Patuxent Wildlife Research Center

NAAMP III Archive - community & regional
Home | Archive by Alphabetical Order | Archive by Category

Amphibian Inventory and Monitoring in the Pacific Northwest
(United States and Canada)

Deana H. Olson

Pacific Northwest Research Station
USDA Forest Service
3200 SW Jefferson Way, Corvallis, OR 97331

[ Abstract ]

Introduction
The need for inventory and long-term monitoring programs of amphibian populations is broadly recognized. Reports of apparently widespread amphibian population declines (e.g., Bury et al. 1980; Blaustein and Wake 1990; Phillips 1990; Wyman 1990; Wake 1991) have resulted in a critical review of the status of amphibian species globally (Vial and Saylor 1993). To assess declines, studies of distribution patterns and population dynamics over time need to be examined. The paucity of multi-site or multi-year amphibian datasets has been apparent for several years (e.g., Blaustein and Wake 1990), and although additional data have become available (review by Blaustein 1994), assessments of the status of amphibian populations have been difficult due to the rarity of comparable datasets and long-term studies, both within and among taxa or regions (Vial and Saylor 1993; Reed and Blaustein 1995). Reviews of the development of this issue (e.g., Pechmann and Wilbur 1994; Blaustein 1994; Blaustein et al. 1994a) call for long-term inventory and monitoring efforts to establish the collection of datasets that adequately assess the status of populations.

Vial and Saylor's (1993) compilation of global data on amphibian populations emphasized the importance of standardized techniques to assess the status of populations. Without clearly described and repeatable survey methods, they found it difficult to attempt an analysis of patterns of population declines among taxonomic groups, geographic areas, or among years at focal localities. The publication of a manual of amphibian survey methods (Heyer et al. 1994) greatly advanced the standardization of techniques for amphibians. Their presentation of ten primary and several supplemental methods covers the many inventory and monitoring approaches used globally for diverse amphibian assemblages.

Our intent with this handbook is to further extend prior methodologies (see Heyer et al. 1994) and offer new approaches field-tested by several biologists. We prioritize among methods for specific objectives, species, and lentic habitats (e.g., ponds, lakes, wetlands) that occur in the Pacific Northwest (PNW) of the United States. While Heyer et al. provides multiple methods that may pertain to numerous habitats, survey conditions and objectives, we have outlined common conditions and objectives with regard to PNW amphibians occurring in lentic habitats, and guide surveyors to a specific recommended suite of methods. Biologists that are trained in the identification of PNW amphibian species can follow this guide to implement standardized surveys for inventory and monitoring in the major lentic habitat types in the region. Although this guide has been based on our knowledge of PNW lentic habitats, it may have broader application to ponded waters and amphibian assemblages globally. We believe the approaches will be instructive for assessments of pond-dwelling amphibians elsewhere, especially in north temperate and montane habitats. Integration of methodologies proposed among regions (e.g., with the efforts of the North American Amphibian Monitoring Program) and taxonomic groups (e.g., desert anurans) are expected to develop. Likewise, the methodologies themselves may be expected to change as their effectiveness is further evaluated. With these caveats stated, this effort addresses the urgent call in the PNW region for a guide to standardized survey methodologies for pond-breeding amphibians.

INVENTORY AND MONITORING IN THE PACIFIC NORTHWEST

In the PNW, there is a dire need to standardize amphibian survey techniques. In particular, there are several factors that are collectively driving the large-scale planning and implementation of amphibian inventory and monitoring efforts (Figure 1).

The large extent of PNW public lands has placed the joint tasks of land management and species-conservation on the state and federal governments. For example, in the recent development of the Northwest Forest Plan (Record of Decision [ROD] 1994), which provides ecosystem protection and management designs for forests within the range of the northern spotted owl (Strix occidentalis), over 40% (24.3 million acres) of the lands were reported to be federally-administered. At this spatial scale, effective protection and management of forest-dwelling organisms can be addressed.

