D. Survey Protocol for the Van Dyke's Salamander (Plethodon vandykei)

Lawrence L. C. Jones. USDA Forest Service, Pacific Northwest Research Station, 3625 93rd Ave. SW, Olympia, WA 98512.

Introduction

This survey protocol was developed for determining the presence/absence of the Van Dyke's salamander, Plethodon vandykei (PLVA) on federally-owned and managed lands. It provides the conceptual framework and steps to conduct surveys in areas where ground disturbing projects have been proposed.

Van Dyke's salamander is a rare species of terrestrial lungless salamander endemic to the state of Washington (Leonard et al. 1993). It is superficially similar to other woodland salamanders, especially when young. Across its range, it is polychromatic with ontogenetic changes and different color phases as adults. In some respects, it is a generalist, because it may be found in a variety of habitats at a large range of elevations (for a woodland salamander). This species has been found in upland forests, along lake shores, at cave (basalt tube) entrances, at seeps and along streams. However, because it is often found restricted to streamside or seep habitats, it may appear to be a rather specialized species. This remains one of the least known amphibians in the western United States.

The Larch Mountain Salamander, Plethodon larselli (PLLA), is another Survey and Manage species that occurs in roughly the same range in Washington. Although PLLA is typically a more upland species, the two are known to occur together at some sites (Aubry et al. 1987, C. Crisafulli, pers. comm.), so both species should be looked for during surveys.

Techniques employed in this protocol are specific to the different habitat types (e.g., streams, seeps, upland forest) occupied by this species. The methods involve area- and time-constrained sampling (Campbell and Christman 1982, Bury and Raphael 1983, Corn and Bury 1990), transect surveys (Organ 1961, Hairston 1987, Jaeger 1994), nocturnal searches (Applegarth 1994), and opportunistic sampling. A glossary is included in this protocol (Appendix 3) to assist the reader with nomenclature associated with PLVA and its habitats.

Read the entire document before initiating any survey.

1. Prerequisites for Performing Surveys.

The unique features of this salamander and its ecology make it imperative that anyone doing surveys for this species be well-versed on its identification and habits. Prior to conducting surveys for PLVA, field personnel need to be trained in (1) amphibian biology, with emphasis on the Northwest; (2) species identification, including all species of salamanders found west of the Cascade Range crest in Washington; and (3) survey methods, including this protocol. Training should have both classroom and field components and should be recognized as sufficient by experts in the field. The observers should be able to positively identify all plethodontid salamander species that may be sympatric with PLVA, with special attention to juvenile identification. Additionally, the observers should have a complete understanding of the polychromatism associated with PLVA and other congeners, a common source of misidentifications.

References for this species include Leonard et al. (1993) and Blaustein et al. (1995), and the management recommendations for Survey and Manage salamanders in Pacific Northwest forests (Olson et al. 1996, in prep.).

2. State Regulations.

PLVA:

In Washington, a Scientific Collection Permit is required to collect wildlife for research or public display. Activities that involve killing, capturing, controlling or attempting to control wildlife requires this permit. The Van Dyke's salamander is a State Sensitive species. Requests to collect specimens are scrutinized.

Survey Procedures

1. Spatial and Temporal Aspects.

a. Site location.

Surveys will be conducted in the State of Washington, from the Columbia River at the south end to the north end of King County. The eastern edge of the survey region is the Cascade Range crest (as indicated by county lines) and the western edge is defined as the lowland physiographic province of the Puget Trough (Nussbaum et al. 1983). However, because salamander ranges are so poorly known, we would encourage crews to search (at least opportunistically) areas outside these boundaries, particularly to the north and east. Any areas above timberline that do not have special habitats need not be surveyed.

All forested habitats (including stream, seep, and upland habitats) and special habitats (cave entrances, montane lake shores, other suspected sites) that fall within the proposed project area and a 180 m area surrounding the site (Fig. 1) should be surveyed for the presence of PLVA before ground-breaking activities proceed. Although streams and seeps fall under the Standards and Guides of Riparian Reserves (ROD 1994), they should be surveyed if they are within 180 m of a project area, since individuals or populations in the area may be affected by projects.

b. Survey timing.

