Data Handling

Field Data

       Field data should be recorded immediately when taken (fig. 45). Data may be recorded on data sheets, preferably in pencil using waterproof paper, or by using preprogrammed palm pilots. ARMI is currently developing a web based data entry program using palm pilots. Park researchers may desire to link their data collection with the DOI-sponsored national amphibian monitoring program. Palm pilot programs with project-specific formats also can be developed. In any case, the following data should be recorded at all sampling sites (note that all measurements should be recorded in metric units):

Date: month/day/year.
Site No.: a unique identifying site number. Example: BB-1 could indicate site 1 on the Bunches Bald Quadrangle. There are many ways this can be done, but site location codes should be consistent.
Personnel: initials or names of those persons conducting the survey.
Weather: at the time of the survey.
Altitude: in meters.
Wind: categorical judgement of wind speed 1 m above sampling area.
General location: a geographic description of the site location. Example: Garretts Gap on the Hemphill Bald Trail on Cataloochee Divide.
Specific location: using GPS or Topo software.
Quadrangle: USGS 7.5' quadrangle map. Start time and End time: in military time (that is, 0800 or 1600 hrs).
Standing water: at aquatic sites, record whether water is present.
Water level: deepest water level at sampling site. Can be estimated (example: > 0.5 m).
Air temperature (AT): recorded at 1 m above substrate in ^oC.
Water temperature (WT): recorded at 30 cm under water in ^oC.
Substrate temperature (ST): recorded at 30 cm under leaf litter in ^oC.
Relative humidity: recorded at 1 m above substrate in ^oC.
pH: when appropriate, recorded in soil/water with a calibrated meter.
Conductivity: when appropriate, recorded in water with a calibrated meter.
Habitat type: a general appraisal of the habitat type (circle one, see appendix II).
Vegetation: a general appraisal of the vegetation types (circle as many as appropriate, see appendix II).
Canopy: a categorical assessment of canopy cover (especially important at wetland sites).
Slope aspect: a compass direction of slope aspect.
Drainage direction: in which compass direction does a stream flow at the sampling location.
Amphibians: The species (using the three letter species code), sex (if discernible), life stage (adult, juvenile), number of individuals, and other notes (for example, reproductive condition, missing limbs) should be recorded. In some cases, the snout-vent length (for salamanders), total length (for frogs), mass, or other individual measurements may be required by a study’s objectives. Measurements should always be in metric units.
Method of capture: specialized capture techniques may require a data form to reflect the types of data taken, in addition to the information listed above. For example, the identifying number of the trap, PVC pipe, or coverboard should always be recorded to discern possible capture biases. The distance an animal is captured or observed from a transect’s origin and baseline helps indicate spatial distribution.
Invertebrates: the type (genus, order, class) and relative abundance of invertebrates may be very important in studies of amphibians, especially amphibians breeding in ponds and woodland pools.
Active sampling effort: the number of observers (exclusive of the person recording data, unless that person is also sampling animals) x the amount of time sampling occurs.

       If data sheets are used, additional information concerning the site can be included on the back of the form, such as drawings of ponds or pools, sketches and notes of unusual color patterns or morphology, notes on the physical description of the sampling site, records of photographs taken, and the presence of unusual plants and animals. A sample data sheet is included as appendix II.

Spreadsheets and Databases

       Most U.S. Government agencies are now using Microsoft Excel^® and Access^® to generate spreadsheets and databases. Data from field data sheets should be transferred into one of these programs as soon as possible following a survey, or entered directly while in the field using palm pilots, using the same conventions as on the data sheets. Both programs are compatible with a variety of statistical programs, such as SAS^® (Statistical Applications Systems). Data accuracy should be checked to ensure quality control and prevent inaccuracy; the field data sheets serve as a backup from which to double check data records. Backup copies of data should be made weekly, at a minimum, and copies should be safely stored at different physical locations or in a fireproof data safe.

Analysis and Software

       The objective of monitoring the amphibians of Great Smoky Mountains National Park is to detect population trends so that actions can be taken, if possible, to reverse declines should they be detected. Inasmuch as many species’ populations fluctuate from one year to the next, especially in unstable habitats such as temporary ponds, and that populations probably go extinct naturally (and vacant habitats are recolonized), trend analysis is not an easy task to apply to amphibian populations. Much ongoing research is focused on amphibian populations; new biometric methods are being developed to analyze trends in light of the complexities of amphibian biology.

       Traditionally, population trends have been measured via changes in numbers or abundance of the animal in question. If the population size can be measured through time, then changes could indicate increasing or decreasing trends and, therefore, reflect changes in conservation status. To determine the size of a population, it is necessary to relate the numbers recorded during periodic counts to the overall population size. The most commonly used method to do this is to individually mark animals and to record the numbers recaptured during a period of extended sampling. Thus, each animal is accorded a capture history. If enough animals are captured and recaptured during a survey, it is possible to relate the counts mathematically to an estimate of actual population size within a certain degree of confidence. Although it is beyond the scope of this manual to discuss the nuances, theory, and assumptions of mark-recapture analysis, there is substantial literature available on this subject (Pollock and others, 1990; Nichols, 1992; Thompson and others, 1998).

       Unfortunately, it is not easy to use markrecapture techniques when studying populations of amphibians for two reasons:

1. Amphibians are not easy to mark “permanently.” Various methods, such as toe clipping, elastomer implants, and photographic identification (ID), have been used, although each technique has limitations. Amphibians lose toes naturally and regrow clipped toes; elastomers are time consuming to apply and are difficult to read under field conditions, and photographic ID is not practical when hundreds or thousands of animals are involved or when animals are uniformly patterned or unpatterned. Observer error is an ever-present bias.
2. In most instances, very few recaptures are recorded in relation to the number of amphibians marked. In such cases, the variance of the population estimate can become quite large, thus negating the reliability of the estimate.