Numerous amphibians occur in or are closely-associated with western forests of this region, and are intended to be protected by the Northwest Forest Plan (Blaustein et al. 1995, Forest Ecosystem Management and Assessment Team [FEMAT] 1993, ROD 1994). For example, of the 31 native amphibian species occurring in Oregon and Washington, 27 occur in forested landscapes of the western portions of these states, 22 are restricted to this region, and 17 have been considered close-associates of old-growth forest conditions (Leonard et al. 1993, Walls et al. 1992, Blaustein et al. 1995, FEMAT 1993). A large proportion of these species have at-risk status, largely because they are rare endemics (Bury 1994) or due to concerns for declining populations. Mandates to protect forest-dwelling species on federal lands places much of the burden of amphibian conservation for the region on federal agencies and the local district offices implementing the Northwest Forest Plan.

Land conversion, urbanization, and development result in habitat loss for native biota, the single most overwhelming cause of amphibian losses globally (Vial and Saylor 1993). Conversion of lands that accompanies both urban and agricultural development has apparently had profound impacts on amphibian habitats in the Pacific Northwest. Loss and alteration of lentic habitats, in particular, is documented. For example, Benner and Sedell

1996) reported changes in the Oregon Willamette River basin since Euro-American settlement in the 1800s. Along a 58-mile section of the Willamette River, 45 to 70 percent of the channel length has been lost, changed from a multiple channel to an often single channel system (also Sedell and Frogatt 1984). Flooding patterns are changed; "A large portion of land that once was within the 10-year floodplain, and flooded multiple times during a 10-year period, is now infrequently flooded" (Benner and Sedell 1996). Lentic habitats provided by river side channels, wooded floodplain areas, and off-channel sloughs and swales have been eliminated. Likewise, in Washington State, loss of wetlands from development, in addition to current wetland management practices, are issues of concern. In 19 palustrine wetlands of the Washington Puget Sound basin, amphibian species richness was correlated with the degree of watershed urbanization and with the mean-monthly water level fluctuation. "Specifically, increases in the duration and frequency of flooding and changes in discharge rates...decreases the success of lentic breeding amphibians" (Richter and Azous 1995).

Demand for the maintenance of biodiversity is a strong factor driving inventory and monitoring programs. Knowledge of species diversity patterns is necessary to address protection, but for PNW amphibians there is limited knowledge of diversity at a variety of biological and spatial scales. For example, over half of the amphibian species in the PNW are found on State lists of sensitive species (OR) or species of concern (WA); many of these occur on State lists becausetheir current status is unknown (e.g., Marshall et al. 1992).

The number of species that are recognized in the region is changing rapidly as biochemical techniques are finding additional species-level distinctions (Good 1989, Good and Wake 1992, Green et al. 1996). As additional species distinctions are made, the status of populations with truncated ranges becomes an issue. For example, the previously known Olympic salamander (Rhyacotriton olympicus) was considered widely distributed from the Olympic Peninsula, Washington, to northern California. But now it is split into four species (Good and Wake 1992) and the status of each geographically restricted form becomes an issue of importance for biodiversity maintenance. Western spotted frogs (Rana pretiosa) are pond-breeding amphibians for which recent genetic assessment may affect state and federal status classification: the western form is genetically distinct ("species A", Green et al. 1996) and apparently is restricted in distribution to a few known sites in southwestern Washington and western Oregon, whereas the rest of this species-complex is more broadly distributed, occurring in 6 western states and Canada.

