Winter mortality of adult nine-banded armadillos (dasypus novemcinctus) on Cumberland Island, Georgia

Georgia Journal of Science, 2002 by Bond, Bobby T, Warren, Robert J, Nelson, Martin I

ABSTRACT

Implantable radio-transmitters were used to collect data to estimate seasonal survival rates (SR) of 16 adult (8 males and 8 females) nine-banded armadillos (Dasypus novemcinctus) on Cumberland Island National Seashore, Georgia, from July 1987 through May 1988 (10 months). Seasonal survival rate (SR) was the same for spring, summer, and fall (SR = 100.0%), with all mortalities occurring in winter (SR = 54.6%, SE = 15.0%). We suggest that poor thermoregulatory ability of armadillos may predispose them to overwinter mortality. There was no evidence that these individuals varied in body condition and weight at time of implantation nor died from predation, vehicle collisions, or parasitism; therefore, we postulate that the armadillos in our study died from cold stress during winter. Armadillos are known to have poor thermoregulatory ability, high thermal conductance, low basal metabolic rate, and low body temperature. Hence, they are extremely sensitive to cold ambient temperatures.

Keywords: Cumberland Island, Dasypus novemcinctus, nine-banded armadillo, radio-telemetry, winter mortality.

INTRODUCTION

Few studies have examined armadillo mortality patterns (1) and road-kill trends (2, 3). Zimmerman (4), Herbst and Redford (5), and Bond et al. (6) have presented radio-telemetry data on habitat-use and home range for the nine-banded armadillo. However, to our knowledge, no study on survival of nine-banded armadillos using radio-transmitters has been published. Therefore, our objective was to determine seasonal survival rates of male and female nine-banded armadillos on Cumberland Island, Georgia.

MATERIALS AND METHODS

Cumberland Island National Seashore was established in 1972 and the island is the southernmost and largest (9,416 ha and approximately 28.2 km long by 4.8 km wide) of Georgia's barrier islands. It is located in Camden County approximately 2 km from the mainland. The natural history and ecology of the island have been described by Hillestad et al. (7). Armadillos arrived on Cumberland Island by unknown means in 1971 (7). The absence of hunters and large predators, lack of motor vehicle traffic, and favorable habitat and climate on Cumberland Island have contributed to an abundance of armadillos, which made this area an ideal setting for our study.

We captured armadillos with a long-handled dip net and retained only individuals that were >3 kg on the assumption that they were adults (8). We marked the carapace of each captured animal using enamel paint and by attaching a numbered stainless steel tag in the ventrolateral pectoral shield. Only healthy captured armadillos were transported to a laboratory for implantation surgery using aseptic procedures. Following sedation via an intramuscular injection of a mixture of xylazine hydrochloride (0.5 mg/kg) and ketamine hydrochloride (3.5 mg/kg), a 2.5-cm incision was made in the ventrolateral abdominal wall, a 25-g radiotransmitter (Model IMP 200; Telonics Inc., Mesa, AZ) was inserted intraperitoneally and left free-floating (9, 10, 11), and the incision was closed with three layers of absorbable sutures. Implantable transmitters were used because the body shape and behavior of armadillos, like other burrowing animals, precludes the use of radio-transmitter collars (9, 12, 13). Also, Nelson and Warren (10) evaluated the effects of surgery and implanted transmitters in armadillos, and reported no detrimental effects (e.g., incision completely healed, no surgical complications, and maintenance of constant body weight) 68 days post-implantation. Herbst (11) concluded that nine-banded armadillos in adequate body condition could be subjected to capture and surgery without adversely affecting reproduction or survival. Armadillos were released at their capture sites approximately 18 hrs post-surgery after they fully recovered from anesthesia.

We implanted radio-transmitters in 12 armadillos in 1987 (11 in July and 1 in September) and 4 in 1988 (1 in January and 3 in March; Table I). Therefore, 5 additional armadillos were implanted after the start of our study because we had additional working transmitters to implant. Armadillos were monitored during July and August 1987 (summer - 15 days), September and November 1987 (fall - 17 days), January 1988 (winter - 8 days), and March and May 1988 (spring - 16 days). Therefore, we monitored radio-transmittered armadillos for 10 months. During each monitoring period we tracked armadillos through the 24-hr diet period, locating each individual approximately 6 times daily.

We used PROC LIFETEST (14) a non-parametric Kaplan-Meier product limit estimator method modified by Pollock et al. (15) to estimate sex-specific and season-specific survival rates along with associated standard errors. We used the midpoint between detection of mortality and the last known date that the armadillo was alive when we did not know the exact date of mortality. If transmitter failure occurred, we used data up to the day of transmitter failure.

RESULTS

All armadillos implanted with radio-transmitters survived >3 weeks post-surgery and mortality was assumed independent of capture, surgery, and handling. We report sex-specific and season-specific survival rate (SR) for 16 armadillos (8 males and 8 females). Six armadillos survived the entire study, 5 died >3 weeks post-implantation (26 - 299 days, x(overscored) = 178 days), 3 were marked in the spring of 1988 and survived to the end of the study, and 2 (Fall 1987) transmitter failures occurred (Table I).