Stephen R. Humphrey

Summer Habitat and Ecology of the Endangered Indiana Bat, Myotis sodalis

A nursery population of M. sodalis was discovered in Indiana and studied for 2 years. The nursery roost was located under the loose bark of a dead tree. Sometimes the bats temporarily moved to bark crevices of a living shagbark hickory tree. Weather strongly affected roost microclimate. The nursery tree was unshaded and received maximal solar warmth during clear, mild or hot weather, whereas temperature at the shaded alternate tree was more stable during spring and autumn cold. The nursery population consisted of adult females and young, and few males were netted in local foraging areas. Each female bore a single young and these incurred eight percent mortality between birth and weaning. Social behaviors included “checking,” mothers carrying young to an apparently warmer portion of the roost tree, and apparent mother-young foraging flights as the young became volant. Unusually cool weather in summer 1974 slowed the growth of young; this delayed recruitment of flying young more than 2 weeks and the completion of migration by 3 weeks. The delay exposed some bats to freezing weather at the nursery and may have affected mortality, autumn mating, or fat storage for winter. No problems were evident during the favorable summer of 1975. Foraging habitat included the foliage of riparian and floodplain trees, and the 50 bats used a 0.82 kilometer (km) linear strip of creek. Similar habitat was used by two other populations in Ohio. Before young were volant, adults fed only about riparian trees; when young began to fly, feeding extended to solitary trees and forest edge on the floodplain. This reduced foraging density from 17 to 11 bats per hectare (ha). Suitable foraging habitat occurs over much of the eastern United States, and use of local trees for nurseries makes a large summer distribution possible. Humans also value floodplain habitat, and a land use conflict exists between the two species.

Guild Structure of Surface-Gleaning Bats in Panama

Food habits data from a complex guild of surface-gleaning, Neotropical bats over an annual cycle show that prominent consumption of beetles is the feature common to all species. Other insect orders, arachnids, vertebrates, and fruits are minor dietary components of various subsets of the guild. Multivariate analyses confirm the unity of the guild and show that at least five types of food constitute important niche dimensions for guild members. A consistent but not ubiquitous pattern of interspecific differences in size of the dentary apparatus indicates that food item size may be another important niche dimension. These results, added to previously documented differences in vertical and horizontal microhabitat use, invoke at least eight niche axes to differentiate these gleaning omnivores.

Fire in key deer habitat improves brownse, prevents succession, and preserves endemic herbs

Habitat quality reduced by fire suppression may have influenced the recent decline of Key deer (Odocoileus virginianus clavium). A major foraging habitat for Key deer, the rockland pine community, as well as several plants endemic to this habitat, depend on periodic fires for continued existence. Refuge burning is opposed by residents of nearby urban development. To quantify effects of fire, we measured use and nutritive content of browse for the year following fire, monitored growth of the vegetation after fire, and documented succession in relation to time since the last fire. Browsing by deer was most intense early in the growing season. Deer chose species high in crude protein (CP), phosphorus (P), and in vitro digestibility (IVOMD)

Use of population viability analysis to evaluate management options for the endangered lower keys marsh rabbit

We conducted a population viability analysis (PVA) for the federally endangered Lower Keys marsh rabbit (Sylvilagus palustris hefucri). We estimated parameters via livetrapping, radiotelemetry, and fecal-pellet counting during a 2.5-year study. Historically, this subspecies ranged thronghout the Lower Keys of Florida. but today the marsh rabbit is limited to 41 subpopulations occurring in 3 distinct metapopulations. To evaluate this subspecies risk of extinction, we used VORTEX, a simulation modeling program, that combined deterministic and stochastic birth and death events, differing migration rates based on spatially explicit patch structure, and extinction and recolonization. Our simulation calculated, the chance that each metapopulation would go extinet imder current conditions and several management scenarios. Under current conditions, the model predicted there is a 100% chance that all metapopulations will go extinct. The metapopulation with many small patches went extinet faster that the metapopulations with fewer, larger patches The most effective strategy to decrease the risk of extinction of all 3 metapopulations was to increase the survival rate an several age and sex classes by eliminating predation by domestic cats.

Home Range and Movements of the Lower Keys Marsh Rabbit in a Highly Fragmented Habitat

The endangered Lower Keys marsh rabbit ( Sylvilagus palustris hefneri ) occurs in a highly fragmented habitat in the Lower Keys of Florida. A primary goal in recovering this subspecies is understanding how individuals interact in this patchy landscape. Home range and movements of marsh rabbits were studied to determine if rabbits are confined within a habitat patch (relictual population), spend most of their lives in a patch but are capable of moving between patches (metapopulation), or regularly move between habitat patches (patchy population). Radiotelemetry data were obtained from 43 rabbits representing all age and sex classes. Seven collared juvenile rabbits remained in their natal patches of habitat until the onset of sexual maturity. All of the collared subadults (five of the surviving juveniles and 12 rabbits collared as subadults) made a relatively long, one-way movement. Ten of the males moved to new patches; all but one of the females remained in their natal patches. After establishing a home range, each of the adult rabbits (12 collared as juveniles or subadults and 11 collared as adults) remained in one patch of habitat until their deaths. These results indicate that S. p. hefneri exists as a metapopulation. Conservation efforts should be aimed at protecting both the rabbits marsh habitat and the lower-mangrove and upland-forest-corridor habitats used during dispersal.

