting flighted insects they encounter. Different species of plants support very different average biomasses of herbivores (Joan Hagar [15], personal communication) due to the different types of anti-herbivore biochemicals present in the leaves. Plants with high herbivore loads may be abundant in the local habitat and plants with few herbivores rare (or vice-versa). Paul Hammond [8] and Jeff Miller [9] (1990, 1993) have documented in several sites in Oregon that >80% of caterpillar species prefer non-coniferous plant hosts to conifers. On an ecosystem-wide leaf biomass basis, conifer leaves account for probably more than 95% of the photosynthesis, but more than 80% of the total caterpillar biomass is produced on the leaves of deciduous understory shrubs and herbs.

Understory shrubs like ninebark (Physocarpus) support both high species richness of caterpillars and total caterpillar biomass, shrubs like Rhododendron support both very low diversity and biomass. In large regions of the forested Northwest, both the abundance of shrubs and caterpillar biomass increase as one approaches a stream channel (Jiquan Chen [10] unpublished data). This phenomenon is often accentuated where beavers have been eliminated (most of the Northwest), and unconstrained stream banks that formerly were under beaver impoundments now support luxuriant shrub growth. A limited extent of streambank habitat within a bird’s feeding territory can presumably compensate for extensive unproductive uplands.

To achieve the necessary preliminary data on prey abundance and biomass, it is logical to “beat” shrubs both at streamside and in the upland. Each woody species should be treated separately. Collections should be taken every two weeks throughout the growing season. Results should enable estimation of the diversity and biomass of arthropod prey throughout the year correlated with host species and habitat placement. An ecological profile of each woody species should be assembled detailing the main herbivores, activity periods, and which arthropod predators are most numerous.

B) Insects on the wing are the prime food resource of fly-catching birds and bats. Though the relative biomass cornucopia of the stream habitat versus the upland has been documented in desert environments, it seldom has been quantified in temperate forest habitats. No ecologist doubts that the same phenomenon occurs in the Pacific Northwest conifer biome. Progar and Moldenke (2001a,b) have shown recently that the numbers of flying insects emerging from Oregon headwater streams vary with canopy cover, time of year, and stream characteristics. During much of the season, more than 100 insects may emerge from a single square meter of stream per week. They also have shown that emerging biomass is sometimes not coupled with emerging densities of aquatic insects. Very little hard data has been collected about the diel flight periods of the major species of aquatic insects (bat-food versus bird-food) (Alex Farrand [11] unpublished data).

Relative to their role as vertebrate-food, the most efficient way to monitor the production of flighted insects from aquatic environments is by the use of “emergence traps” (Leif Gundersen [12] unpublished data). Traps would be set in a design to quantify the effect of canopy presence, stream flow and seasonal trends. Traps would be collected and reset once every two weeks. In addition to quantifying the effect of aquatic insects as flycatcher food, basic natural history data on the life cycles and identification of the species present would be collated.

C) There is a greater diversity of vertebrates utilizing the terrestrial ground-surface arthropod food resource than either of the two former categories. Studies by Brenner (2000) and Moldenke (in prep) in the past decade have shown that the abundance of large terrestrial arthropods increases dramatically as one approaches a stream. These higher densities are comprised of two distinct types of species: 1) those restricted in all of their activities to the terrestrial-riparian zone; and 2) those typical of the upland but which visit the riparian zone presumably in search of both more abundant food and available drinking water. There is probably a far higher level of available arthropod biomass in the ground-surface stratum than in either the foliage-gleaning or the fly-catching strata. It is known that the arthropod species inhabiting this zone are active at distinct times of the year, and that species richness of arthropods is extraordinarily high.

Ground-dwelling mammals, amphibians, and reptiles would be most affected by the distribution and abundance of potential terrestrial macro-arthropod prey. This food resource is most easily quantified by pitfall-trapping. Traps should be harvested and re-set once every two weeks.

3) POLLINATION INTERRELATIONSHIPS

Which species of insects are the most important pollinators of plants in the Klamath/Siskiyou region? Which localities still maintain viable populations of about 300 different pollinating native bee species in the lower elevation oak savanna and prairie communities? Do the serpentine vegetation types at higher elevations act as refuges for the majority of bee species that may have been totally extirpated at lower elevations?

One exemplary research project would involve the pollination ecology of native plants. The Klamath/Siskiyou is the logical center of this focus, though the implications will be felt region-wide.

Case in point: As documented in any number of publications (recently reviewed: Oregon State of the Environment Report, Paul Risser editor in chief (2000); Wildlife-Habitat Relationships in Oregon & Washington, DH Johnson and TA O’Neill, 2001), the Westside Valley Prairie and Oak Savanna habitat types have been the most severely impacted of all native Westside Oregon communities. The extent of the impact is usually understated, since remnant oak trees persist in many localities where the entirety of the native herbaceous flora and arthropod fauna has been extirpated (Williams and others 2001). Overgrazing by livestock in the mid-1800’s replaced the native perennial grasses with annual introduced species that indirectly changed competition, eliminating most of the forbs as well. Remaining herbaceous flora was gradually eliminated through the shrub and sapling buildup consequent to fire suppression. The pollinator fauna co-dependent on these plants was eliminated as well. Though there are some remnant native herbaceous plants in certain locations, it is very likely that they no longer support the majority of native pollinator species (William Stephen [4], Moldenke & Ver Linden, current research). The recent decline (almost complete elimination) of feral honeybee colonies due to the respiratory mite has placed these remaining plant populations in jeopardy. Though there has been an increase in bumblebee populations as a result (pers. obs.), this usually does not assist the native plants in question, since bumblebees are too large to efficiently pollinate many species of native plants.

Which species of insects are the most important pollinators of plants in the Klamath/Siskiyou region? Originally more than 400 different species of bees inhabited the lower elevation of westside Oregon and Washington (Moldenke 1976b).Relict areas most likely to continue to support diverse populations of these native bees are in the Klamath/Siskiyou region. There are several reasons for this: 1) the species richness of bees was originally far higher there (see Moldenke database listing of bees of region); 2) the amount of habitat conversion is least there, since human populations remain low and fires still occur rather frequently; and 3) the presence of edaphic conditions limiting tree cover are most frequent there. A large percentage of the original bee fauna of the Westside was restricted to the low elevation savannas and prairies. We fully expect that many of these species of both plants and bees have been able to persist at higher elevations that are less disturbed by civilization in the serpentine soils of southwestern Oregon. Though ecologists generally emphasize the poisonous nature of serpentine soils and the specialized adaptations required to inhabit them (note any list of endemic plant species in the region), they do function as island preserves for native flora and fauna, since the introduced European weeds generally can not invade them successfully. Perhaps there are even hitherto undescribed specialized endemics in these community types following the pattern of neoendemic grasshopper radiation (Strohecker and others 1968).

