Recipient Organization
STATE UNIV OF NEW YORK
(N/A)
SYRACUSE,NY 13210
Performing Department
Environmental & Forest Biology
Non Technical Summary
Once widely distributed across North America, river otter (Lontra canadensis) were extirpated by themid-20th century from much of their historic range due to habitat loss, pollution, and over-trapping (J).Otter have persisted in northern and eastern New York State, where their populations today areconsidered to be stable. Attempts to restore otter in the rest of the state have included closure of someareas to harvest since the early 1990s and releases of 279 individuals from the Adirondacks into westernNY between 1995-2000 (Figure 1). Translocations appear to have been successful with released animalsestablishing home ranges within two years (2) and surveys and public repmis documenting otterthroughout the region. Neve1iheless, monitoring of otter since their reintroduction has been insufficient toassess the current status of otter populations throughout the desired recovery area, especially whether theyhave achieved harvestable population levels.Pressure is mounting to reopen certain areas to harvest due to the perceived recove1y of otter andthe high market value of otter pelts. Without a baseline assessment of the status ofreintroduced otterpopulations, and a reliable methodology for monitoring changes in their population status, managementplans remain specious. The need for an effective, non-harvest related monitoring plan was identified as apriority for otter in NY's 2005 Comprehensive Wildlife Conservation Strategy (CWCS): "The primary conservation need for river otter and American marten is the development of robustmeasures of population status to inform management actions, primarily adjustment of trappingregulations and repotiing requirements. Moreover, non-harvest-based data are needed to developharvest independent measures of population status." (see Appendix A6, page 15)The CWCS futiher stressed a need to understand the genetic structure of the reintroduced otterpopulation (see Appendix A6, page 14). Some otter populations sustain low genetic diversity (3), raisinga concern that translocated populations may suffer from founder effects, deleterious genetic drift, andproblems with homozygosity ( 4). Thus, understanding the status of reintroduced otter populationsrequires consideration of their genetic status in addition to their distribution and relative abundance.Awarded Start Date: 4/1/15Sponsor Name: NYS Department of Environmental Conservation
Animal Health Component
(N/A)
Research Effort Categories
Basic
100%
Applied
(N/A)
Developmental
(N/A)
Goals / Objectives
The primary objectives of this research are to:I. Provide a statistically-rigorous, baseline assessment of the distribution and abundance of otter incentral and western NY State (Figure I).2. Develop an efficient sampling protocol to reliably detect a 20% change in population status.The first two objectives will be funded by the DEC MOU. As additional funds are secured, we alsointend to:3. Assess otter diets concurrent with statewide sampling to document their feeding habits andpotentially identify factors limiting otter abundance.4. Assess genetic diversity of the reintroduced population in comparison to other extant populations.
Project Methods
Distribution and potential abundance -- Otter are notoriously elusive creatures (5). Despite a thoroughreview of the published literature and personal discussions with biologists studying otter throughoutNorth America, we identified no feasible approach to directly quantify otter abundance over a spatialscale as broad as our assessment area. Spatial-capture-recapture based on genetic sampling shows promise for otter, but the cost may be too prohibitive over our entire project area. As an alternative, otteroccurrence in an area might be readily detected (6, 7) and predictable based on local and regionalenvironmental variables (8, 9). When tied to data in a GIS, a spatially-explicit model of otter distributionmay also reflect their potential abundance when considering the proportion of aquatic habitats predictedto be occupied by otter. Thus, we propose to develop a rigorous empirical model of habitat occupancy byotter to assess their distribution and potential abundance in central and western NY State.Occupancy models are based fundamentally on a comparison of sites used by animals to thosethat remain unused (10). The most reliable indications of otter use of an area are the presence of activelatrine sites and otter tracks (7). However, weather, soil, vegetation, and otter behavior influence theprobability of detecting sign during any given survey (11). Thus, repeated surveys in an area will berequired to develop a detection function for otter so as to produce an unbiased model of otter habitatoccupancy (10). Based on studies of other species (12), we estimate that 4-5 repeat surveys will berequired to provide precise estimates of habitat occupancy for otter. To maximize sample size we willemploy a 'removal' design (10) in which sites will be surveyed a maximum of 5 times if otter are notdetected, but with surveys discontinued at a site once otter are detected. Following