The importance of patch attributes and context to the management and recovery of an endangered lagomorph

We investigated the role of patch attributes and context on patch occupancy of the Lower Keys marsh rabbit (Sylvilagus palustris hefneri). The Lower Keys marsh rabbit is a federally endangered lagomorph endemic to the Lower Keys of Florida. The marsh rabbit occurs in subpopulations on patches of high marsh that interact to form a metapopulation. Between March 1991 and July 1993, all known patches of high marsh in the Lower Keys were surveyed for presence or absence of marsh rabbit pellets three times per year. Of the 59 habitat patches, 20 had pellets present during all of the surveys (occupied patches), 22 had pellets present during at least one survey (variable patches), and 17 never had any pellets present (empty). Ten variables were measured at each of the 59 patches; seven of these variables concerned attributes of the patch (food, cover, patch size), and three were patch context variables (distance of patch to other patches, distance of patch to other features). Two discriminant function analysis (DFA) were performed. The first DFA compared empty patches to occupied patches (both variably and consistently occupied). Patch isolation explained the most variation in patch occupancy followed by area. The second DFA compared the variably occupied sites with the consistently occupied sites, and patch attributes variables involving the type and height of vegetation were significant. Management efforts for the Lower Keys marsh rabbit should be aimed at both improving habitat quality and decreasing distance between patches.

Seasonal Habitat Use by River Otters and Everglades Mink in Florida

Seasonal habitat use by river otters (Lutra canadensis) and Everglades mink (Mustela vison evergladensis) was documented in the Big Cypress Swamp, Florida, using line transects of chalk-dusted trackboards and anal scent attractant. Response to scent was species-specific in late wet season, indicating that mating occurs in autumn, when water levels are high in marshes and swamps. Mustelids retreated from marshland as it dried, and most had moved to permanent ponds in the late dry season. This period of concentrated aquatic food coincides with lactation in mink. Use of line transects with olfactory attractants may enable monitoring of otter and mink population levels and trends on a regional or statewide scale, but rigorous testing of the technique is needed. J. WILDL. MANAGE. 46(2):375-381 The river otter is included in Appendix II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (U.S. Fish and Wildlife Service 1977). Treaty obligations require determination that otter harvest for export is not detrimental to the species before export permits are issued (Endangered Species Sci. Auth. 1978). The Everglades mink is listed as threatened by the Florida Game and Fresh Water Fish Commission (1980). Protected status has increased interest in these mustelids, but little is known of their biology in Florida. Information from reports by otter trappers was presented by McDaniel (1963). The scattered information available on Everglades mink has been reported by Layne (1974) and Brown (1978), and new data were presented by Smith (1980). Both species occur in the Big Cypress Swamp in southwestern Florida. The objective of our study was to use a measure of relative abundance, adapted from the scent station technique reported by Linhart and Knowlton (1975) and Linhart et al. (1977), to document seasonal habitat use by otters and mink within a large wetland in South Florida. Our research was funded by an NIH Biomedical Grant through the University of Florida Division of Sponsored Research. Work in the Fakahatchee Strand State Preserve was encouraged by J. A. Stevenson and K. C. Alvarez and supported in the field by R. Goble and R. Baker, all of the Florida Department of Natural Resources. We thank D. B. Barbour, J. Jones, B. Tatje, and J. Thompson for assistance in the field.

Activity and diet of an urban population of big cypress fox squirrels

The Big Cypress fox squirrel (Sciurus niger avicennia) is listed as threatened by the state of Florida, individuals are secretive, and their biology is poorly known. However, populations in urbanizing areas persist on golf courses, where they are protected, habituated to humans, and readily observed. These populations offer opportunities to study and promote the survival of the subspecies. Consequently, we studied activity and diet of Big Cypress fox squirrels for 1 year on 4 golf courses in Naples, Florida

Comparison of 2 Methods to Estimate Density of an Endangered Lagomorph

We estimated densities of the endangered Lower Keys marsh rabbit (Sylvilagus palustris hefneri) at 5 habitat patches on Boca Chica Key, Monroe County, Florida, using pellet counting and direct enumeration from livetrapping. We compared density estimates from these 2 methods to an independent estimation of population size based on radiotracking data to determine which method would be best for long-term population monitoring. Density estimates were made 3 times per year for 2.5 years from March 1991 to July 1993. Significant correlations were found between pellet counting and radiotelemetry estimates (r 2 = 0.89, P < 0.0001) and between the livetrapping and radiotelemetry estimates (r 2 = 0.99, P < 0.0001). Pellet-counting estimates took 80% less time to perform than livetrapping estimates. Although livetrapping may have a higher correlation with actual population density, pellet counting provides an economic and efficient estimate that is well suited to long-term population monitoring.

Survival Rates of the Endangered Indiana Bat, Myotis sodalis

Rates of survival of Myotis sodalis are calculated from recaptures of bats marked in cohorts of unknown age. Such data can be misleading, so the proper types ot interpretation are identified. Survival rates are high for 10 years after marking in females and 6 years in males. Females can live as long as 14.8 years and males as long as 13.5 years. Survival of the species is good in undisturbed environments, and endangerment results from loss of habitat.