All plants benefit from, and many require, pollination. Even plants with tiny flowers normally subjected to habitual self-pollination benefit from the outcrossing provided by bee (and other insect groups) pollination. In the long-term, this is true of all species in all environments (Rathcke and Jules 1993; Bond 1994). Certain plants in any community, including people’s gardens, have large showy flowers, produce copious nectar and attract sufficient numbers of social bees to insure pollination under most eventualities. However, it surprises most people to learn that the majority of species of wild plants in most communities are seriously pollinator-limited (Moldenke; 1976a). The two most important reasons for this are: 1) the paucity of pollinators (of all taxa) in many plant communities; and 2) the competition from the showiest species with the biggest rewards. Though most such plant species with relatively “unrewarding” flowers in the Pacific Coast States have evolved the insurance policy to self-pollinate under the conditions of pollinator limitation, they still rely heavily upon two classes of pollinators where available (Moldenke 1976a, 1979c).

The first class of pollinators is composed of bees evolved to visit only one species of flower; these specialist bees are frequent in the Northwest, comprising between 25-33% of the total number of bee species (Moldenke 1976a). They will visit their target-flower faithfully, regardless of what other plants are blooming at the same time. Most importantly, they will faithfully visit their target-flower even when plant populations are low and individual plants are sparsely distributed (unlike generalist-pollinators, who will never do this). There is no doubt that specialist-feeding bees have been the most seriously impacted (pers. obs.). The second class of bees is the small (in size) generalist-pollinator group. If conditions allow bee populations in general to thrive, these small-sized generalist-pollinators can efficiently pollinate dense populations of many plants. Fully 37% of the flora in central California is completely dependent on this type of pollinator (Moldenke 1976a, 1979b).

Even if there is a remnant seed bank in some of the oak savannas that can be activated through fire management, one still must have the pollinators for successful plant establishment (Spira 2001).

The Westside must be inventoried for remaining populations of native bees. One most logical place to look is in extensive federal land holdings not immediately subject to potential habitat conversion. Others are: 1) documenting which bees have survived in isolated hot-spots of remnant plants (e.g., Nature Conservancy sites); and 2) documenting the bee fauna in large undisturbed sites most likely to have retained a large portion of the original fauna (i.e., serpentine communities within southwestern Oregon). The second type of site might be able to serve as a source for reintroduction in the future.

It is noteworthy that in Europe, probably the largest percentage of any group of organisms cited in the Red Book of sensitive species is the nearly 150 species of bees (Matheson 1996). Bees as a group greatly benefited from the habitat conversions associated with agriculture for thousands of years, but the introduction of mechanized systems and the coalescence to larger land holdings during the past 50 years has turned the tables and now imperiled most species both in northern Europe and western Oregon. Jurisdictional responsibilities are complex in low-elevation Oregon; the ownership of most remaining populations is hopefully the Forest Service, the lands most likely to respond to oak savanna mitigation belong to the BLM, and the lands most likely to be receptive soonest to the possible reintroduction of native bees belong to the Nature Conservancy. Regardless of future negotiations, the first task is to jointly determine which species remain and where.

4) CHARACTERIZATION OF ARTHROPOD COMMUNITIES

This literature review process supports the conclusion that the Klamath/Siskiyou region supports the largest number of species with limited distributions in the entire Pacific Northwest. This region clearly has the greatest potential for new as-yet-undescribed species as well. Regardless of whether a significant portion of the region becomes a National Monument, the most fundamental information requirement remains: a comprehensive listing of all arthropod species associated with their appropriate habitats and abundances. This is by far the most significant of the research recommendations. However, it will require considerable effort and integration of field sampling design, literature analysis and newly collected data.

As previously highlighted in this report – old-growth and oak-savanna communities, in particular, require characterization (relative to arthropod fauna). The span of serpentine communities requires analysis both as the location of potential previously “undescribed” species and as refugia for arthropods previously more widespread in open-canopy low-elevation plant communities. Research under the direction of Marcot (see http://www.SpiritOne.com/~brucem) has documented the necessity of a comparative research design for arthropods whose biologies are so poorly known. The identity of old-growth forest interior species can be determined most easily by directly comparing old-growth insect communities with those of either mature forest or young plantation stands immediately adjacent. Many arthropods are readily mobile, and their mere presence in a certain type of stand does not necessarily imply a close ecological tie to that particular community; only comparison of relative densities between plant community types (in the absence of documented direct ecological relationships) will accomplish that.

Entomology cannot provide any faunistic resolution comparable to that of botany at the present time. This requires that new information must be gathered subsequently in a systematic fashion. Once gathered, in broad terms, the most frequent species in each habitat (especially if they are BSHT species) could provide the basis for long-term monitoring, if desired. Because there are both theoretical and procedural difficulties in monitoring directly for the rarest species, such common focal species would be used in the original sense of “indicators”. If the most frequent species, or abundant species most typical of a certain association, showed no significant changes to forest management practices, then it is less likely that the rare species would have changed significantly. Though this synecological type of long-term monitoring would require resolution at the expert species-level of identification, logistic difficulties would be minimized by focusing on the most common species. Even a list of the commonest focal species from a broad range of the different components of any community (e.g., soil-inhabiting, understory-inhabiting, canopy-inhabiting collected with a variety of trapping devices) would encompass more than 100 species. Hopefully, this extensive level of resolution would render credence to an indirect monitoring with focal species.

Synecological description of the insect community requires systematic collecting and considerable sample processing before one even gets to the stage of authoritative identifications and efficient databasing. A limiting factor in the collection phase is sampling design; a statistically robust design will allow quantification of alpha- and beta-diversity relationships. A limiting factor for sample processing is time (or money), pure and simple. The identification process will always unavoidably be to multiple levels of resolution (some taxa will be identified to species readily, others may never reach that level of resolution because no taxonomic specialists are available). As long as all the information is entered into the database, more sophisticated comparative analyses can always be programmed at a later time as more authoritative identifications are gradually obtained.

Practicality should be the first consideration when it comes to fieldwork. If the goal is to describe the insect community at a specific site, sampling must encompass all the seasons. Since the majority of individuals collected by many sampling techniques are immature (and hence often not possible to identify), only consecutive sampling will permit correct identifications, and at the same time allow archiving of immature stages associated with adult specimens so that subsequent researchers will be able to identify immature stages as well.

Another practical aspect concerns multiple sampling protocols. If one has gone to the trouble of laying out an appropriate statistical sampling design, one can maximize effort by employing several types of traps simultaneously that sample different segments of the fauna. Any one sampling device collects only a small percentage of the species present. To gain a sufficient level of description of the entire arthropod community necessary as a benchmark for long-term validation monitoring, several different techniques must be employed. Each type of standard insect trapping technique has its own characteristic cost to benefit ratio (see Table 2). Extensive literature searches have been made as part of this report; the world literature on the “pros” and “cons” of different sampling devices is now readily available in our appended data base.

INTEGRATED INVENTORY DESIGN

Implementation Work-up

Initial Equip $ Cost Information speed

Pitfall Trapping Lo Lo Hi Fast

Berlese Extraction Lo Lo Hi Slow

Beating/Sweeping Lo Lo Lo Mod

Light Trapping Hi Mod Hi Fast

Malaise Trapping Mod Lo Hi Slow

Aquatic Emergence Mod Lo Hi Mod/Slow

Table 2. Relative costs and information gains from different standardized insect trapping devices. (There are many other forms of arthropod trapping, but none are as efficient as these for sampling a broad range of diversity.)