Further, at the spatial scales of a single site, a watershed, or a subregion, we often have poor knowledge of species occurrences because the animals tend to be poorly censused, cryptic, and patchily distributed. Moreover, distributions have been influenced by various types of anthropogenic disturbances over the last several decades. Many amphibian decline issues related to human disturbances are being dealt with in this region, including concerns for native endemic species in forests as well as other habitats (Leonard and McAllister 1996; McAllister and Leonard 1990, 1991; McAllister et al. 1993; Marshall et al. 1992; Walls et al. 1992; Blaustein et al. 1994a, 1994b, 1995; Olson et al. 1996, 1997). Recent survey efforts have improved our understanding of the geographic ranges of some species (e.g., Leonard and McAllister 1996; McAllister and Leonard 1990, 1991). However, inventories are needed to understand regional biodiversity patterns.The need to survey amphibians in lentic habitats, in particular, stems from the combined effects of the heterogeneous fauna and habitat types, and the historical neglect of this assemblage. There are three general amphibian assemblages in the PNW: 1) terrestrial obligates, salamanders which do not require aquatic habitats, 2) stream-dependent species, which live in or beside lotic waters for part or all of their life, and 3) "pond-breeders", which require lentic habitats for egg-laying and larval development and which may or may not be restricted to lentic sites as juveniles or adults (Cissel et al. in prep.). Over a third of PNW species are found in association with lentic habitats (Table 1). Although there are methods for each of these assemblages (e.g., Corn and Bury 1990, Bury and Corn 1991, Fellers and Freel 1995) a consensus of standardized survey approaches consistent with various survey objectives is being sought. Field data forms developed by the National Biological Service (e.g., page 181 in Heyer et al. 1994) have helped to initiate standardized data collection among assemblages, but methodologies specific to lentic habitats and species are lacking. Thus, our focus here on pond-breeders meets two objectives: 1) to present methods specific to PNW pond-breeding assemblages, and 2) to prioritize these methods to facilitate a standardized approach that can be fit to varying survey objectives, habitat conditions, and pond assemblages. Methods should be applicable to other regions.

Lentic habitats in the PNW are diverse in character. Climate, geomorphology, and vegetation patterns interact to produce an array of ecoregions in the area (USEPA 1995 (Map M-1), Omernik, 1987) within which numerous lentic habitat types occur. There are gradients in lentic habitats related to their spatial extent (size, depth), temporal extent (ephemeral, permanent), latitude, altitude, distance from the coast, and vegetation type (emergent or in surrounding habitat). For example, in temperate rain forests in the Coast Range and western Cascade Range, habitats for pond-breeding amphibians include ephemeral and permanent ponds, and stream backwaters and deep unvegetated pools (Fig. 2; photographs of contrasting pond habitats, not included in Review Draft). There are many types of wetlands, including forested, scrub/shrub, emergent wetlands, and sphagnum bogs (Fig. 3; photographs of contrasting wetlands, not included in Review Draft). In valleys, large lakes and several wetland types may be found. In the Cascade Range, lentic habitats range from flooded meadows to cirque lakes (Fig. 4; photographs of lakes, not included in Review Draft). High desert ephemeral ponds occur east of the Cascade Range crest. Anthropogenic disturbance has impacted some lentic habitats in this region. For example, stream channelization has reduced off-channel pools, oxbow ponds, and associated wetlands (e.g., Sedell and Froggatt 1984, Benner and Sedell 1996). The character of lentic habitats influences effective survey methodology (e.g., visual surveys may not be effective in heavily vegetated sites or extremely large sites)Amphibian assemblages in lentic habitats are highly variable in the Pacific Northwest. There are 14 species of pond-breeders in the region, with 3 additional species that are sometimes found in association with high elevation lakes (Table 1). Up to five species (up to 2-3 anurans and 2-3 salamanders) are often found at a single site, due to their range restrictions, associations withspecific habitat features that vary geographically, and biotic interactions. Thus, assemblages can vary across the region. Numerous species-specific considerations may affect surveys for these animals. Breeding seasons (Table 1), and nocturnal and diurnal activity patterns may varywith species or site. Some may be easily observed, while others may require observer stealth or experience. Habitat requirements, the tendency for some species to use certain microhabitats, and biotic interactions can affect distribution patterns within sites as well as among sites (e.g., Olson 1988). Within large sites (e.g., a lake), pond-breeding amphibians can be patchily distributed. Pond-breeding amphibians have complex life histories, thus multiple age or stage classes of each species may occur at a location. Seasonally, effective survey methods can vary with the species and life history stages that are anticipated.Choice of sampling technique is guided by several factors: survey objective, resources available, and fauna and habitat to be sampled. Survey objectivescan be numerous, but generally include general collecting, inventory, monitoring, and research. These categories are a continuum with corresponding gradients in the intensity of sampling methods and subsequent analyses conducted. For example, while general collecting or inventories may yield presence data or data on species richness (total number of species) for a single time period or a single site, monitoring extends sampling geographically and/or temporally, and may provide more extensive population data from specified sites. Research may have a completely different focus and obtain population data secondarily or incidentally. While relative abundance data can be taken during well-planned inventory and monitoring surveys, density estimates of populations may be possible only from more intensive monitoring and research studies. How intensive or extensive the surveys are intended to be (per site or per region) should be one of the first objectives delineated. In the choice of methods, other study objectives must also be considered. For example, if comparisons are to be made with historical datasets, attention must be paid to matching approaches with those used previously. Prior surveys may have been more labor-intensive research studies, because relatively few inventories have been conducted with amphibian assemblages historically. As new survey programs are developed, their intent, geographic extent, temporal extent, and availability of resources will drive the types and intensities of surveys possible, which will affect methodologies chosen. Fellers (this volume) provides additional guidance for the design of amphibian surveys.

