Author: Erin Muths, USGS Fort Collins Science Center, 2150 Centre Avenue, Bldg C, Fort Collins, CO 80526-8118 20708, erin_muths@usgs.gov

Species list
Description of technique
       Field time components
       Office time
       Equipment
Things that could bias your counts
       Detectability
       Effort
       Weather
       Time of year/time of day
       Habitat change
       Observer effects
Advantages and disadvantages
Approaches to analyzing your data
Existing protocols and programs using this technique
Studies that have used this technique
Literature cited
Send a comment on this technique (this takes you to another page)
See existing comments (this takes you to another page)

This list is not exhaustive, but provides a few examples. The success of VES may depend on breeding season, or most recent rain shower.

Most salamanders
Bufo boreas, Boreal (western) toad (Males don't call, so call surveys ineffective)
Pseudacris triseriata, Chorus frog
Rana pipiens, Leopard frog
Rana sylvatica, Wood frog
Spea bombifrons, Spade foot toad (Presence often associated with rain events)
Bufo cognatus, Great Plains toad
Bufo punctatus, Red-spotted toad
Rana catesbeiana, Bull frog

Visual encounter surveys (VES) are a time honored field technique. VES was formalized by Campbell and Christman (1982) and Corn and Bury (1990), both using time as the constraint. Visual Encounter Surveys are used to document the presence of amphibians and are effective in most habitats and for most species that breed in lentic (non-flowing) water. There are a number of assumptions inherent in VES that should be considered when designing a program using this technique. In brief, the assumptions are: 1) equal observability among species and among individuals, 2) no between-sampling visit effects e.g. there is an equal likelihood of being observed for each species for each sampling visit, 3) individuals are recorded only once per survey, and 4) no observer related effects. A more detailed description of the assumptions and limitations of VES can be found in Crump and Scott (1994).

Visual encounter surveys are conducted by observers walking through a designated area for a prescribed time, visually searching (in a systematic way, e.g. transects), for animals. The number of animals encountered are noted along with time elapsed during the survey.

Visual encounter surveys are effective in easily identifiable habitats, such as riparian zones or ponds or in uniform habitats with good visibility. Species that are highly clumped are also good targets for VES, for example, pond breeding amphibians. In this case, the survey is restricted to the particular sites (ponds) of interest. VES is also useful in surveying species that are difficult to catch in traps or are rare. It is an inappropriate technique to use in searching for fossorial or canopy dwelling species. VES can be applied in monitoring and surveying designs. Visual encounter surveys can determine species richness, provide information for compilation of a species list, and provide data used to estimate proportion of area surveyed that is occupied by target species. Data collected yields information on the presence of a species but does not establish absence, nor does it give reliable estimates of abundance.

VES can be used along transects, streams, ponds, in quadrats or larger areas. There are three standard sampling designs for VES, randomized walk, transects, or a quadrat design (see Crump and Scott 1994 for details).

The scale of inference from this technique depends on the scale of the surveys. For instance if the level of inference is an entire refuge, locations for VES should be chosen randomly, but within strata that provide even spatial coverage of the entire area encompassed by the refuge. The level of inference can scale up to watershed or region with appropriate site selection and sampling effort.

VES can be supplemented with dipnetting and aural identification where appropriate. VES used in conjunction with pitfall arrays may be more effective in some habitats. More than one person can participate, number of minutes searching is always the number of minutes searched multiplied by the number of people searching. Ancillary data such as air and water temperatures, weather conditions, date, and time of survey should also be recorded.

Minimum data collected during VES includes, number of each species encountered, size (e.g. length or acreage) of the area searched and total search time.

• Record data: depends on the amount of area designated, generally based on the outline of a pond or wetland or length of stream reach.
• Record weather and ancillary information: 2-10 minutes per area.
• Travel between areas to be surveyed. This can be considerable when surveying in rough terrain with very discreet wetlands or streams are located several kilometers apart.
• Vehicle travel to survey area or to trailhead.
• Return trip: out of wilderness/drainage, travel back to office.