The first three methods have been employed by Marcot in the past several years in his study of old-growth patch size biodiversity (http://www.SpiritOne.com/~brucem). Realistic cost estimates could be based on his studies. The method most frequently employed in the literature is pitfall trapping (Ruzicka and Bohac 1993; Spence and Niemela 1994; Rykken and others 1997). However, it is quite apparent that any synecological arthropod studies will have to be undertaken with volunteer citizen-scientist field sampling and initial laboratory processing (see COMTESA p. 7, www.ent.orst.edu/comtesa/ and Appendix 17). This research paradigm is already in practice in southern Oregon through the Applegate Alliance, and will soon be available through the Cold Spring Conservancy in the Columbia Gorge. Moldenke’s summer science teacher consortium with the Forest Service has demonstrated that all field trapping techniques and initial laboratory workup are performed with excellent quality control by novice volunteers. Entomologist consultants will always be needed for the final steps of authoritative identification, however. In this regard, the potential for automated arthropod identification through computer image analysis techniques is at hand. It is quite likely that identification of adult insects in the orders Hymenoptera, Diptera and Lepidoptera is already doable; the process requires only the filling of the appropriate databases with reference pictures from identified museum specimens (Moldenke, personal communication and current grant proposal). The Klamath/Siskiyou region is liable to be so incredibly unique with respect to arthropod diversity that no opportunity should be overlooked to obtain this requisite management responsibility.

Throughout the region, however, the two most pressing aspects of arthropod diversity vis-a-vis management will continue to be: (1) the landscape pattern of patch size and successional stages, and (2) the delineation of special microhabitats embedded within the true forest mosaic.

The ecological integrity of any island-like habitat is of critical concern as the diameter of the patch decreases. This has been one of the most discussed aspects of spotted owl conservation plans. Spotted owls are on the larger patch size end of the spectrum, while many arthropods (especially those associated with the soil) can persist in much smaller islands. How large a patch of old-growth is necessary to maintain the ecological functions of old-growth is an oft-debated issue. Research by Marcot (current) is analyzing the arthropod community integrity of small patches of mature forests in southwestern Washington. Some ecosystem processes, like cross-pollination of understory plants, are facilitated by edge-effects, but other faunal components may be equally hindered. Further discussion of ecological integrity would benefit from analytic field data.

When special habitats, with their associated endemic arthropod associations, occur under the forest canopy (like some swamps, bogs and waterfall splash-zones, etc.), they will need to be actively incorporated into management plans, unlike the fauna associated with balds, meadows (etc.) which should thrive under a system of benign educated neglect. A list of these former cryptic-sensitive species within the range of the Northern Spotted Owl will encompass many hundreds, though any one habitat type is likely to be characterized by a limited list. From a realistic point of view, these will be the species most likely to be seriously affected negatively by management practices. Identification of the fauna to a species-level of precision is necessary to establish what proportion of the fauna inhabits such island habitats within the greater forest mosaic, however, once the habitat islands themselves are identified in the GIS database (see Dimling & McCain 1996), these units (or a subset) can be managed on a habitat scale. The isolated nature of the habitats is likely to produce species eliminations through time given even the most benign of “natural conditions,” under the basic laws of island biogeography. Therefore, these habitats (as surrogates for the localized species themselves) need to be managed on a landscape scale. Assuming that these habitats will do well through time is very different from actually establishing quantitatively that they will.

Summary for Part III:

This list of topics represents a prioritization of the subjects requiring analysis. It is by no means meant to be all-inclusive. Ecosystem management requires information-rich databases. There are many times more species of native arthropods than all the vertebrates, higher plants and macro-fungi combined. These 4 subjects represent consensus priorities where the next steps in arthropod analysis should be focussed.
Footnotes:

[1] John Richardson, Department of Entomology, University of British Columbia, Vancouver, BC.

[2] Christine Niwa, Research Scientist, Forest and Range Experiment Station, US Forest Service, Corvallis, OR.

[3] Bruce Caldwell, Department of Fish and Wildlife, Oregon State University, Corvallis, OR.

[4] William Stephen, Department of Entomology, Oregon State University, Corvallis, OR.

[5] Nancy Diaz, Regional Office, US Forest Service, Portland, OR.

[6] Steven Gisler, Ph.D. graduate student, Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR.

[7] Nancy Rappaport, US Forest Service Experiment Station, Albany, CA.

[8] Paul Hammond, Department of Entomology, Oregon State University, Corvallis, OR.

[9] Jeffrey Miller, Department of Entomology, Oregon State University, Corvallis, OR.

[10] Jiquan Chen, Department of Environmental Sciences, Michigan Technical University, Kalamazoo, MI.

[11] Alex Farrand, Master of Science graduate student, Department of Fish and Wildlife, Oregon State University, Corvallis, OR.

[12] Leif Gundersen, Master of Science graduate student, Department of Environmental Sciences, Oregon State University, Corvallis, OR; and Jessica Rykken, Ph.D. graduate student, Department of Entomology, Oregon State University, Corvallis, OR.

[13] David Kavanaugh, Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, CA.

[14] Mark Harmon, Department of Forest Science, Oregon State University, Corvallis, OR.

[15] Joan Hagar, Ph.D. graduate student, Department of Forest Science, Oregon State University, Corvallis, OR.

Literature Cited:

Anderson, J.M. 1975. Succession, diversity and trophic relationships of some soil animals in decomposing leaf litter. Journal of Animal Ecology 44: 475‑495.

Anderson, J.M. 1988. Invertebrate mediated transport processes in soils. Agriculture, Ecosystems and Environment 24: 5‑19.

Anderson, J.M., S.A. Huish, P. Ineson, M.A. Leonard, and P.R. Splatt. 1985. Interactions of invertebrates, microorganisms and tree roots. In: Fitter, A.H., ed. Ecological interactions in the soil. Oxford, England: Blackwell Science; 377‑392.

Anderson, I.M. and P. Ineson. 1982. A soil microcosm system and its application to measurements of respiration and nutrient leaching. Soil Biology and Biochemistry 14: 415‑416.

Antvogel, H. and A. Bonn. 2001. Environmental parameters and microspatial distribution of insects: a case study of carabids in an alluvial forest. Ecography 24: 470-482.

Axelrod, D.I. 1976. History of the conifer forests, California and Nevada. University of California Publications in Botany 70:1-62.

Bal, L. 1982. Zoological ripening of soils. Wageningen, the Netherlands: Centre for Agricultural Publishing and Documentation; 365 p.

Barois, I., G. Villemin, P. Lavelle, and F. Toutain. 1993. Transformation of the soil structure through Pontoscolex corethrurus (Oligochaeta) intestinal tract. Geoderma 56: 57‑66.

Baumeister, N. 2002. The nutritional ecology of millipedes in Pacific Northwest conifer forests. Ph.D. Thesis, Oregon State University, Corvallis, OR.