DEVELOPMENT OF RECOMMENDED METHODS

To achieve a regional consensus of recommended methodologies for pond-breeders, experts from agencies, universities, private institutions, and industry were convened to discuss currently-used techniques, published or established methods, and ongoing and planned studies within which methodology development could be addressed. The scope of the task was restricted to addressing basic inventory and monitoring methods for use in the field to assess populations and assemblages. Not directly included in the scope of this effort were the development or assessment of statistical designs for monitoring programs, habitat assessment using aerial photography, videography, or remotely sensed data, and techniques used exclusively for research purposes. To cover various survey objectives, faunal assemblages, and habitat types, we identified several methods to serve as guides for regional inventory and monitoring. After initial identification of methods, they have undergone a year of further development and field validation by various field crews regionwide. Although evaluations of some techniques are continuing, results-to-date are presented here and are believed to be sufficient to guide currently-needed surveys.Many earlier methods (see Heyer et al. 1994) can be applied to PNWpond-breeders. However, we found a subset of the methods more applicable to the fauna, habitats, and general regional objectives anticipated. Also, some methods have been further developed for regional use. Of the primary ten methods presented for inventory and monitoring (Heyer et a. 1994), six are most pertinent for PNW pond-breeder surveys: visual estimation surveys, quadrat-, transect-, and patch-sampling, breeding surveys, and larval sampling. These provide the foundation of our proposed standardized surveys (Thoms and Corkran, this volume), and rely heavily on visual observations of animals, and hand and net captures. Due to the large spatial extent of some PNW habitats (large lakes or wetlands), the patchiness in species-distributions within habitats, and cryptic species, we have placed a greater emphasis on funnel trapping than was presented in Heyer et al. (see Adams, Richter, and Leonard, this volume). Specific methods often are needed to survey different life history stages or habitat areas, thus a combination of techniques may be needed (Thoms, Corkran and Olson, Bury and Major, Crisafulli, this volume).

Additional supplemental methods of use in the PNW, but not advocated in Heyer et al. nor by us as primary methods for most sites, include snorkeling and electrofishing (Thoms, Corkran, and Olson, this volume). These techniques are used extensively in the region by fish biologists to assess fish populations (e.g., Dolloff et al. 1993), and incidentally detect amphibians. Aquatic phases of lentic amphibians in streams or in inaccessible portions of lakes have been discovered by these methods. Snorkeling, in particular, can be effective in wetlands with low structural diversity, montane lakes with inaccessible shorelines, or lakes and wetlands with soft-bottomed substrates that make visual surveys conducted by walking difficult. We found methods less suitable to PNW pond conditions and fauna for inventory and monitoring surveys included auditory (acoustic) surveys, drift fences and pitfall traps (Thoms, Corkran and Olsonthis volume). Auditory surveys have limited use in the PNW because 11 of the 17 species in PNW lentic habitats (Table 1) do not have generally-audible male advertisement calls during breeding. However, calls of the remaining species can be used to detect species presence or categories of relative abundance. More generally, calls are useful to locate habitats. The use of drift fences and pit-fall traps have several drawbacks. First, they are time consuming to install and to run. Second, although fences with traps can be effective to sample amphibians (e.g., Bury and Corn 1989), it may not be possible for them to sufficiently cover an extensive area, characteristic of many PNW lentic sites. Third, small mammals also become trapped in pitfalls, many of which suffer mortality after a few hours in a trap. The number of sensitive small mammal species in the western portions of the PNW has required biologists to reduce their use of this technique, restricting its use primarily for research.