Seasons - PLVA may have a longer period of surface activity than other lungless salamanders, especially when associated with lotic conditions. This means that it may be possible to survey for this species during the summer or winter, if the environmental requirements are met. However, in general, the best time to survey is during the relatively warm fall and spring rains, as with other lungless species. For upland sites, this is the only time acceptable for survey. Fall is usually the preferred season because: (1) there is an additional cohort available to detect (hatchlings) and (2) salamanders that have reduced activity during dry times are more readily encountered on the surface with the onset of fall rains. However, the threat of snow in the Washington Cascades is ever-present during the fall; in spring, once the snow has melted, the ground usually stays wet until June or so. Thus, sampling opportunities may be up to 3 times longer during the spring. Surveying can be conducted both spring and fall, if presence is not established.

In general, the optimal times are March to June and October to December. The very best months are usually May and October or November.

Number of Site-Visits - Sampling can be stopped once presence is established. To be reasonably assured of adequate sampling to detect presence, sites should be visited a minimum of 3 times, at least once in the spring and once in the fall during acceptable environmental conditions. Each survey should be separated by at least 4 days. Since summer and winter conditions cannot be predicted, these seasons are not part of the survey period outlined by this protocol. Off-season surveys are supplemental and do not replace surveys in the required seasons.

2. Environmental Conditions.

Moisture - The soil under the upper layer of surface objects (rocks, logs, etc) should be moist or wet to the touch in order to conduct a survey. For lotic situations, these conditions should be found 10 m from the edge of the stream or seep. (Note: given the fact that seeps or special habitats may not be able meet the 10-m requirements when they may be otherwise conducive to survey, we suggest sampling them anyway, as a supplemental survey).

Relative Humidity - The air relative humidity at the site should be a minimum of 45%, when measured 1 m above the ground and, in lotic situations, 10 m from the edge of the water.

Temperature - The soil temperature should be between 4 and 15 ^oC at the time of the survey. The air temperature should not have reached freezing during the preceding night.

3. Reference Sites.

It is difficult to completely predict the temperature-moisture regime at any particular site. Therefore, a reference site may be established to help decide whether or not to proceed with a survey. However, reference sites are not entirely accurate. For example, an upland site may be extremely dry, but a streamside site may be acceptable. Similarly, if a reference site is chosen that is in a streamside situation, good conditions may erroneously suggest that the survey site will also be acceptable. Given these difficulties, it is not mandatory to establish reference sites. Good judgement may be more cost-effective in the long run, but this usually requires experience and some trial and error. If reference sites are to be used, here are some suggestions.

a. Two reference sites should be selected: one upland and one streamside.

b. The reference sites should be near one another.

c. They should be at elevations similar to the ones to be surveyed.

d. They should be on the same aspect as the ones be surveyed.

e. They should be otherwise comparable in form as they ones to be surveyed. For example, if a stream is to be surveyed, it should be of similar gradient and have similar cover.

In an ideal situation, if the reference site meets the criteria of acceptable environmental conditions, the conditions of the survey site will be similar. If the area being surveyed has been receiving substantial warm (>3 ^oC) rain and the ground is essentially saturated, you can hardly go wrong.

4. Detection.

Presence is determined when one or more PLVA are found and positively identified. For our purposes, it is assumed a population occurs at that site. Further surveying will help determine the extent of that population by delineating the boundaries. The site will be managed according to the most recent guidelines.

Absence is determined only after three mandatory surveys have been conducted, all resulting in no PLVA being observed.

5. Voucher Specimens.

A voucher specimen should be retained when the animals are detected. Photographic vouchers are encouraged, however photographic specimens are only acceptable if the quality of photographs is good and the animal can be positively identified based on current knowledge of the characteristics of the species. Photographs should be taken of the animal's ventral, dorsal, and lateral surfaces. Photographs should be close-up; e.g., they should nearly fill the frame. Avoid over- or under-exposure and flashback. The colors and patterns should be true to life. Whenever possible a series of individuals representing different age classes, color patterns, and sexes should be photographed. As a minimum, adult animals should be photographed, because they are more readily identified than younger animals. A preserved voucher specimen is an animal that is permanently preserved and archived in a major museum, such as the University of Washington's Burke Museum or National Museum of Natural History. Preservation can be done either by collecting a live specimen and preserving it in fluid or photographing the specimen in life. Pisani (1973) gives guidelines for museum preparation of fluid specimens.