Bell, G. 2001. Neutral macroecology. Science 293: 2413-2418.

Benckiser, G. 1997. Fauna in Soil Ecosystems: recycling processes, nutrient fluxes and agricultural production. Dekker Press, Monticello, N.Y.

Benedict, E.M. and E.G. McEvoy. 1995. Malheur cave status survey: the distribution, habitat and status of the Malheur isopod, Amerigoniscus malheurensis Schulz and the Malheur pseudoscorpion, Apochthonius malheuri Benedict and Malcolm with notes on special species. Unpublished report on file with: Natural Heritage Program, The Nature Conservancy, Portland, OR.

Bird, S., R.N. Coulson, and D.A. Crossley, Jr. 2000. Impacts of silvicultural practices on soil and litter arthropod diversity in a Texas pine plantation. Forest Ecology and Management 131(1-3):65-80.

Bond, W.J. 1994. Do mutualisms matter? Assessing the impact of pollinator and disperser disruption on plant extinction. Philosophical Transactions of the Royal Society, London B 334:83-90.

Bouche, M.B. 1975. Action de la faune sur les états de la matière organique dans les ecosystemes. In: Kilbertius, G., ed. Biodégradation et Humification. Sarrugemines, France: Pierron; 157‑168.

Brenner, G. J. 2000. Riparian and adjacent upslope beetle communities along a third order stream in the western Cascade Mountain Range, OR. Ph.D. Thesis, Oregon State Univ.

Brown, K.S., Jr. 1991. Conservation of neotropical environments: insects as indicators. IN: Collins, N.M. and J.A. Thomas (eds.) Conservation of insects and their habitats. Academic Press, London.

Brown, K.S., Jr. 1997. Diversity, distributiuon and sustainable use of Neotropical forests: insects as bioindicators for conservation monitoring. Journal of Insect Conservation 1:25-42.

Bury, R.B. and C.A. Pearl. 1999. Klamath-Siskiyou herpetofauna: biogeographic patterns and conservation strategies. Natural Areas Journal 19:341-350.

Cadenasso, M.L. and S.T.A. Pickett. 2001. Effect of edge structure on the flux of species into forest interiors. Conservation Biology 15(1):91-97.

Canterbury, G.E., T.E. Martin, D.R. Petit, L.J. Petit, and D F. Bradford. 2000. Bird communities and habitat as ecological indicators of forest condition in regional monitoring. Conservation Biology 14:544-558.

Carcamo, H.A., T.A. Abe, C.E. Prescott, F.B. Holl, and C.P. Chanway. 2000. Influence of millipedes on litter decomposition, N mineralization, and microbial communities in a coastal forest in British Columbia, Canada. Canadian Journal of Forest Research 30: 817-826.

Chase, M.K., W.B. Kristan, III, A.J. Lynam, M.V. Price, and J.T. Rotenberry. 2000. Single species as indicators of species richness and composition in California coastal sage scrub birds and small mammals. Conservation Biology 14:474-487.

Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microclimates among clearcut, edge, and interior of old-growth Douglas-fir forest. Agricultural and Forest Meteorology 63:219-237.

Coleman, David C. and D.A. Crossley, Jr. 1996. Fundamentals of soil ecology. New York: Academic Press; 202 p.

Collinge, S.K., M. Holyoak, C.B. Barr, and J.T. Marty. 2001. Riparian habitat fragmentation and population persistence of the threatened valley elderberry longhorn beetle in central California. Biological Conservation 100: 103-113.

Crowson, R.A. 1981. The biology of the Coleoptera. London: Academic Press.

Curl, E.A. and B. Truelove. 1986. The rhizosphere. Berlin: Springer‑Verlag; 288 p.

Daniel, S.L., T.M. Jamieson, L.D. Hoover, E.A. McGee, G. DeNitto, and R.M. Creasy. 1995. A field guide to serpentine plant associations and sensitive ploants in northern California. USDA Forest Service, Washington, DC.

Dawod, V. and E.A. Fitzpatrick. 1993. Some population sizes and effects of the Enchytraeidae (Oligochaeta) on soil structure in a selection of Scottish soils. Geoderma 56: 173‑178.

Deharveng, L. 1996. Soil Collembola diversity, endemism, and reforestation: a case study in the Pyrenees (France). Conservation Biology 10:74-84.

DellaSala, D.A., S.B. Reid, T.J. Frest, J.R. Strittholt, and D.M. Olson. 1999. A global perspective of the biodiversity of the Klamath-Siskiyou ecoregion. Natural Areas Journal 19:300-319.

de Ruiter, P.C.; J.C. Moore, K.B. Zwart, L.A. Bouwman, J. Hassink, J. Bloem, J.A. de Vos, J.C.Y. Marinissen, W.A.M. Didden, G. Lebbink, G., and L. Brussaard. 1993. Simulation of nitrogen mineralization in the below ground food webs of two winter wheat fields. Journal of Applied Ecology 30: 95‑106.

Desender, K. 1994. Carabid beetles: ecology and evolution. Kluwer Academic Press, Dordrecht.

Dimling, J. and C. McCain. 1996. Willamette National Forest Special Habitat Management Guide. 43 pp. unpublished report.

Dobson, F.S., J. Yu, and A.T. Smith. 1995. The importance of evaluating rarity. Conservation Biology 9:1648-1651.

Edwards, C.A. and J.R. Lofty. 1972. Biology of earthworms. London: Chapman and Hall; 333 p.

Ehrlich, Paul

Erwin, T.L., G.E. Ball, and D.R. Whitehead. 1979. Carabid beetles: their evolution, natural history and classification. Proc. 1^st Int. Symp. Carabidology, Smithsonian Institution, Washington, D. C. Aug 21-25, 1976. Junk Publ., The Hague.

Estrada-Venegas, E.G. 1995. Soil arthropods in the central Cascades: slash burning effects and biology of some species. M.Sci. Thesis, Oregon State University.

Faith, D.P. and P.A. Walker. 1996. Environmental diversity: on the best-possible use of surrogate data for assessing the relative biodiversity of sets of areas. Biodiversity and Conservation 5: 399-415.

FEMAT, Forest Ecosystem Management Assessment Team. 1993. Forest ecosystem management: an ecological, economic and social assessment. US Department of Interior, US Department of Agriculture, US Department of Commerce, and US Environmental Protection Agency, Washington DC.

Finlay, R.D. 1985. Interactions between soil microarthropods and endomycorrhizal associations of higher plants. In: Fitter, A.H.; Atkinson, D.; Read, D.J.; Usher, M.B., eds. Ecological interactions in the soil. Oxford: Blackwell Science; 319‑332.

Fitter, A.H., D. Atkinson, D.J. Read, and M. B. Usher. 1985. Ecological Interections in the Soil. Blackwell Sci., Oxford.

Fox, J.L. 1994. Microbial diversity: low profile, immense breadth. American Society of Microbiology News 60: 533‑536.

France, R.L. 1997. Macroinvertebrate colonization of woody debris in Canadian shield lakes following riparian clearcutting. Conservation Biology 11:513-521.

