Growth Rates of Juvenile Nine-banded Armadillos

Growth rates were determined for two captive litters (one male and one female) of nine-banded armadillos (Dasypus novemcinctus) born to the same parents in two successive years. For females, gain in mass from birth to about 2 years old was best described by a fourth order polynomial function. However, increase in mass was linear from birth up to at least 400 days of age. The male litter was born the year after the female litter and. also showed a linear increase in mass from birth through their first year. Males were significantly heavier than females of the same age, even though the female litter was born earlier and weaned later than the male litter. Comparisons between masses of wild-caught and captive juveniles weighed during the same time period (June to August) revealed few differences, but comparisons of growth rates suggested that wildcaught animals may have been born later and grown faster than their captive counterparts. As with data from captives, long-term (greater than 1 year) changes in mass of wild-caught animals were best described by a fourth order polynomial. Data on growth rates may have some value in predicting ages of juveniles from their body mass. RESUMEN-Se determinaron tasas de crecimiento para dos camadas (una de machos y una de hembras) de armadillos de nueve band as (Dasypus novemcinctus) nacidos en cautiverio de los mismos padres en dos alios sucesivos. Para la camada femenina, el aumento en la masa desde el nacimiento hasta dos an os de edad fue descrito mejor por una funci6n de polinomio del cuarto orden. Sin embargo, el aumento en la masa fue lineal desde el nacimiento hasta al menos 400 dias de edad. La camada masculina naci6 el ano despues de la camada femenina y tambien mostr6 un aumento lineal en la masa desde el nacimiento hasta el primer ano. Los machos pesaron significativamente mas que las hembras de la misma edad, aunque la camada femenina naci6 antes y destet6 despues de la camada masculina. La comparaci6n de la masa de juveniles silvestres con juveniles cautivos pesados durante el mismo periodo (junio a agosto) no revel6 mucha diferen cia, pero se not6 una diferencia en la tasa de crecimiento. Esto sugiere que los animales silvestres nacieron mas tarde y crecieron mas rapido que sus contrapartes cautivos. Semejante a los datos de los animales cautivos, los cambios en la masa a largo plazo (mas que un ano) fueron descritos mejor por un polinomio del cuarto orden. Los datos de la tasa de crecimiento quizas tienen valor en predecir las edades de juveniles basados en su masa corporal. " Patterns of growth are of interest to evolutionary biologists because they can provide insights into many aspects of a species' life history such as duration of time spent in the juvenile stage, cost of reproduction, and sex-ratio allocation (Trivers, 1972; Trivers and Willard, 1973; Glutton-Brock and lason, 1986; CluttonBrock, 1991). To address these issues, quantitative descriptions of growth rates are needed for individual species. Herein, we describe growth rates of nine-banded armadillos (Dasypus novemcinctus) by examining gains in mass over time. Nine-banded armadillos are relatively asocial, nocturnal, burrowing mammals (New... man, 1913; Kalmbach, 1943; Taber, 1945; Talmage and Buchanan, 1954; Galbreath, 1982; McBee and Baker, 1982). Breeding usually occurs during summer, with delayed implantation of the blastocyst occurring in autumn (Hamlett, 1932; Enders, 1966; Storrs et aI., 1988; McDonough, 1992). Young are born in litters of genetically identical quadruplets the following spring (Newman and Patterson, 1910; Patterson, 1913; Hamlett, 1932; Enders, 1966; Storrs and Williams, 1968; Prodohl et aI., 1996) and first emerge from their natal burrows in Mayor early June. Previous work on development in nine-banded armadillos has focused on embryological December 1998 McDonough et al.-Growth rates of armadillos 463 eve~ts associated with delayed implantation of the blastocyst and subsequent polyembrony (Newman and Patterson, 1910; Patterson, 1913; Hamlett, 1932; Storrs et aI., 1988). However, Jacobs (1979) estimated post-partum growth rates to average 10.6 g/dayforjuveniles caught in the wild. Her data are limited because they cover only the period from emergence from natal burrows to about 3 to 4 months later, and because juveniles were recaptured infrequently, forcing Jacobs (1979) to assume a linear growth rate between widely separated points in time. We recorded gains in mass for two litters (one male and one female) of nine-banded armadillos born in captivity to the same parents. Our study was designed to extend Jacobs' (1979) analysis by obtaining numerous