If specimens are to be collected for fluid preservation, a Washington State collecting permit is needed. Because PLVA is considered a sensitive species in Washington, special reference should be made to collection of this species on the application for state permit. The following sentence should be added onto the justification for study: "This study is one of several large scale surveys to detect the presence of Van Dyke's Salamander on federal lands in Washington, in accordance with federal laws and the Northwest Forest Plan".

Survey Method

Survey methods for PLVA are specific to habitat types and site-constraints. Which particular method to use depends on where the ground-disturbing activity is proposed to occur and the features of that area. There are several types of surveys that can be used to look for PLVA, but there is only one recommended for each set of circumstances (Table 3). For example, PLVA may occur in association with streams or seeps, but there is considerable variation in the character of these aquatic habitats. Figure 3 illustrates some basic features of stream morphology in respect to PLVA habitat. Specific methods for each survey type listed in Table 3 are described below.

Table 4 offers suggestions for supplemental surveys. Supplemental surveys may have less rigor than surveys outlined in Table 3, do not fulfill the survey requirements for Van Dyke's salamanders, but may be a quicker way to detect their presence. Supplemental surveys may be used as proposed project areas are being prioritized and general field reconnaissance of habitats is being conducted. Supplemental surveys can establish Presence, but cannot be used to designate Absence.

Table 3. Recommended survey types to assess presence of PLVA prior to ground-breaking activities. The habitat and constraints will determine which type is most appropriate.

Table 5. Outline of supplemental surveys.

Although the survey methods may be considered Area- or Time-constrained searches, it should be realized that the terms area- and time-constrained may be misleading. Constrained implies that a unit of measure is confined, or not exceeded. In the case of time-constrained surveys (e.g., Corn and Bury 1990), however, the measure of effort is standardized across the sampling units rather than maximized, to allow duplication of effort. A standardized area survey is synonymous with a quadrat survey. In this protocol, both time and area will be recorded, so there is accountability of effort. Standardized time surveys are simply referred to as timed surveys in this protocol. Area surveys in this protocol are neither standardized nor constrained, but rely upon the ability of the observer(s) to record area searched.

Transect sampling is a bit more complicated. Jaeger (1994) states that transect surveys are appropriate when working across gradients. The primary gradient in this case is the transition from stream to upland habitat. Surveying this gradient is important because streamside amphibians may shift toward or away from the stream during different seasons or during different climatic conditions (Dumas 1956, Organ 1961, Hairston 1987). PLVA can also be have a spotty distribution along the axis of the stream. Scott and Woodward (1994) point out some other difficulties associated with streamside transects for amphibians. The streamside transect guidelines in this protocol (especially the grouped belt transects) were prepared in accordance to these issues by sampling gradients parallel and perpendicular to the stream.

Regardless of survey method used, there is a chance of not detecting PLVA when it is actually present. Some of this is out of the control of the observer(s). However, it has often been recognized that a good search image may prove more effective than randomized or systematic sampling. To accommodate a good search image and unsampled habitat, a 1 person-hour (p-h) timed survey is tacked on to the end of 3 of the 4 survey types (there is no need for it with the fourth).

For any recommended survey (Table 3), certain guidelines will help the surveyor know when and where to survey:

a. Survey only during acceptable environmental conditions within the survey region.

b. In lotic situations, only the area within 10 m (11 m for grouped belt survey) of the edge of the water needs to be searched.

c. In lotic situations, only survey where water is < 1 cm deep. However, remember that a displaced rock in a seep or overflow area will fill with water.

d. Areas that cannot be logistically searched or searched without severely degrading the habitat, may be skipped. This may mean relocating a search area somewhat and searching before or after the skipped section is reached.

e. Side channels and other areas not surveyed initially are searched at the end with a 1 p-h timed search. Because there is an 8 p-h time limit plus the 1 p-h search, this is known as an 8+1 search.

All of the Survey and Manage surveys will be done in a relatively non-destructive manner. Cover objects will be lifted and replaced to their original position. Bark may be pulled off of logs, but it should be done carefully, so that it can be replaced (this is a favorite haunt of PLVA, so it would be difficult to rationalize not searching this microhabitat). Logs that are moderately decayed into large chunks or splits may be separated, but again, the pieces should be replaced as best possible. However, logs should not be completely destroyed. The intent here is not to keep from altering the habitat at all; that would be next to impossible and would make it difficult to detect salamanders. The intent is to be conscientious about minimizing disturbance to the habitat.