SUMMARY OF CHAPTERS

This handbook has been developed for the inventory and monitoring methodsexpected to be needed by area biologists to conduct surveys of lentic habitats for biodiversity assessment or prior to proposed land management activities at a site. A "toolbox" of recommended sampling techniques is presented. Guidance is provided to select suitable methods for use in various situations, contingent on sampling objectives, site-specific lentic habitat conditions, and anticipated fauna (Chapters 3 and 4). Examples are given of how these methods have been integrated into current monitoring designs (Chapters 5 and 6). Further discussion of survey objectives and site selection procedures is provided by Fellers (Chapter 2). Fellers outlines issues common to the general design of surveys for inventory and monitoring of lentic-breeding amphibians, including standards for data collection, data analysis, and voucher specimens. Once a site has been chosen for sampling, Chapters 3 and 4 provide screening procedures to guide the choice of sampling techniques. Corkran,Thoms and Olson (Chapter 3) summarize the types of pre-field planning (e.g., map-work, compiling known information, safety issues, data form design) and post-field work that can accompany any field sampling protocol for amphibians occurring in lentic-habitats. In particular, they outline the role of information gathered from known sources and field reconnaissance to tier sampling technique choice. When a habitat can be completely surveyed, our recommended inventory method is the "Basic Survey", presented by Corkran and Thoms (Chapter 3). This method uses primarily visual searching and dipnetting to sample amphibians. Less ritualized forms of this method have been used by biologists for years for haphazard collecting or inventories in small ponds; here, a standardized approach is described for regionwide use. As habitats or pond fauna cannot be completely surveyed, additional methods and subsampling techniques are recommended (Thoms and Corkran, Chapter 3). Adams, Richter, and Leonard (Chapter 4) present aquatic funnel trapping designs and sampling schemes. Funnel trapping can be conducted in most lentic habitats, and is

particularly suitable at sites with limited visibility or access, and at

sites with nocturnally-active and cryptic amphibians. At large lakes or

wetlands, the combined use of funnel trapping with other techniques may

effectively sample amphibians. Bury and Major (Chapter 5) present their study design for the inventory and monitoring of lentic habitats in the Olympic National Park, Washington, and the PNW Cascade Range, integrating many elements discussed in Chapters 2, 3, and 4. In their study, access and safety are issues of concern for site-selection in remote areas. Surveys of sites selected from maps and aerial photographs are augmented by surveys of areas immediately adjacent to these sites, which apparently yields an important subset of lentic habitat types that may have gone unrecognized. Their sampling procedures include the Basic Survey, funnel trapping, and subsampling techniques. Also, Bury and Major summarize observer bias and detectability issues. Lastly, Crisafulli (Chapter 6) provides a more intensive habitat-based methodology for combined habitat and faunal monitoring in high elevation cirque lakes of the Mount St. Helens blast zone. Crisafulli presents a more detailed approach to lentic habitat characterization, which may be easily used for broader study objectives than examining associations with the amphibian fauna. For amphibian sampling, Crisafulli uses primarily visual searching and dipnetting along the periphery of a lentic habitat, with an approach similar to the Basic Survey, augmented by visual searches from a float-tube or by snorkeling in deeper water, and funnel trapping.

U.S. Department of the Interior
U.S. Geological Survey
Patuxent Wildlife Research Center
Laurel, MD, USA 20708-4038
http://www.pwrc.usgs.gov/naamp3/naamp3.html
Contact: Sam Droege, email: Sam_Droege@usgs.gov
Last Modified: June 2002