       In the Great Smokies, there is only one species, the Hellbender, that is probably amenable to reliable mark-recapture population estimation. These large salamanders are territorial and relatively confined to a circumscribed habitat (only large rivers and streams) in a few areas of the Park. They can be permanently identified through implantation of an injectable passive integrated transponder (PIT) tag. As such, resurveys should be possible to track populations within certain sections of streams. Nickerson and others (2002) have marked Hellbenders in Little River using PIT tags, and National Park Service biologists should be able to track the status and size of this population annually using a transect-based snorkeling protocol.

       Another technique that is gaining favor is to conduct repeated sampling at locations throughout a designated area, such as a Park or refuge, or in a particular subset of a habitat type within such an area. Through time, researchers can record a capture history for each species at each location. Thus, a data set is developed that in practice looks very much like the capture history of individuals in a typical mark-recapture study. By recording changes in these species’ capture histories through time, biometricians can determine detection probabilities for each species. Trends can be determined by changes in the “percent of area occupied” (PAO) by a species and by changes in detection probabilities. More information on applying PAO analyses to monitoring amphibians is contained in MacKenzie and others (2002), and at:

http://www.mbr-pwrc.usgs.gov/software.html#presence

SOFTWARE

       Program MONITOR  Power analysis basically tells the researcher how reliable his or her data are considering a number of variables, such as sample size and the length of time that a program is conducted. Important caveats for interpreting the results of a monitoring program are contained in “Power Analysis of Wildlife Monitoring Programs: Exploring the Trade-Offs Between Survey Design Variables and Sample Size Requirements” by Paige C. Eagle, James P. Gibbs, and Sam Droege (http://www.pwrc.usgs.gov/resshow/droege3rs/salpower.htm). The USGS has developed a free software program, MONITOR, which uses linear regression to estimate the statistical power of population monitoring programs relative to: the number of plots monitored, the magnitude of counts per plot, count variation, plot weighting schemes, the duration of monitoring, the interval of monitoring, the strength and nature of ongoing population trends, and the significance level associated with trend determination. MONITOR is available at:

http://www.mbr-pwrc.usgs.gov/software.html
(then click on POWER)

       Program MARK Program MARK provides population parameter estimates (for example, survivorship and population rate changes) based on mark-recapture data. Re-encounters (captures or observations) can be recorded from animals found dead, live recaptures (for example, the animal is retrapped or resighted), radio tracking of an animal’s movements, or from some combination of these sources. The time intervals between re-encounters do not have to be equal, but are assumed to be one time unit if not specified (for example, every week or month). Data can be subsetted, such as by sex or life history stage, so that population parameters can be estimated for the designated group. The basic input to program MARK is the encounter history for each animal (for example, the entry 1001101001 could result for an animal caught 5 times during 10 sampling periods where 1 = captured, 0 = not captured). MARK also can be used to provide estimates of population size for closed populations. Capture and recapture probabilities for closed models can be modeled by attribute groups and as a function of time, but not as a function of individual-specific covariates. Program MARK is available free from Colorado State University at:

http://www.cnr.colostate.edu/~gwhite/mark/mark.htm

       Program PRESENCE The number and diversity of amphibians in the Great Smokies and elsewhere in the southeast makes monitoring all species difficult, if not impossible. Nonetheless, high species richness of amphibians is a hallmark of ecosystems in southeastern North America. Changes in ecosystems through disturbance, human activities, disease, environmental contaminants, or other factors could negatively impact the composition and richness of amphibian communities. Estimating variation in species richness through time and among different locations is one means of tracking the status of amphibians as a group. This type of analysis, termed percent of area occupied (PAO), may be more effective than focusing on abundance measures of individual species, which have been shown in most studies to lack statistical power because, in part, of the low recapture probabilities in markrecapture studies of amphibians.

       In the past, the main hindrance to making reliable inferences about variation in species richness has been the inability to count all species present in an area during a survey. Weather conditions, the behavior of different species, cryptic coloration, and observer skill are just some factors affecting detection (also see Things to Consider During Planning). Invariably, some species will be missed, thus biasing the estimates (Boulinier and others, 1998a,b). However, methods are now available which account for variation in detection probabilities, and which estimate species richness, standard error, and 95 percent confidence intervals (Nichols and Conroy, 1996). These methods have been extended to estimate several important vital rates in animal communities, which would be useful to assessing status, for example, rates of local species extinction, turnover, and colonization (Nichols and others, 1998a). They also have been used to test hypotheses concerning factors affecting temporal (Boulinier and others, 1998a,b) and spatial variation (Nichols and others, 1998b) in species richness.

       The application of PAO methods to amphibian survey data is promising, not only because these methods can address important questions, but also because they may easily be applied to inventory surveys, intensive monitoring at preselected sites, and in extensive surveys (MacKenzie and others, 2002). Furthermore, detection of a change in species richness can alert biologists and managers to potential problems that may require more focused study. To facilitate PAO analyses in amphibian monitoring studies, USGS researchers have developed Program PRESENCE. This program is available free at: http://www.mbr-pwrc.usgs.gov/software.html#presence. This program is still being tested and developed; undoubtedly improvements will be forthcoming to enhance its performance and ease of use.

Equipment and Training