Franklin, J.F. and C.T. Dyrness. 1973. Natural Vegetation of Oregon and Washington. USDA Forest Service General Technical Report PNW-8, Portland, OR,

Furniss, R.L. and V.M. Carolin. 1977. Western Forest Insects. USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, Misc. Pub. No. 1339. 654 pp.

Gavrilets, S. 1999. A dynamical theory of speciation on holey adaptive landscapes. American Naturalist 154:1-22.

Hansen, A.J., T.A. Spies, F.J. Swanson and J.L. Ohmann. 1991. Conserving biodiversity in managed forests, lessons from natural forests. Bioscience 41:382-392.

Haskell, D.G. 2000. Effects of forest roads on macroinvertebrate soil fauna of the southern Apalachian Mountains. Conservation Biology 14:57-63.

Hawksworth, D.L. 1991. The fungal dimension of biodiversity: magnitude, significance and conservation. Mycological Research 95: 641‑655.

Haynes, R.W., R.T. Graham, and T.M. Quigley (editors). 1996. A framework for ecosystem management in the Interior Columbia Basin. Gen. Tech. Rpt. PNW-GTR-374. USDA Forest Service, Pacific Northwest Research Station, Portland, Oregon. 68 pp.

Heliölä, J., M. Koivula, and J. Niemelä. 2001. Distribution of carabid beetles (Coleoptera, Carabidae) across a boreal forest-clearcut ecotone. Conservation Biology 15(2):370-377.

Heyborne, W.H. 2000. Ground dwelling betles in early and late successional forests in the western Cascades of Oregon. M.Sci. Thesis, Oregon State University.

Hinchliff, J. 1994. An atlas of Oregon butterflies. The Evergreen Aurelians.

Hopkin, S.P. and H.J. Read. 1992. The biology of millipedes. Oxford. England: Oxford Scientific Publications; 233 p.

Hughes, J.B., G.C. Daily, and P. Ehrlich. 2000. Conservation of insect diversity: a habitat approach. Conservation Biology 14: 1788-1797.

Hunt, H.W., D.C. Coleman, and E.R. Ingham. 1987. The detrital food‑web in a short‑grass prairie. Biology and Fertility of Soil 3: 57‑68.

Ingham E.R. and A.R. Moldenke. 1995. Microflora and microfauna on stems and trunks. In: Gartner, B., ed. Plant stems: physiology and functional morphology. New York: Academic Press; 241‑256.

Izco, J. 1998. Types of rarity in plant communities. Journal of Vegetation Science 9:641-646.

Johnson, D.H. and T.A. O’Neil. 2001. Wildlife-habitat relationships in Oregon and Washington. Oregon State University Press, Corvallis, OR.

Kalisz, P.J. and J.E. Powell. 2000. Effects of prescribed fire on soil invertebrates in upland forests on the Cumberland Plateau of Kentucky, USA. Natural Areas Journal 20:336-341.

Kerr, J.T. 1997. Species richness, endemism, and the choice of areas for conservation. Conservation Biology 11:1094-1100.

Kubiena, W.L. 1938. Micropedology. Arms, IA: College Press.

Kuikman, P.J., A.G. Jansen, J.A. van Veen, and A.J.B. Zehnder. 1990. Protozoan predation and the turnover of soil organic carbon and nitrogen in the presence of plants. Biology and Fertility of Soils 10: 22‑28.

Lang, F.A. 1999. Klamath-Siskiyou natural history. Natural Areas Journal 19(4):298-299.

Lattin, J.D. and A.R. Moldenke. 1992. Ecologically sensitive invertebrate taxa of Pacific Northwest old-growth forests. Report to the Northern Spotted Owl Recovery Team on Other Species and Ecosystems Committee. Portland, OR. U.S. Department of the Interior.

Lavelle, P. 1988. Earthworm activities and the soil system. Biology and Fertility of Soil 6: 237‑251.

Lavelle, P. 1994. Faunal activities and soil processes: adaptive strategies that determine ecosystem function. IN: Transactions of the 15th World Congress of Soil Science; 1994 July 10‑16; Acapulco, Mexico. Volume 1. Mexico City, Mexico: Mexican Society of Soil Science; 189‑220.

Lee, K.E. 1985. Earthworms: their ecology and relationships with soils and land use. Sydney, Australia: Academic Press; 411 p.

Leach, R. 1972. Monograph of the Amaurobiidae of North America. Mem. Ent. Soc. Canada, vol. 84.

Lewandowski, A.M. (editor) 1999. Soil Biology Primer. Natural Resounce Conservation Service, US Department of Agriculture, Publication PA-1637.

Lindroth, C.H. 1961-69. The ground-beetles of Canada and Alaska. Supplements 1-6. Opuscula Entomologica. p 1-1192.

Linsley, E.G. 1961-63. The Cerambycidae of North America. Parts 1-3. University of California Press, Berkeley, CA.

Madson, S.L. 1997. Correlation between structural heterogeneity and arthropod biodiversity: implications for forest management of Pacific Northwest forests. M.Sci. Thesis, Oregon State University.

Manel, S., S.T. Buckton, and S.J. Ormerod. 2000. Testing large-scale hypotheses using surveys: the effects of land use on the habitats, invertebrates and birds of Himalayan rivers. Journal of Applied Ecology 37(5):756-770.

Marcot, B.G., L.K. Croft, J.F. Lehmkuhl, R.H. Naney, C.G. Niwa, W.R. Owen, and R.E. Sandquist. 1998. Macroecology, paleoecology, and ecological integrity of terrestrial species and communities of the interior Columbia River Basin and portions of the Klamath and Great Basins. General Technical Report PNW-GTR-410. USDA Forest Service, Portland OR. 131 pp.

Martin, A. and J.C.Y. Marinssen. 1993. Biological and physio‑chemical processes in excrements of soil animals. Geoderma 56: 331‑347.

Martin, M. 1987. Invertebrate‑microbial interactions: ingested fungal enzymes in arthropod ecology. Ithaca, NY: Cornell University Press; 148 p.

Matheson, A. 1996. The conservation of bees. Academic Press, London.

May, R.M. 1975. IN: Ecology and evolution of Communities. M.L. Cody and J.L. Diamond, eds. Harvard University Press, Cambridge.

McAlpine, J.F. (editor). 1987. Manual of the Nearctic Diptera. Monograph 28, Research Branch, Agriculture Canada.

McIver, J.D., A.R. Moldenke, and G.L. Parsons. 1990. Litter spiders as “biological indicators” of recovery after clear-cutting in a western coniferous forest. Northwest Environmental Journal 6: 410-412.

McIver, J.D., G.L. Parsons, and A.R. Moldenke. 1992. Litter spider succession after clearcutting in a western coniferous forest. Canadian Journal of Forest Research 22: 964-982.

Miller, J.C. 1990. Field assessment of the effects of a microbial pest control agent on non-target Lepidoptera. American Entomologist 36: 135-139.

Miller, J.C. 1993. Insect natural history, multi-species interactions and biodiversity in ecosystems. Biodiversity and Conservation 2: 233-241.