the 'removal' of onesite, an additional survey location can be added without increasing the field effort required. Under aremoval design, the number of surveys required until otter are detected is used to estimate the detectionfunction. Removal designs provide the greatest precision under most circumstances (12), and here havethe added benefit of maximizing the number (and spatial extent) of our sampled locations.A large sample of rivers, streams, lakes, ponds, and beaver impoundments that occur within closeproximity to an access road will be surveyed (8, 16). A stratified survey may be employed based on theoutcome of the present niche-mapping effort being completed by A.J. MacDuff (NY DEC biologist andgraduate student at SUNY ESP). To optimize field efforts, sampling will be restricted to late winter andspring when otter most actively mark their territories (13-15) and otter sign is most detectable. Surveyswill consist of walking or canoeing along a 500-m section of shoreline while looking for otter tracks,slides, and latrine sites or otters proper. We will also simultaneously conduct bridge-based surveys,following the standardized DEC protocol, to evaluate how well these low-intensity surveys perform incomparison to our more robust survey design.The detection function and probability of otter occurrence will be estimated simultaneously usingProgram Presence (17). Plausible models to predict otter occupancy will include variables representingphysical habitat features (e.g., shoreline diversity, forest cover, and density of beaver impoundments),surrounding land use, and human disturbances (e.g., summer recreation levels). These data will bederived from existing GIS layers, aerial photo interpretation, and other existing spatial data. Informationtheoretictechniques will be used to select the most parsimonious model ( 18). The best model will belinked to data in a GIS to estimate otter occupancy throughout the recove1y area. We will validate modelpredictions using the DEC otter sighting database, information from previous bridge-based surveys, andother existing data on otter occurrence (e.g., road killed otter, hunter surveys). Ultimately, we willsummarize the total amount and connectedness of habitat having a high likelihood of otter occurrence ineach watershed to assess both the current status of populations and habitat limitations that might threatentheir persistence.Diet -- Mortality, body condition and reproductive success are closely related to seasonal environmentalchanges, especially in food sources (11, 23) thus food availability and quality have direct implications forthe success ofreintroduced otters. River otters rely primarily on aquatic prey (24) including fish,crustaceans and insects, but also are known to consume amphibians, mammals, and small amounts ofvegetation (25). The diet of otters reintroduced into the Black, Honeoye, and Oatka creeks has beencomprised entirely of fish in the winter and spring, with an increase in crayfish during the summer and fall (26). Sample sizes have been insufficient to compare otter diets among areas, thus a larger scaleeffort is needed to fully describe the otter diet in western NY State.All scat encountered during our statewide surveys will be collected and frozen for diet and DNAanalyses. Latrine sites will be revisited during summer and fall to increase sample sizes and captureseasonal dietary differences. After washing and smting scats, hard-part remains will be identified tospecies with the aid of an inverted 50x microscope and dichotomous key reference collection (27). Preysize will be estimated by counting annual growth rings on fish scales (28) and from a regression relatingcrayfish body length to cheliped size (26). Diet will be evaluated using the frequency of occurrencemethod, in which use of given prey item is expressed as a percentage of scats in which that item wasdetected. Volume or weight of remains will not be used as a measure of prey impmtance because ofdifferences in the density and size of the various diagnostic items (29, 30). Principal ComponentsAnalysis will be used to investigate diet in relationship to seasonal, regional, and habitat-based trends.Related to diet, as apex predators, river otters are known to be susceptible to water pollution (31).Although it is infeasible to sample water quality over the entire study area, we will investigate waterquality for a subsample of areas - specifically to compare high occupancy areas to areas where ourmodels may predict high occupancy but where we failed to actually detect otter occurrence. Waterquality (dissolved oxygen, temperature, pH, ammonia, phosphates, and nitrates) will be assessed usingboth an YSI probe and EPA approved test kits following the protocol in the Quality Assurance Work Planfor Biological Stream Monitoring in New York State (32). In addition, a traveling kick sample will beused to collect benthic organisms using a standard D-frame dipnet (with specimens preserved in 75%ethanol for later analysis). All collected macroinvertebrates will be identified to the genus level. Speciescomposition and water chemistry will be combined into an index of water quality for comparison amongareas (32).