• Mapping and site selection: variable, depends on what data and maps are available. May want to use National Wetland Inventory Maps or other existing data the provide coverages for wet areas. Can designate VES areas by watershed or by some combination of hydrological units, assuring an approximately equal number of water bodies in each sampling unit (= VES area).
• Planning: Depending on the accessibility of sites, advance logistical planning is critical. VES can be particularly useful during the breeding season when amphibians tend to be more obvious. If there are numerous sites with difficult access, careful planning to minimize travel time and nights in the field is critical.
• Data entry: Data entry can be minimized with the use of PDA (personal data assistants) (studies using PDAs in the field include the USGS Amphibian Research and Monitoring Initiative, Rocky Mountain Region (http://www.mesc.usgs.gov/research/rarmi/) and Southern California (http://www.werc.usgs.gov/index.html).
• Editing and crosschecks of entered data: QA/QC time depends on the amount of data, whether or not PDAs were used and how much QA/QC was built into the electronic form.
• Analysis: Depends on the amount of data (number of sites and number of visits) and how complex the analyses. Proportion or area occupied can be analyzed using (for example) program MARK (http://www.cnr.colostate.edu/~gwhite/mark/mark.htm) or program PRESENCE (http://www.proteus.co.nz).

• Protocol hard copy
• Data sheet (use "rite in the rain" paper or a clipboard covered with plastic)/ pencils or indelible ink pen, or PDA / extra batteries / rain cover
• Thermometer
• Compass
• GPS unit (optional)
• Maps
• Watch
• Rain gear
• Spring scale (e.g. pesola) and bag* to hold animal while it is being weighed (if animals are to be weighed)
• Calipers (if animals are to be measured)
• Headlamps if data to be collected at night
• General gear as appropriate to your survey sites
• First aid equipment

* Because of disease considerations, individual bags should be used for each animal. At least use one bag per species per site.

Things that could bias your counts

A number of variables influence detectability, and detectability can, in turn, bias VES. These include, but are not limited to: reproductive cycle, unpredictable seasonal events (e.g. drought or blizzard), animal activity patterns, air, water and substrate temperatures, relative humidity, soil moisture, rainfall, barometric pressure, cloud cover, moon phase, prey (food) availability, and predator activity.

Small numbers of sites visited and / or low number of visits per site can affect estimates of proportion of area occupied. In general, the greater the number of sites and visits, the lower the variance and the smaller the confidence interval around the probability of finding an amphibian on at least one visit and therefore, the more robust the estimate. Few sites surveyed and low numbers of visits will generate higher variance and larger confidence intervals around the probability of finding an amphibian on at least one visit.

Weather can compromise detectability during VES as well as affect amphibian behavior. For example, rain can affect the ability of the observer to see animals both in and out of the water. Weather may also influence amphibians to be more active and therefore more detectable in the case of warm spring rain, or less detectable in the case of sleet or snow. In general, most protocols strive to minimize the effects of weather by determining a priori what the guidelines are for conducting surveys. For example, don’t do a VES if it snows (although I’ve done VES in the snow and found toads at high elevation breeding sites). Over time, weather biases smooth out, but can be confounding in the short term (e.g. 2-5 years).

Breeding season is the optimal time for conducting VES because most amphibians (pond dwelling, stream dwelling) are more visible during this time especially in the case of anurans that are explosive breeders and return to specific natal ponds in the spring. Visits that occur before the breeding season has started or after it has concluded may bias results. This potential source of bias should be considered carefully. Often, breeding or activity seasons can vary by several weeks with elevation even though the sites are in the same general area, (Corn and Muths 2002). Time of day can also bias results and can be linked to temperature factors, sunlight and activity cycles of particular amphibians.

Changes in habitat over the years may provide a subtle bias in the results if those changes increase or decrease the relative detectability of the amphibians. For example, if over time a site's habitat shifted from dense and nearly complete grass cover to brush with an open understory and lots of downed wood, counts would be biased for some species as they would be easier to find and spot in the now more open and accessible environment. In most circumstances, however, vegetation and habitat changes will have a minor influence on detectability, even though they may have major changes in population size and status.

These are biases brought to the project by the observers who participate. Changes in personnel doing the survey from one visit to the next, particularly the visual acuity and “search image” of a particular observer can influence the results. Errors in identification, both by visual inspection and by sound can occur. This is more likely in areas where amphibians are more abundant such as the southeastern United States. In the Rocky Mountains, the amphibian fauna is relatively depauperate, making identification fairly simple (generally needing to differentiate between 6 – 10 species rather than dozens). Observer bias, in whatever form it takes can be factored into some modeling exercises by adding the observer as a covariate.

Advantages:

• Little equipment is needed.
• Can be accomplished with only one person (depending on scale of project).
• Amenable to stratification by habitat, presence of water or other factors.
• Will easily scale up from small areas to regional or continental scale (with enough resources).
• A focus on numbers of populations in a landscape rather than numbers of individuals in a particular population is the more appropriate focus when asking questions about amphibian decline. This technique provides data to track amphibian populations at the landscape level. Finding a decline in the number of populations of amphibians across an entire landscape is more informative than tracking one population whose numbers are decreasing. A single population may be going through a natural boom-bust cycle whereas a decrease in the number of populations across a region suggests a more widespread problem.