Moldenke, A.R. 1975. Niche specialization and species diversity along a California transect. Oecologia 21: 219-242.

Moldenke, A.R. 1976a. California pollination ecology and vegetation types. Phytologia 34: 305-361.

Moldenke, A.R. 1976b. Evolutionary history and diversity of the bee faunas of Chile and Pacific North America. Wasmann J. Biol. 34: 147-178.

Moldenke, A.R. 1979a. Pollination ecology in the Sierra Nevada. Phytologia 42: 223-282.

Moldenke, A.R. 1979b. Pollination ecology as an assay of ecosystemic structure: convergent evolution in California and Chile. Phytologia 42: 415-454.

Moldenke, A.R. 1979c. Host-plant coevolution and the diversity of bees in relation to the flora of North America. Phytologia 43: 357-419.

Moldenke, A.R. 1979d. The role of host-plant selection in bee speciation processes. Phytologia 43: 433-460.

Moldenke, A.R. 1990. One‑hundred twenty‑thousand little legs. Wings 15: 11 ‑ 14.

Moldenke, A.R. 1999. Soil-dwelling arthropods: their diversity and functional roles. P. 33-44. IN: Proceedings: Pacific Northwest Forest and Rangeland Soil Organism Symposium, March 17-19, 1998; ed.: R.T. Meurisse, W.G. Ypsilantis, and C. Seybold. Gen. Tech. Rept. PNW-GTR-461, US Dept. Agric. Forest Service, Pacific Northwest Research Station, 215 pp.

Moldenke, A.R., N. Baumeister, E. Estrada-Venegas, and J. Wernz. 1994. Linkages between soil biodiversity and above-ground plant performance. Trans. 15^th World Cong. Soil Sci., Vol 4a (III): 186-204, Acapulco.

Moldenke, A.R. and J.D. Lattin. 1990a. Density and diversity of soil arthropods as "biological probes" of complex soil phenomena. The Northwest Environmental Journal 6:409-410.

Moldenke, A.R. and J.D. Lattin. 1990b. Dispersal characteristics of old-growth soil arthropods: the potential for loss of diversity and biological function. Northwest Environmental Journal 6: 408-409.

Moldenke, A.R., M. Pajutee, and E. Ingham. 1999. The functional roles of forest soil arthropods: the soil is a lively place. Pp. 7-22, IN: Proceedings of the California Forest Soils Council Conference on Forest Soils Biology and Forest management, Feb. 23-24, 1996.

Moore, I. and E.F. Legner. 1975. A catalogue of the Staphylinidae of America north of Mexico. University of California Division of Agricultural Sciences, Special Publication 3015: 1-514.

Moore, I. and E.F. Legner. 1975. An illustrated guide to the genera of the Staphylinidae of America north of Mexico. University of California Division of Agricultural Sciences, Special Publication 4093: 1-332.

Moore, J.C., D.E. Walter, and H.W. Hunt. 1988. Arthropod regulation of micro- and mesobiota in belowground food webs. Annual Review of Entomology 33: 419-439.

Morrison, M.L. 1986. Bird populations as indicators of environmental change. Pp. 429-451 in: R. F. Johnston, ed. Current Ornithology, Vol. 3. Plenum Press, New York.

Murphy, D.D. and B.A. Wilcox. 1986. Butterfly diversity in natural habitat fragments: a test of the validity of vertebrate-based management. pp. 287-292 IN: Verner, J, M.L. Morison and C.J. Ralph (eds.) Wildlife 2000, modeling habitat relationships of terrestrial vertebrates. University of Wisconsin Press, Madison, WI.

Nakano, S. and M. Murakami. 2001. Reciprocal subsidies: dynamic interdependence between terrestrial and aquatic foodwebs. Proceedings of the National Academy of Sciences 98: 166-170.

Neher, D.A., K.N. Easterling, D. Fiscus, and C.L. Campbell. 1998. Comparison of nematode communities in agricultural soils of North Carolina and Nebraska. Ecological Applications 8(1):213-223.

New, T.R. 1999. Limits to species focusing in insect conservation. Annals of the Entomological Society of America 92:853-860.

Newell, K. 1984. Interaction between two decomposer basidiomycetes and Collembola under sitka spruce. I & II. Soil Biology and Biochemistry 16: 227‑240.

Newton, A.F., M.K. Thayer, J.S. Ashe, and D.S. Chandler. 2000. Staphylinidae IN: R. Arnett (ed.) American Beetles. CRC Press, Boca Raton, FL.

Niemela, J., D. Langor, and J.R. Spence. 1993. Effects of clear-cut harvesting on boreal ground-beetle assemblages (Coleoptera: Carabidae) in western Canada. Conservation Biology 7:551-561.

Nilsson, S.G., U. Arup, R. Baranowski, and S. Ekman. 1995. Tree-dependent lichens and beetles as indicators in conservation forests. Conservation Biology 9:1208-1215.

Olson, D.H., S.S. Chan, P. Cunningham, B. Hansen, A. Moldenke, R. Progar, P.S. Muir, B. McCune, A. Rosso, and E.B. Peterson. 2000a. Characterizing managed headwater forests - integration of stream, riparian, and upslope habitats and sepecies in western Oregon: Companion projects to the BLM Density Management Studies. Pp. 539-540 In: Proceedings of the Society of American Foresters 1999 National Convention, 11-15 September, Portland, OR. SAF Publication 00-1, Bethesda, MD; ISBN 0-939970-81-3.

Olson, D.H., S.S. Chan, G. Weaver, P. Cunningham, A. Moldenke, R. Progar, P.S. Muir, B. McCune, A. Rosso, and E.B. Peterson. 2000b. Characterizing stream, riparian, upslope habitats and species in Oregon managed headwater forests. Pp. 83-88 In: Wiggington, J. and R. Beschta (eds.). Riparian Ecology and Management in Multi-Land Use Watersheds. International conference of the American Water Resources Association, 30 August, Portland, OR. AWRA Publication TPS-00-2, Middleburg, VA. 616 pp.

Olson, D.H., J.C. Hagar, A.B. Carey, J.H. Cissel, and F.J. Swanson. 2000c. Wildlife communities in westside and high montane forest. Chapter 7, pp. 187-212 In: Johnson, D.H. and T.A. O'Neil (Managing Directors), Wildlife-habitat relationships in Oregon and Washington, Oregon State University Press, Corvallis, OR.

O'Neil, T.A., D.H. Johnson, C. Barrett, M. Trevithick, K.A. Bettinger, C. Kiilsgaard, M. Vander Heyden, E.L. Greda, D. Stinson, B.G. Marcot, P.J. Doran, S. Tank, and L. Wunder. 2001. Matrixes for wildlife-habitat relationships in Oregon and Washington. CD-ROM. in: D. H. Johnson and T. A. O'Neil, ed. Wildlife-habitat relationships in Oregon and Washington. Oregon State University Press, Corvallis OR.

Opler, P.A. and J.D. Lattin. 2001. Narrative on arthropods and annelid worms of old-growth and late successional forests, mature riparian woods and of coarse woody debris associated arthropods within the range of the northern spotted owl. http://www.mesc.usgs.gov/pubs/online/arthropods/ arthropods_annelids.

