Midwest Fish & Wildlife Conference 2016

AUTHORS: Joseph Altobelli*, Patrick B. Laarman, Jennifer Moore – Grand Valley State Univeristy

ABSTRACT: Turtles and tortoises are experiencing global declines in population numbers due to their sensitivity to environmental conditions as well as anthropogenic interactions. Climate change not only influences patterns of speciation and extinction within taxonomic groups, but also affects the phenology, demography, and habitat selection of many species including turtles. We examined factors that determine nest selection by female eastern box turtles Terrapene carolina carolina and modeled environmental factors that affect nest fate of eastern box turtles in northern Michigan. Using radio telemetry, we tracked hatchlings during their first year of life to construct known fate models to estimate annual survivorship. Preliminary results show low nest success and low neonate survival. We predict that survival will increase with age and will depend on temperatures during overwintering. The results of this study will aid in forming more efficient management strategies for box turtle conservation and similarly susceptible ectothermic species worldwide.

AUTHORS: Tracy Swem, Michigan State University; Alicia Ihnken, Michigan DNR; Gary Roloff, Michigan State University

ABSTRACT: It is generally accepted that Eastern box turtles are declining range wide. Land management activities typically conducted before the growing season, such as prescribed fire, may be exacerbating the decline of already fragmented populations by causing direct mortality from intense, fast moving fires. Suitable fire weather in the spring also tends to correspond with the timing of box turtle emergence from winter hibernacula. We hypothesized that slower, growing season fires may allow turtles to evade advancing flame fronts during their active season while simultaneously achieving vegetative goals. Our objectives were to determine if adult box turtles exhibited evasive behavior associated with a growing season prescribed fire within a southwestern Michigan recreation area. Radio-tagged animals (6 females, 6 males) were located 12 hours before ignition. Once ignition occurred, researchers located individuals within the path of the advancing fire and remained 10-20 meters away to observe behavior as the flame front approached. Direct behavior was observed for individuals subjected to the flame front (3 females, 1 male); the other 8 turtles were either in unburnable areas or outside burn boundaries. All but one individual subjected to the flame front exhibited evasive, actionable maneuvers when the flame was 10-20 meters away and without smoke signaling, included negotiating multiple flame fronts and completely burying into mineral soil. Within the burn unit, all adult research animals were in cryptic locations 12 hours after extinguishment, and in water or wetlands 24 hours after the fire was extinguished. Understanding direct and indirect behavioral effects of growing season fires on this species has important implications for the sustained use of fire vegetation types of southern Michigan associated with box turtles.

AUTHORS: Melissa Szymanski*, University of Michigan–Flint; Teresa Yoder-Nowak, University of Michigan–Flint

ABSTRACT: Conservation of long lived reptiles (e.g., turtles) has many different management strategies, one of them being headstarting. For this study, turtle eggs were collected, hatched, and hatchlings were raised in captivity at the Detroit Zoological Society for a year and a half before we released them into their native habitat within Shiawassee National Wildlife Refuge in Saginaw, MI, USA. Headstarted Blanding’s turtles were released in varying microhabitats. This research looked at different habitat factors within a protected wetland including water depth, vegetation, substrate depth, water temperature, and interspecies interactions to assess the efficacy of release locations by measuring movement patterns and habitat use of headstarted Blanding’s turtles. Blanding’s turtles were tracked for eighteen months, as a part of an ongoing conservation effort, using radio telemetry to measure survival and to investigate the movements among habitats. Data from the turtles and their habitats were analyzed using regression analysis. By analyzing the habitat variables in this manner, relationships between habitat variables can be determined and used in future releases of headstarted turtles. This research will aid in making management decisions for this species which is a species of special concern in the state of Michigan.

AUTHORS: Ethan J. Kessler*, Illinois Natural History Survey (INHS); Jason P. Ross, INHS; Christopher A. Phillips, INHS; Michael J. Dreslik, INHS

ABSTRACT: Accurate home range estimates inform management and are vital to studies of habitat use and selection. Currently, minimum convex polygons (MCP) and kernel density estimators (KDE) remain the predominant home range estimators for chelonian studies despite invalid assumptions (i.e. equal spatial distribution of points (MCP) and uncorrelated relocations (KDE)). The recent popularity of GPS satellite transmitters and automated telemetry systems has led to the development of new home range estimators to handle the numerous, spatially autocorrelated locations produced. One of these methods, Brownian bridge movement models (BBMM), uses the time between successive locations to incorporate movement behavior into the home range estimate. BBMM are most widely used in avian and mammalian studies using GPS satellite transmitter data, but are likely robust enough for wide use with smaller radio telemetry sample sizes. We investigated the application of BBMM using radio-telemetry data for 25 Eastern box turtles Terrapene carolina with an average of 92.8 radio-locations between May and November 2009. We calculated 50% and 95% BBMM for comparison with MCP and 50% and 95% KDE smoothed with the reference h value (KDE href). Despite the small sample size, the BBMM smoothing parameter reached convergence in all individuals, including one individual with only 32 relocations. The 95% KDE href produced the largest home range size followed by the MCP and 95% BBMM (F(2,72)=3.69, p=0.03). KDE href was also significantly larger than BBMM for estimates of 50% core areas (paired-t=-2.69, p=0.01). Additionally, many turtles utilized wooded corridors to move between forest patches. These corridors were represented in BBMM home range estimates whereas MCP included surrounding, unused habitats, and KDE href was simply unable to represent this behavior. Although the BBMM has a reputation as a data-hungry estimator, it can provide robust, biologically relevant results for chelonian radio-telemetry studies with realistic relocation schedules.