Report given to U.S. Fish and Wildlife Service, 1 January 1999, All Rights Reserved.
Final Report: Population Estimates of Florida Key Deer
Field studies of Florida Key deer (Odocoileus virginianus clavium) were initiated by the Department of Wildlife and Fisheries Sciences, Texas A&M University (TAMU). Funding for the project provided by TAMU, U.S. Fish and Wildlife Service (USFWS), and the Welder Wildlife Foundation. Much of the information known about Key deer movements, habitat use, and herd size comes from the work of a research project conducted nearly 30 years ago under the direction of Dr. W. D. Klimstra at Southern Illinois University. TAMU researchers are replicating that historic study allowing a comparison of historic Key deer population dynamics (e.g., deer movements, habitat use, population numbers) to the current deer herd. This report presents preliminary findings for a part of research activities conducted by TAMU personnel during January-December 1998.
Objectives
- Estimate the population size of Florida Key deer on Big Pine and No Name Keys.
- Compare current deer population estimates to historic estimates.
- Determine the population structure (i.e., sex, age) and dynamics (i.e., mortality, reproduction) of Key deer.
- Develop a computer simulation model to evaluate the precision of our estimate.
Trapping and Radio-tagging
Between January-December 1998, 114 deer were live-trapped and radio transmitters were attached using neck collars (Table 1). Deer were trapped using portable drive nets, drop nets, and/or hand-capture techniques.
|
Table 1. Sex and age-classes of collared Florida Key deer, January-December 1998. |
| |
Adult |
Yearling |
Fawn |
Total |
|
Male |
24 |
15 |
2 |
41 |
|
Female |
45 |
21 |
7 |
73 |
|
Total |
69 |
36 |
9 |
114 |
Radiotelemetry
Radio-tagged deer were monitored 6-7 times per week since February 1998. Deer were monitored for 24-hour periods at random intervals (24-hour period was divided into 6 equal 4-hour segments; one 4-hour segment was randomly selected and during that time all deer were located).
Population Density
Silvy (1975) conducted a weekly 10-mile road census and a bi-monthly 44-mile road census (both 1 hour before sunrise and again at sunset) over a predetermined route on Big Pine Key, 1968-72. A road census also was established for No Name Key in our study. These census routes were designed to estimate deer density on the island using a Lincoln-Peterson index** (White and Garrott 1990). In our study, similar methods (i.e., same census routes, method of estimating population) are being used to compare the current population size to historic numbers. Annual population estimates (mean and 95% confidence interval) were calculated using the program NOREMARK (White and Garrott 1990).
**FYI.....A Lincoln-Peterson index uses the number of marked versus unmarked animals to estimate a population. For example, if 100 deer were marked prior to a census, and approximately 50% of deer seen during the census were marked, the Lincoln-Peterson index would estimate the population to be approximately 200 deer. This estimate assumes the ratio of marked/unmarked deer in the population is equal to the ratio of marked/unmarked deer seen during the census.
Road census data (March-December 1998) estimates the deer population to be between 460-536 deer on Big Pine Key (mean 496, 44-mile census combined) and 99-117 deer on No Name Key (mean 107) (Table 2). Approximately 603 Key deer are estimated for Big Pine and No Name Keys. This estimate does not include other keys in the deer’s range (note: estimates for other keys are to be obtained in 1999). Historically, Silvy (1975) estimated approximately 177 and 34 deer (211 total) on Big Pine and No Name Keys, respectively (Table 3). In comparing historic and current population estimates, the population on Big Pine Key has increased approximately 250% whereas the No Name Key population has increased by 318% (Table 3).
|
Table 2. Results of 44-mile sunrise/sunset route for Big Pine and No Name Key, March-December 1998. |
|
Route Type |
Number of Censuses |
Population Estimate |
95% C.I.* |
|
44-mile (sunrise) |
34 |
504 |
442-579 |
|
44-mile (sunset) |
34 |
398 |
361-442 |
|
44-mile (combined) |
68 |
496 |
460-536 |
|
No Name (sunrise) |
34 |
126 |
108-149 |
|
No Name (sunset) |
34 |
94 |
83-107 |
|
No Name (combined) |
68 |
107 |
99-117 |
*95% C.I. - represents a 95% confidence interval. For example, the estimated confidence interval for No Name Key (combined) is 119-142. If 100 census routes were conducted, 95 of 100 estimates would fall within this interval or range. A confidence interval is an indicator of precision.
|
Table 3. Comparison of historic and current Key deer population estimates. |
|
Area |
Present |
Historic |
Increase
(%) |
|
Big Pine |
|
|
|
|
44-mile (sunrise) |
504 |
213 |
237 |
|
44-mile (sunset) |
398 |
165 |
241 |
|
44-mile (combined) |
496 |
177 |
280 |
|
No Name (combined) |
107 |
34 |
318 |
Model Simulation
We developed a computer model to simulate the population dynamics (i.e., growth, death) of the current Key deer herd (Figure 2). Factors such as mortality and reproductive rates were determined from radio-collared deer and incorporated into the model. The model was used to predict the number of deer deaths for the last 5 years (1993-98), allowing predicted numbers to be compared to actual deer deaths. It is important to note model predictions were independent of road census data, thus, allowing the precision of the estimate and current population structure to be evaluated.
The model consisted of 2 variables representing males and females on Big Pine Key (boxes labeled "males" and "females"). We assumed the current population to be our estimate (146 males, 349 females, 496 total) in our model. The model also included an output (i.e., mortality) and input variable (i.e., reproduction) in the simulated population (Table 4).
Since the beginning of our simulation (1993), the total number of deer deaths has exceeded 60 per year with a record number of deer deaths being observed for each subsequent year (USFWS, unpubl. data). In comparing predicted deer deaths from our computer model to actual deer deaths as reported by USFWS biologists, actual and predicted mortality was found to be similar (Figure 3). These data supports our population estimate for Big Pine Key. Preliminary data suggests the Key deer population has been increasing for the last 5 years (the time period simulated with our model) (Figure 4). It is important to note that no inferences are being made to the status of the population before 1993.
|
Table 4. Key deer model assumptions, 1998. Mortality and reproductive information from radio-collared Key deer. |
|
Population size
- 146 males
- 349 females
- 496 total
|
|
Reproduction
- Fawn:doe ratio = 0.53:1
- Buck:doe fetal ratio = 1.40:1 (35/49)
|
|
Mortality
- Male – 50% (7/14)
- Female – 12% (2/17)
|
|
Simulation period – 1993-98 |
Literature Cited
Silvy, N. J. 1975. Population density, movements, and habitat utilization of Key deer, Odocoileus virginianus clavium. Ph.D. Diss. Southern Illinois Univ., Carbondale. 152pp.
White, G. C., and R. A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press, San Diego, Calif. 383pp.