Pacific Analytics. 2001. Literature synthesis and recommendations for general surveys for forest understory and canopy gap herbivores pertinent to the southern range of the northern spotted owl. Unpublished report by G. Brenner and J. Halaj. http://www.fs.fed.us/r6/nr/fid/pubsweb/litsurvey/sect7.htm.

Palleroni, N.J. 1994. Some reflections on bacterial diversity. American Society of Microbiology News 60: 537‑540.

Pankhurst, C., B. Doube, and V. Gupta, (editors). 1997. Biological indicators of soil health. Oxford University Press, New York.

Paoletti, M.G., W. Foissner, and D. Coleman. 1993. Soil Biota, Nutrient Cycling and Farming Systems. Lewis Publ., London.

Parsons, G.L., G. Cassis, A.R. Moldenke, J.D. Lattin, N. H. Anderson, J.C. Miller, P. Hammond, and T D. Schowalter. 1991. Invertebrates of the H. J. Andrews Experimental Forest, Western Casde Range, Oregon. V: An annotated list of insects and other arthropods. Gen Tech. Report PNW-GTR-290, US Dept. Agric. Forest Service, Pacific Northwest Research Station, Corvallis, OR.

Pawluk, S. 1985. Soil micromorphology and soil fauna: problems and importance. Questiones Entomologicae 21: 473‑496.

Pearson, D.L. and F. Cassola. 1992. World-scale species richness patterns of tiger beetles (Coleoptera: Cicindelidae): indicator taxon for biodiversity and conservation studies. Conservation Biology 6: 376-391.

Petersen, H. and M. Luxton. 1982. A comparative analysis of soil fauna populations and their role in decomposition processes. Oikos 39: 287‑388.

Prendergast, J.R., R.M. Quinn, J.H. Lawton, B.G. Eversham, and D.W. Gibbons. 1993. Rare species, the coincidence of diversity and conservation strategies. Nature 365: 335-337.

Pongprasert, W. 2000. Phylogenetic relationships of the bumblebees (Apoidae; Hymenoptera) in the Pacific Northwest of America inferred from mitochondrial cytochrome oxidase gene sequences. PhD. Thesis, Oregon State University, Corvallis, OR.

Postle, A., J. Majer, and D. Bell. 1991. A survey of selected soil and litter invertebrate species from the northern jarrah (Eucalyptus marginata) forest of western Australia, with particular reference to soil-type, stratum, seasonality and the conservation of forest fauna. Pp. 193-204 in: D. Lunney, ed. Conservation of Australia's forest fauna. The Royal Zoological Society of New South Wales, New South Wales, Australia.

Powers, D.L. Hauxwell, and G.M. Nakamura. 1999. Gen. Tech. Report PSW-GTR-178. US Dept. Agric. Forest Service, Pacific Southwest Research Sta, Redding, CA. 113 pp.

Prendergast, J.R., R.M. Quinn, J.H. Lawton, B.C. Eversham, and D.W. Gibbons. 1993. Rare species, the coincidence of diversity hotspots and conservation strategies. Nature 365: 335-337.

Progar, R. and A. Moldenke. 2001a. Insect production from temporary and perennially flowing headwater streams in western Oregon. Journal of Freshwater Ecology (submitted).

Progar, R. and A. Moldenke. 2001b. Insect production from clearcut and closed canopy headwater streams in western Oregon. Journal of the North American Benthological Society (submitted).

Rabinowitz, D. 1981. Seven forms of rarity. Pp. 205-217 in: H. Synge, ed. The biological aspects of rare plant conservation. Wiley, Chichester.

Raphael, M. 1988. Long-term trends in abundance of amphibians, reptiles, and mammals in Douglas fir forests of northwest California. Gen. Tech. Rpt. RM-166. (pp. 23-31) In: R. C. Szaro, K. E. Severson, and D. R. Patton (Ed.). Management of amphibians, reptiles, and small mammals in North America. Rocky Mountain Forest and Range Experiment Station, Flagstaff, Arizona.

Rathcke, B. and E.S. Jules. 1993. Habitat fragmentation and plant-pollinator interactions. Current Science 65: 273-277.

Risser, P.G. (editor). 2000. Oregon State of the Environment Report 2000: Statewide Summary. Oregon Progress Board, Salem, OR.

Robbins, R.K., and P.A. Opler. 1997. Butterfly diversity and a preliminary comparison with bird and mammal diversity. Pp. 69-82 in: M. L. Reaka-Kudla, D. E. Wilson, and E. O. Wilson, ed. Biodiversity II: understanding and protecting our biological resources. John Henry Press, Washington, D.C.

Rundio, D.E. and D.H. Olson. 2001. Predator-prey relations between southern torrent salamanders (Rhyacotriton variegatus) and Pacific giant salamander (Dicamptodon tenebrosus) larvae. Journal of Herpetology 35:133-136.

Rusek, J. 1985. Soil microstructures—contributions on specific soil organisms. Questiones; Entomologicae 21: 497‑514.

Ruzicka, V. and J. Bohac. 1993. The utilization of epigeic invertebrate communities as bioindicators of terrestrial environmental quality. pp. 79-86 IN: Salanki, J, D. Jeffrey and G.M. Hughes. Biological monitoring of the environment. A manual of methods. CAB International, London.

Rykken, J.J., D.E. Capen, and S.P. Mahabir. 1997. Ground beetles as indicators of land type diversity in the Green Mountains of Vermont. Conservation Biology 11: 522-530.

Sayer, J.D. and D.A. Thornburgh. 1977. Montane and subalpine vegetation of the Klamath Mountains. pp 669-732 IN: Barbour, MG. and J. Majors. Terrestrial vegetation of California. Wiley Interscience, New York.

Schowalter, T.D. 1989. Canopy arthropod community structure and herbivory in old-growth and regenerating forests in western Oregon. Canadian Journal of Forest Research 19: 318-322.

Schowalter, T.D. 1995. Canopy arthropod communities in relation to forest age and alternative harvest practices in western Oregon. Forest Ecology and Management 78: 115-125.

Schowalter, T.D. 2000. Insect Ecology: an Ecosystem Approach. Academic Press, New York.

Schowalter, T.D., B.A. Caldwell,; S.E. Carpenter, R.P. Griffiths, M.E. Harmon, E.R. Ingham, R.G. Kelsey, J.D. Lattin, and A.R. Moldenke. 1988. Decomposition of fallen trees: effects of initial conditions and heterotroph colonization rates. In: Singh, K.P.; Singh, J.S., (eds.) Tropical ecosystems: ecology and management. New Delhi, India: Wiley Eastern Publishers; 373‑383.

Seastedt, T.R. 1984. The role of microarthropods in decomposition and mineralization processes. Annual Review of Entomology 29: 25-46.

Setala, H. and V. Huhta. 1991. Soil fauna increase Betula pendula growth: laboratory experiments with coniferous forest floor. Ecology 72: 665‑671.