To forward our ecological knowledge of this species, additional information may be gathered at capture locations (Appendix 4). Collection of these data is optional because the primary objective of surveys is to detect presence, however collection of several site characteristics and microhabitat features will aid our understanding of this species.

There are two options for when to stop surveys: 1) stop when a specimen has been detected, or 2) stop at the end of the allotted time or area. From a biological standpoint, the latter is always the preferred method because it yields more information that may be useful later. However, it is realized that if a detection is all that is needed to establish a population, then the survey can be completed as soon as one verifiable specimen is obtained. NOTE: since both PLVA and PLLA are target species within the survey region of PLVA, the survey cannot end until both species are detected, or until the normal stopping point is reached. The lead biologist and crew leaders need to aware of which approach they will be taking ahead of time. This should be coordinated with other biologists on a regional scale.

1. Prior to Sampling.

A contour map covering the specific proposed survey area is needed when conducting a survey. Features such as pockets of late seral forest, talus, streams, and seeps should be delineated within the survey area boundary, as these areas may be the most likely to support PLVA populations.

Prior to initiating a survey, air and soil/substrate microclimate conditions need to be measured. If ambient conditions are outside the range described under the "environmental conditions" section then the survey should be halted for the day.

2. Total Area Searches.

This is the default survey method. It is used whenever all cover objects that fit within the guidelines can be searched within 9 p-h s (8 + 1 search). If the area cannot be searched within this time frame, another method should be selected.

a. Fill out header information on data form as per code sheet (Appendix 4).

b. Mentally delineate survey area as per guidelines and make a judgement call on how long the search will take, particularly with respect to the 8+1 p-h time frame.

c. Search all cover objects that you are ethically allowed to search (see section on Survey Ethics), replacing them as you proceed. Look under rocks, bark, logs, and similar debris, including bark on logs and other stacked objects (e.g., all bark in a bark pile or in slabs of wood). Search down to the top surface of the soil.

d. Whenever an amphibian is found, data are recorded as per code sheet. This time is subtracted from the total time. When a PLVA is captured, the location is marked on a map, the site is flagged and labeled. The animal is released at the point of capture.

e. After the survey is completed, record information on time and area searched and sketch the area, highlighting any areas with a detection.

3. Belt Transect.

This is done in upland areas, when surveying (primarily) for Larch Mt. Salamanders. The belt transect survey design is discussed in the protocol for that species (see Figure 2). Crews will always be looking for both species in Washington when doing this or any other type of survey.

The procedure for implementing a belt-transect survey is as follows.

a. Fill out header information on data form as per code sheet (Appendix 4).

b. First, the surveyor(s) locates one of the lower corners of the survey area and walks 25m upslope (from the corner) and attaches flagging that has the observer's initials, date and line # (lines will be numbered one through X, where X = the last 25m interval possible, before going beyond the upslope boundary of the survey area). This indicates the beginning point for a given transect survey line. If there is more than one surveyor, then each crew member will position themselves at 25 m intervals upslope from one another and indicate their beginning point with flagging (as described above). The crew member(s) will travel along the contour of the slope searching under all available cover objects [woody debris (bark, branches, logs) and rocks] within a 10m wide swath (5m on each side of the transect line). In talus habitats, where cover objects are usually too abundant to completely search, the width of the swath is reduced to 4m (i.e., 2m on each side of the line). In either case, the surveyor searches for animals while walking in a zigzag fashion along the route of travel. Additional flagging is labeled and hung on vegetation at about 50m intervals along each transect route. Once a surveyor has traveled across the site (along a transect line) reaching the distal side, they position themselves upslope at the next 25m location and survey back across the slope as described above. This is continued until the entire area has been surveyed at the 25m resolution. All cover objects searched should be carefully returned to their original position.

Moisture condition is evaluated and temperature is measured at six positions spaced at 25 m intervals along the transect line. At distances of 25, 50, 75, 100, 125 and 150m from the beginning (0m) of the transect determine the closest cover object (e.g., wood, rock) to the line and measure the temperature by slipping a thermometer beneath it, then turn the object and determine moisture condition.