Disadvantages:

Naïve estimates (number of sites searched / number of sites searched where amphibians found) provide a similar answer as the more complicated modeling procedures but without a confidence interval. Both the naïve estimate and any modeling procedure (e.g. Proportion of Area Occupied (PAO)) with too few sites and/or visits may be imprecise and inaccurate.

Proportion of area occupied is a method of analysis using data generated by VES methods. PAO by breeding populations of each species of amphibian detected is the primary response variable used for detecting trends in amphibian populations at mid-level monitoring sites at multiple study areas under the Amphibian Research and Monitoring Initiative (ARMI). Proportion of Area Occupied is based on a closed-population model that incorporates detection probabilities along with the number of water bodies where each species is detected to estimate the area occupied by each species. For common species, PAO is based on the proportion of sites where breeding populations are detected. For rare species such as the boreal toad in the southern Rocky Mountains, PAO may be scaled up to the proportion of watersheds (or other large units of land) occupied rather than sites.

This is not an exhaustive list, but provides some examples of programs and projects that are have used or are currently using VES as part of their standard protocols.

Analysis:

Proportion of area occupied (PAO). Program PRESENCE (http://www.proteus.co.nz). PAO can also be computed using program MARK (http://www.cnr.colostate.edu/~gwhite/mark/mark.htm)

Data collection and design:

Crump, M.A. and N.J. Scott Jr., 1994. Visual Encounter Surveys. In: Heyer, W.R., M.A. Donnelly, R.W. McDiarmid, L.C. Hayek and M.S. Foster (Eds). 1994. Measuring and monitoring biological diversity: standard methods for amphibians. Smithsonian Institution Press, Washington. Pp 84-92.

Olson, D. H., W.P. Leopnard, and R.B. Bury. 1997. Standardized survey methodologies for pond breeding amphibians in the Pacific Northwest including methods, design and suppliers.

Heyer, W.R., M.A. Donnelly, R.W. McDiarmid, L.C. Hayek and M.S. Foster (Eds). 1994. Measuring and monitoring biological diversity: standard methods for amphibians. Smithsonian Institution Press, Washington. Pp 364.

Programs:

Amphibian Research and Monitoring Initiative (U.S. Department of the Interior)

NOCA NRPP Amphibian Inventory, North Cascades National Park Service Complex. R. E. Holmes and R. S. Glesne. May 1999 (http://www.nps.gov/noca/amphrep98.htm).

NOCA NRPP Amphibian Inventory Bridge Creek Watershed. R. E. Holmes and R. S. Glesne. February 1998. (http://www.nps.gov/noca/bcamph.htm).

Bury, B. and M. Adams. Inventory and Monitoring of Amphibians in North Cascades and Olympic National Parks, 1995-998. (http://fresc.fsl.orst.edu/online/online_docs/amphib_ONP.pdf).

Jung, R.E., S. Droege, J.R. Sauer, and R. Landy. 2000. Evaluation of terrestrial and streamside salamander monitoring techniques at Shenandoah National Park. Environmental Monitoring and Assessment 63:65-79.

Standardized monitoring methods for amphibians in National Parks and associations between amphibian abundance and environmental stressors (http://www.pwrc.usgs.gov/amphib/primenet).

Wetland Amphibian and Reptile Community Structure Project Manager: J. W. Koebel Jr. Principle Investigator: Dr. M. A. Donnelly
(http://www.sfwmd.gov/org/erd/krr/krrep/component/4_herpframe.html).

Campbell, H.W. and S.P. Christman. 1982. Field techniques for herpetofaunal community analysis. Pp. 193-200. In: N.J. Scott, Jr. (ed.), Herpetological Communities. U.S. Department of the Interior, Fish and Wildlife Service, Wildlife Research Report 13.

Corn, P. S., and R. P. Bury. 1989. Logging in western Oregon: responses of headwater habitats and stream amphibians. Forest Ecology and Management 29:39-57.

Corn, P.S. and E. Muths. 2002. Earlier breeding in dry years can ameliorate exposure of amphibian embryos to UV radiation. Ecology 83(11): 2958-2963.

Crump, M.L. and N.J. Scott, Jr. Visual encounter surveys. In: Heyer, W.R., M.A. Donnelly, R.W. McDiarmid, L.C. Hayek, M.S. Foster (Eds.). 1994. Measuring and monitoring biological diversity: standard methods for amphibians. Smithsonian Institution Press, Washington.

MacKenzie, D. I., J. D. Nichols, G. B. Lachman, S. Droege, J. A. Royle, and C. A. Langtimm. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology 83:2248-2255.