Shaw, C.H., H. Lundkvist, A.R. Moldenke, and J.R. Boyle. 1991. The relationships of soil fauna to long-term forest productivity in temperate and boreal ecosystems: processes and research strategies. Pp. 39-77, IN: W. J. Dyck and C. A. Mees (eds.) Long-term field trials to assess environmental impacts of harvesting. Forest Research Inst., Rotorua, N.Z. FRI Bulletin No. 161.

Spence, J.R. and J.K. Niemela. 1994. Sampling carabid assemblages with pitfall traps: the madness and the method. Canadian Entomologist 126: 881-894.

Spira, T.P. 2001. Plant-pollinator interactions: a threatened mutualism with implications for the ecology and management of rare plants. Northwest Areas Journal 21: 71-81.

Stebbins, G.L. and J. Major. 1965. Endemism and speciation in the California flora. Ecological Monographs 35: 1-35.

Stephen, W.P. 1957. Bumbleebees of northwestern North America. Oregon Agr. Expt Sta. Tech. Bul. 40.

Stork, N.E. 1990. The role of ground beetles in ecological and environmental studies. Intercept press, Andover.

Strohecker, H.F., W.W. Middlekauff and D.C. Rentz. 1968. Grasshoppers of California. Bulletin of the California Insect Survey #10.

Swengel, S.R. and A.B. Swengel. 1999. Correlations in abundance of grassland songbirds and prairie butterflies. Biological Conservation 90:1-11.

Tappeiner, J.C. II, D.H. Olson, and C.R. Thompson. 2000. Density management studies of western Oregon. Pp. 556-557 In: Proceedings of the Society of American Foresters 1999 National Convention, 11-15 September, Portland, OR. SAF Publication 00-1, Bethesda, MD; ISBN 0-939970-81-3.

Thiele, H.U. 1977. Carabid beetles in their environments. Springer-Verlag, Berlin.

Tiedje, J.M. 1994. Microbial diversity: of value to whom? American Society of Microbiology News 60: 524‑525.

USDA (US Derpartment of Agriculture). 1988. Final Supplement to the environmental impact statement for an amendment to the Pacific Northwest regional guide: spotted owl guidelines. USDA, Forest Service, PNW Region, Portland, OR.

USDA (US Derpartment of Agriculture). 1994. Record of decision for amendments to Forest Service and Bureau of Land Management to the survey and manage, protection buffer and other mitigation measures standards and guidelines. USDA, Forest Service, PNW Region, Portland, OR.

USDA (US Derpartment of Agriculture). 2000. Final supplemental environmental impact statement for amendment to the survey and manage, protection buffer and other mitigation measures standards and guidelines. USDA, Forest Service, PNW Region, Portland, OR.

Vance-Borland, K.W. 1999. Physical habitat classification for conservation planning in the Klamath Mountain region. Ph.D. Thesis. Oregon State University.

Visser, S. 1985. The role of soil invertebrates in determining the composition of soil microbial communities. In: Fitter, A.H., ed. Ecological interactions in the soil. Oxford, England: Blackwell Science; 297‑317.

Webb, D.P. 1977. Regulation of deciduous forest decomposition by soil arthropod feces. In: Mattson, William J., ed. The role of arthropods in forest ecosystems. New York: Springer; 57‑69.

Welsh, H.H. 1990. Relictual amphibians and old-growth forest. Conservation Biology 4:309-319.

Whittaker, R.H. 1960. The vegetation of the Siskiyou Mountains of Oregon and California. Ecological Monographs 30:279-338.

Whittaker, R.H. 1961. Vegetation history of the Pacific Coast states and the central significance of the Klamath Region. Madrono 16: 5-23.

Wikars, L. O. and J. Schimmel. 2001. Immediate effects of fire-severity on soil invertebrates in cut and uncut pine forests. Forest Ecology and Management 141:189-200.

Williams, N.M., R.L. Minckley, and F.A. Silveira. 2001. Variation in native bee faunas and its implications for detecting community changes. Conservation Ecology 5:7 [online]. Web: http://www.consecol.org/vol5/iss1/art7

Willis, H.L. 1968. Artificial key to the species of Cicindela of North America (north of Mexico). Journal of the Kansas Entomological Society 41: 303-317.

Wilson, M.V. 1988. Within-community vegetation structure in the conifer woodlands of the Siskiyou Mountains, Oregon. Vegetatio 78: 61-72.

Wolfe, J.A. 1969. Neogene floristic and vegetational history of the Pacific Northwest. Madrono 20: 83-110.

Wolfe, J.A. 1978. A paleobotanical interpretation of Tertiary climates in the northern hemisphere. American Scientist 66: 694-703.

APPENDIX ONE:

GLOSSARY OF TERMS

Allopatric – two species which live in non-overlapping geographic distributions.

Alpha-diversity – the total number of species inhabiting a particular small sample site.

Beta-diversity – (variously employed in the literature) generally referring to the total number of species within a region encompassing several different habitat types.

Biogeographically significant higher taxa (BSHT)—Any taxon, including or above the species-group or subgenus level (e.g., subgenera, genera, tribe, family), that occurs almost entirely within the Vancouveran Bioregion. These taxa may either be paleoendemics associated with the Klamath Mountain Miocene orogeny or neoendemic adaptive radiations of phyletic lineages that reached the region subsequent to glaciation. (= the platypuses and giraffes of the Pacific Northwest)

Community composition – the spectrum of relative abundances of individual species within a community.

Cryptic-sensitive species—Microsite-adapted sensitive species which are liable to be characterized by broad distributions throughout the entire region, yet often infrequent in museum collections. Their apparent scarcity may be real because they are actually associated with extremely limited microsites rather than the generalized forest mosaic; as such, they represent a critical subset of the generally poorly documented entomofauna of the Pacific Northwest forests. Such species associated with environments under a forest canopy (unlike meadows and balds) are especially good candidates for cryptic-sensitives since it is easy to overlook their less obvious microhabitat requirements.

Endemic—occurring only within a limited geographic region, i.e., the species is endemic to the Oregon Cascades.

Focal species—A species whose occurrence is used as a surrogate to indicate the presence of other species. The association of the species is not based on a priori knowledge of specific limiting environmental features shared between the species.

Gamma-diversity – refers to the total fauna of a large region, which is comprised of more than one biogeographic zone.

Higher taxa – nomenclaturial categories more inclusive than the species; generally used for subgenera, genera, tribes and families.

Indicator species—a useful taxon for inventory studies, used often indirectly as a surrogate for many others. A species that has a high degree of habitat specificity and can represent the presence of an otherwise difficult to distinguish environmental condition (i.e. Jeffrey Pine on serpentine soils). Often used invalidly as a direct correlate of population levels of associated species; been replaced in Forest Service literature by the more precise term “focal species”.

Keystone species—A species, which by its absence from a former habitat, causes an upset in an ecosystem process. Usually a species belonging to a species-depauperate functional guild, hence its absence affects the entire guild function.

Neoendemic—refers to a species with limited geographic distribution which has only recently (in geologic terms) invaded the region of concern, for instance, since the latest glaciation.