In most cases a survey could be performed by a single surveyor, but if the area is >10 acres, a 2-4 member crew is strongly recommended. This would ensure that sites in the 20-40 acre range would be finished in a single day, thus reducing the likelihood that changes in salamander detectability would occur due to variations in prevailing weather conditions.

c. Whenever an amphibian is found, data are recorded as per code sheet. This time is subtracted from the total time. When a PLVA is captured, the location is marked on a map, the site is flagged and labeled. The animal is released at the point of capture.

d. After the survey is completed, record information on time and area searched and sketch the area, highlighting any areas with a detection.

4. Grouped Belt-Transects

Van Dyke's salamander is usually associated with banks of source or transport stream reaches [stream morphology nomenclature follows Montgomery and Buffington (1993) but is illustrated in Figure 3 of this protocol and terms are given in the glossary, Appendix 3]. They are normally found under rock or wood debris from colluvial and alluvial slumping (Wilson 1993) on the streambank or valley wall. This survey method is used on creeks that are not so steep that they cannot be adequately surveyed and require > 8 + 1 p-h to search all cover objects.

a. Fill out header information on data form as per code sheet (Appendix 4).

b. Begin the transect: start at the downhill margin of the survey area and proceed upstream.

c. Groups (=segment on data form; denoted by sequential letters) of 8 belt transects (sequentially numbered), each 5-m long and 2-m wide will be surveyed parallel to the creek. Numbers 1 and 2 will be on each side of the creek, starting at the edge. Logs or other objects that lie across creeks between 1 and 2 (also 5 & 6) are counted as being in the transect to the side of the creek centerline. Transect numbers 3 and 4 will be on the left side of the stream, centered 5 and 10 m away from the edge of the stream, respectively. Number 5 is parallel to number 4, but along the stream edge. Number 6 is on the opposite bank and numbers 7 and 8 are 5 and 10 m away from the edge, but on the side opposite from numbers 2-5. Each of the transects (except 1 and 2) begin where the last left off at a line perpendicular to the creek. This is illustrated in Figure 4.

d. Walk along the center of each transect, turning over all cover objects within 1 m to each side of the path. Carefully replace each object. Look under rocks, bark, logs, and similar debris, including bark on logs and other stacked objects (e.g., all bark in a bark pile or in slabs of wood). Search down to the top surface of the soil.

e. After group A has been completed, repeat the process for group B and subsequent groups. Group B will start at the end of group A (on the opposite bank) unless it is apparent that the crew cannot cover the length of the streamside in this manner within 8 p-h. In that case, separate the groups by a consistent distance that will allow the distance to be covered. This intergroup distance can be determined by following the instructions in Appendix 5.

f. Run the transects in this manner for as long as it takes (and in as many groups), until about 8 p-h has elapsed. The survey should end after a group has been completed, even if it means going over the 8 p-h mark somewhat.

g. The crew will then spend 1 p-h searching in places that were not searched during the transect survey. The observer(s) will look in what they consider the most likely habitat or any areas they feel they should not have missed earlier.

h. Whenever an amphibian is found, data are recorded as per code sheet. This time is subtracted from the total time. When a PLVA is captured, the location is marked on a map, the site is flagged and labeled. The animal is released at the point of capture.

I. After the survey is completed, record information on time and area searched and sketch the area, highlighting any areas with a detection.

5. Timed Survey--Streamside

The previous survey will not work well if the stream channel is in a deeply cut valley, because it would mean attempting to run a transect where an observer cannot easily walk, or putting some of the transects above the valley, or in steep valley walls where no debris occur. The upland habitat above a deeply cut valley need not be surveyed unless the observer(s) chooses to use the 1 p-h timed search to explore that area. This method is illustrated in Figure 5.

a. Fill out header information on data form as per code sheet (Appendix 4).

b. Starting at the bottom of the stream in the survey area, walk up the streambank, zig-zagging from cover object to cover object, looking under and replacing the objects. Stay on one side of the stream.

c. After 2 p-h, cross the stream and move upstream to the top of the survey area and search under cover objects while zig-zagging downstream for another 2 p-h. This is segment 2. Repeat 5b.

d. Cross the stream and repeat the process for segment 3, walking downstream. Segment 3 should put you between 1 and 2 on the mid to upper end of the creek. Cross the creek again to do segment 4, walking downstream. The four segments should end up being somewhat equally spaced along the stream, as illustrated in Figure 5.

e. When finished with the 8 p-h timed search, conduct the additional 1 p-h timed search, looking in any suspected or missed areas.

f. Whenever an amphibian is found, data are recorded as per code sheet. This time is subtracted from the total time. When a PLVA is captured, the location is marked on a map, the site is flagged and labeled. The animal is released at the point of capture.

g. After the survey is completed, record information on time and area searched and sketch the area, highlighting any areas with a detection.

