Estimating and modeling survival rates of a white-tailed deer population after 7 years of fertility control: combining Lefkovitch's and Siler's methods.
Kathy L. Schwager
The Trade-off
Reproduction can be very energetically costly in mammals. A trade-off between reproduction and survival has been documented in 
wild ungulates and theory predicts that contraception may shunt energy reserves into increased longevity. This may have serious implications in assessing the effects of immunocontraception on a population. Current models do not take this trade-off into account and consequently show decreased fertility having a much greater effect on deer numbers than is truly the case. This finding would exert important biological, political, and economic constraints on the application of immunocontraception for managing problem deer herds.
Description of Study Area:
Fire Island is a barrier island located along the southern coast of Long Island, NY
(FIGURE 1). It is separated from Long Island by the Great South Bay and the Moriches Bay. It is bordered on the east by the Moriches Inlet and by Fire Island Inlet to the west. 17 towns and villages can be found throughout the western half of Fire Island which can be accessed by boat or by car via the Robert Moses Causeway. The study area (K-L) extends from the western edge of Kismet to the western edge of Lonelyville and includes th
e communities of Kismet, Saltaire, Navy Bay, Fair Harbor, and Dunewood (FIGURE 2).
The physiognomy of Fire Island is typical of most Atlantic barrier islands which grade from a primary dune along the ocean to salt marsh along the bay. The development of vegetation is affected by wind, salt spray, erosion, and other environmental factors. The dominant forms of vegetation are pitch pine (Pinus rigida), beach grass (Ammophilia breviligulata), wax myrtle (Myrica cerifera), bayberry (M. pensylvanica), shadbush (Amelanchier canadensis), and common greenbrier (Smilax rotundifolia). T
his composition of vegetation is typical of Fire Island except within the various communities where residents have planted non-indigenous vegetation (FIGURE 3).
Methods
Mature females of this free-ranging population have been treated with Porcine Zona Pellucida (PZP) for 7 years. Since July of 1995, population density and herd composition have been routinely monitored in K-L and other study areas. Distance sampling methods were used to monitor the population. With this method, perpendicular distances are measured from a designated 
transect to the deer. These distances are used to generate a function which describes how the probability of detection changes with distance. From this function, the area around the transect is then computed. This method assumes that not all deer will be detected, thereby eliminating the need to correct for missed animals(FIGURE 4). As herd composition was calculated a standard error for each proportion was computed based on a formula which incorporates the "group" as the sampling unit, not the individual. The population was reconstructed 1000 times for each survey.
Estimating survival of a free-ranging population will be no easy task. Most methods require following a number of live subjects through time usually through the capture and marking of individuals. Before the existence of these methods, however, rates of transition from one life stage to another were measured. Because deer on my study area have not been uniquely identified, I will estimate survival using Lefkovitch's method, which estimates transition probabilities among fawn, yearling and adult age groups based on herd composition counts (FIGURE 5).
Survival rates will be converted to lx schedules using Siler's life-history models, where we will map one or more px value from our sample into the early hazards, prime-aged, and senescence components of the general mammalian survival pattern (FIGURE 6 ). This will be accomplished using non-linear regression models.
Projections of population change will be simulated using an individual-based population model (IBPM). An IBPM will be used because it will not treat population processes as large sample averages. By virtue of their low numbers, small populations have dynamics that are best portrayed as lottery analogs where chance plays a much greater role in population persistence. We will use GAPPS II, version 1.3, which stands for Generalized Animal Population Projection System, to construct a simulation model of deer life history. GAPPS has an event-driven syntax that can model just about any life history. We will use it to explore the consequences of fertility control on population processes.
Crossing time graphs (FIGURE 7) will allow us to place questions about specific management goals into a probabilistic framework. Instead of predicting precisely "when" a population effect will be observed, we will compute the chances that such an effect will be observed within a specific time frame.
Results to Date:
A significant increase in the K-L deer population was reported from July of 1995 to March of 1998. Since then, however, the population has undergone a downturn (FIGURE 8).
Thus far, only the population reconstruction has been performed. This trend is reflected in the reconstructed population of adult females (FIGURE 9 ). Again, there appears to be first an increase in the number of females and then a marked decrease. A similar trend can be seen in the number of adult males in the population, though not as distinct (FIGURE 10 ). The number of fawns has varied substantially as efficacy changed over the course of the study (FIGURE 11). One explanation for the increase in deer numbers may be that females actually increased their longevity by forgoing reproduction, which exerts a number of "survival" costs. This is our hypothesis and the previously outlined steps will help us to test this.
Printable Version (Adobe Acrobat Reader) If you do not have pictures in the printable document then you need to down load a more recent version of Acrobat Reader.
Contacts
Please feel free to contact us with any questions or comment. Any feedback is welcomed and appreciated.
Kathy L. Schwager
e-mail: k_schwag@yahoo.com
Dr. H. Brian Underwood
e-mail: hbunderw@mailbox.syr.edu
Address:
133 Illick Hall
State University of New York College of Environmental Science & Forestry Syracuse,
NY 13210
About The Author: Kathy L. Schwager
Kathy L. Schwager. B.S. Environmental Resources at Hofstra University, currently M.S. student at SUNY-ESF. Currently research involves estimation of white-tailed deer survival rates after 7 years of PZP treatments and beginning research on using
silvicultural scale to mitigate deer impact on forest regeneration.