measurements of individual juveniles from birth through the first year of life. These data allow us to evaluate Jacobs' claim of a linear growth rate of 10.6 g/day for juveniles. In addition, to evaluate how realistic our captive data were, we compared patterns of mass gain in our captive litters with those observed among juveniles living in the wild. MATERIALSAND METHODs-We obtained data from three litters of nine-banded armadillos born in captivity. The first litter was born 10 March 1995 and was female. A male litter was born 2 April 1996 and another male litter was born 7 March 1997. This last litter died at birth, so birth weights were the only data available. The same adult female (7 years old in 1997) produced all three litters. This female had been in captivity from approximately 4 months of age and had reproduced unsuccessfully in 1993 (the litter was stillborn) and 1994 (ultrasound confirmed pregnancy, but the litter was apparently reabsorbed or aborted). The female was housed with two adult males that also were captured as juveniles. These males were 6 and 7 years old in 1997. Although the female was housed with both males, only one (the 7 year old) exhibited reproductive behavior (Jacobs, 1979; McDonough, 1992, 1997) and so is presumed to have fathered all the juveniles. . Growth rates were assessed by recording gains in mass over time. Mass is a convenient index of body size and is significantly correlated with other measurements (Loughry and McDonough, 1996). The female litter was first weighed 6 days after birth and thereafter every other day up to 61 days of age. Mass was recorded weekly from day 65 to 540, then every 2 to 3 weeks up to day 724 (n = 109 measurements per juvenile). The male litter was weighed at birth and daily up to day 94. Males were then weighed every 10 to 12 days up to day 345 (n = 114 measurements per male). Juveniles were weighed to the nearest 3.5 g with a Lita-Minder scale until they weighed 2:454 g (day 50 for males, day 59 for females). They then were measured to the nearest 50 g with either a 12 kg capacity Hanson Utility scale, a Pelouze digital scale, or a Sunbeam Nursery scale. After birth, juveniles were housed with the mother and allowed to nurse at will. Juveniles were provided access to the same solid food as adults (about 1 I of Science Diet dry feline chow, mixed with a hardboiled egg and small amounts of chopped fruits and vegetables) ad lib. In addition, juveniles were allowed to forage in the ground during daily releases into an outdoor enclosure. Thus, food was not limiting and growth rates may be near maximum. Measurements of mass were plotted against age (days since birth) and growth rates were determined by identifying the curve of best fit through these points with regression models, using the Statview statistical package (Abacus Concepts, 1992). We examined differences in masses of males and females with paired (-tests by comparing females and males weighed within 3 days of the same date (i.e., days since birth; n = 57 days total). We used data on masses of juveniles caught in the wild at the Tall Timbers Research Station, located just north of Tallahassee, Florida (for a description of the field site and methods of capture, see McDonough and Loughry, 1997 a; Loughry and McDonough, 1998) in three comparisons. First, we compared masses of juveniles caught in the wild (n = 120) with those of our captive animals. All wild juveniles were captured during the summer (June to August), and we used just the data from captive animals that covered the same time period. Because masses changed over time, data for individuals weighed more than once (n = 14 wild-caughtjuveniles weighed 35 times, range = 2-4 reweighings per individual) were treated as independent. We compared masses within each month with (-tests. As we show below, this comparison resulted in few differences in mass between captive and wild juveniles. However, this comparison is problematic because we did not know the exact ages of wild juveniles (in days since birth). Thus, our result could have occurred because wild juveniles were born earlier and grew more slowly or were born later and grew more rapidly than our captive animals. Such an argument assumes that growth rates differed between the two groups. Our second comparison tested this assumption by examining growth rates of captive and wild juveniles using regression models in which gains in mass were plotted against calendar date (1 June as day 0 through 31 August as day 92). Differences in growth rates between captive and wild-caught juveniles were analyzed with an ANCOVA in which mass was the dependent variable and calendar date was 464 The Southwestern Naturalist vol. 43, no. 4