6. Timed Survey--Other habitats

Most other habitats will likely be searched using the total area survey. However, if the number of cover objects is great, it is possible that > 8 + 1 p-h would be needed to completely search all cover objects. In particular, seeps and waterfalls in talus have large numbers of cover objects (small rocks). They may fall within this category, particularly because of the extensive thin water habitat .

a. Fill out header information on data form as per code sheet (Appendix 4).

b. Mentally delineate the extent of the searchable habitat, including the splash, seep, or overflow area, and surrounding 10 m of habitat. Special habitats may lack the lotic component.

c. Search part of the area for 2 p-h (segment 1), in the manner discussed for the other survey types.

d. Move to another part of the area and repeat (segment 2). Repeat this until 4 segments have been searched. The 4 partial area searches should cover the total area in a roughly systematic manner.

e. Conduct the 1 p-h timed search.

f. Fill out the time and area effort data, and sketch the search area, highlighting any locations of any detections.

g. Whenever an amphibian is found, data are recorded as per code sheet. This time is subtracted from the total time. When a PLVA is captured, the location is marked on a map, the site is flagged and labeled. The animal is released at the point of capture.

h. After the survey is completed, record information on time and area searched and sketch the area, highlighting any areas with a detection.

7. Night Searches (supplemental)

Although PLVA and nearly all other salamanders are nocturnal, searching for them at night with the aid of artificial light has met with varied success. However, since the technique can be productive, it should be tried if the conditions are right and the opportunity presents itself (e.g., en route to other night-time activities). Nussbaum (pers. comm.) has found PLVA to be numerous at some sites at night. This method is most readily used if the site is near a road, not only because of logistical and safety reasons, but also because roadcuts may expose rocky habitat. Searching seeps in fractured rock may be particularly rewarding. Slowly search the site with a flashlight, looking for the salamanders on the surface or in cracks. Search carefully, because only the head may be visible in dark recesses. As with other methods, relatively warm rainy nights in spring or fall are the preferred time. Driving slowly on back roads is productive for some animals such as snakes and ambystomatid salamanders, but is not likely to be so for PLVA due to their small size and supposed site fidelity.

8. Off-season Searches (supplemental)

Although the best time to survey is during the fall and spring, PLVA may be active almost any time of the year if the conditions are conducive. During the summer, they may be active near water where conditions may remain moist. Perennial waterfalls and seeps fall into this category. If a winter is mild, PLVA may be surface active (PLVA in coastal areas are active almost year-round due to maritime influence). It is recommended that these searches be done according to the most appropriate survey method.

9. Opportunistic Searches (supplemental)

Quite simply, an opportunistic search is done anywhere, anytime, by any method. It usually translates into just going out and looking . However, as previously mentioned, it is always preferred to at least document the time and/or area effort (or preferably both).

Delineation of Occupied Sites

Once the Van Dyke's salamander is found at a site, conservative site-delineation is recommended for this rare species. Until surveys are conducted throughout a site, all suitable habitat that is contiguous with an occupied site will be considered occupied and should be delineated. If there is doubt as to the occupancy status of suitable habitat near a known site, survey the site using this same survey protocol. This may include contiguous or adjacent habitat areas.

A finer-scale known site delineation of apparently disjunct populations may be accomplished by additional, more intensive surveys conducted at flagged locations of individual captures. Such surveys can be conducted along multiple transects outward from known site locations, or in a circular path spiraling-out from the flagged site. This more intensive survey effort should include obvious patches of suitable microhabitats. The method used, area searched (transect or path width and length, total area covered), and time spent searching (excluding time spent handling animals and recording data) during these additional surveys should be recorded.

The potential ecological value of contiguous but apparently unoccupied habitat should be considered, especially in regard to desired future conditions, population dynamics, and connectivity issues. Conservative measures are recommended when dealing with this type of rare endemic vertebrate species.