Journal of Herpetology · 1998 · 58 citations · 20 references
Life SciencesEcologyZoologyHerpetologyOrnithologyPopulation EcologyMarine Mammal ReproductionReproductive EcologyReproductive OutputReproductive PerformanceClutch SizeEgg Size DeterminationNest PlacementOviposition EcologySea Turtle EcologySea Turtle Nesting BiologySea Turtle Nesting EcologySea Turtle Reproductive Biology
-Reproduction in spotted turtles, Clemmys guttata was studied in Georgian Bay, Ontario from 1992 to 1995 using mark-recapture, radiotelemetry, and x-ray photography. The nesting season occurred in mid to late June and lasted approximately 2 wk. Oviposition was primarily nocturnal and required up to 8 hr to complete. All 11 observed nests were placed in shallow soil on exposed Precambrian Shield rock outcrops. Over 4 yr, an average of 58% of 109 females captured in June were gravid. No turtle laid more than one clutch per year, most females did not oviposit every year, and some (16%) did not produce eggs in three consecutive years. There was a positive relationship between female body size and nesting frequency. Mean clutch size, determined from x-rays and direct counting of oviposited eggs was 5.3 eggs. As predicted from optimal egg size theory, clutch size was positively correlated with female body size, whereas egg size was independent of female body size. However, correlations adjusted for body size indicated no trade-offs between clutch size and egg mass, length, or width. Compared to data from a more southern population in Pennsylvania, Georgian Bay spotted turtles had a larger mean clutch size and relatively smaller eggs. Most of the variation in reproductive output between the two populations was from differences in clutch size, not egg size. Northern spotted turtles are similar to other northern species of turtles in having more delayed maturation, larger body and clutch size, and eggs that are relatively (to body size) smaller than those in more southern populations. Spotted turtles (Clemmys guttata) range from southern Ontario, Quebec, and Maine southward along the Atlantic Coastal Plain to central Florida (Berry and Gidden, 1974; Ernst et al., 1994; Barnwell et al., 1997). The species has been assigned varying degrees of declining and/or protected status throughout its range in North America (Lovich and Jaworski, 1988; Ernst et al., 1994) and in Canada was listed as vulnerable in 1991 (Oldham, 1991). Other than brief notes (Blake, 1922; Alder, 1961; Chippindale, 1989; Lovich, 1989; Wilson, 1989; Graham, 1995), nesting behavior, clutch size, and egg size of spotted turtles have been reported from only a single I Present Address: Department of Zoology, Miami University, Oxford, Ohio 45056, USA. population in Pennsylvania (Ernst, 1970; Ernst and Zug, 1994). Turtles are an excellent group on which to test hypotheses concerning evolutionary patterns of reproductive effort because both body and reproductive measures, and patterns of partitioning of reproductive output, are readily quantified. Latitudinal variation in clutch and egg size occurs in many vertebrate species, including turtles (Moll, 1979; Fitch, 1980; Iverson, 19 2; Iverson et al., 1993, 1997; Lindeman, 1997), but its significance has been poorly understood (Fleming and Gross, 1990). Latitudinal patterns in egg and clutch size in turtles may reflect local selection for egg size (Tinkle, 1961) and environmental influences on clutch size (Iverson and S ith, 1993). However, these patterns may be confounded by correlations with body size 252 This content downloaded from 157.55.39.55 on Tue, 23 Aug 2016 04:23:02 UTC All use subject to http://about.jstor.org/terms REPRODUCTION IN CLEMMYS GUTTATA (Iverson, 1992; Iverson and Smith, 1993; Iverson et al., 1993, 1997). Optimal egg size (OES) theory assumes that parents have limited resources and energy for a given reproductive bout, and that a minimum amount of energy per egg is required to produce viable offspring (Smith and Fretwell, 1974; Brockelman, 1975; Congdon and Gibbons, 1987, 1990). We tested two predictions of OES theory. First, we tested whether there was independence of egg size and female body size, and a negative correlation between clutch size and egg size. Also, we examined whether the amount of variation in reproductive output among females resulted primarily from variation in the number or size of eggs (Gibbons, 1982; Schwarzkopf and Brooks, 1986; Congdon and Gibbons, 1990). Turtles at high latitudes experience shorter nesting seasons, and tend to have lower clutch frequencies and increased individual clutch size (Moll, 1979; Iverson, 1992; Iverson and Smith, 1993; Iverson et al., 1993, 1997). Clutch sizes may be larger in northern populations because larger female body size allows for more eggs, but if egg size is optimized, then one would expect a decrease in egg size relative to body size with increasing latitude. Alternatively, egg size could change with latitude so as to be isometric with female size if, for example, there is a morphological constraint on egg size rather than a single optimal size (Congdon and Gibbons, 1987; Iverson and Smith, 1993). Wild populations of spotted turtles have only one clutch per year throughout their range (Ernst, 1976; although captive females may have two clutches per year: Wilson, 1989; Ernst and Zug, 1994), so females in Georgian Bay should not vary their clutch size to compensate for a shorter active season as has been suggested for turtles that have multiple clutches per year (Iverson, 1992). Therefore, clutch size in spotted turtles should be similar throughout their range after adjustment for body size variation. Alternatively, in this northern population, clutch size could be constrainted by a short growing season. In this paper, we briefly describe nesting behavior of a northern population of spotted turtles from Georgian Bay, Ontario, particularly with reference to location of nest sites in a rocky, Precambrian Shield habitat, and to rates of recruitment. In addition, we tested for intraspecific differences in body, clutch, and egg size relationships similar to those reported for wood turtles (Clemmys insculpta) (Brooks et al., 1992), common snapping turtles (Chelydra serpentina) (Galbraith and Brooks, 1987; Galbraith et al., 1989; Iverson et al., 1997), painted turtles (Chrysemys picta) (Christiansen and Moll, 1973; Moll, 1973; Iverson and Smith, 1993; Lindeman, 1997), musk turtles (Sternotherus odoratus) (Tinkle, 1961; Edmonds and Brooks, 1996), and other turtle species (Wilbur and Morin, 1988). MATERIALS AND METHODS The study took place on a 300 ha island near the east shore of Georgian Bay, Ontario (45?N, 80?W), near the northern extreme of the spotted turtle's distribution (Ernst et al., 1994). For the most part, the island consists of Precambrian Shield rock outcroppings that are sparsely veget te with lichens and moss, interspersed with low-lying sphagnum (Sphagnum spp.), fern (Osmunda cinnamoea, 0. regalis), and shrub swamps. The most common aquatic shrub species include speckl d alder (Alnus rugosa), mountain holly (Nemopanthus mucronata), leatherleaf (Chamaedaphne calyculata), sweet gale (Myrica gale), and pale laurel (Kalmia polifolia). The dominant tree species are white pine (Pinus strobus), juniper (Juniperus communis), tamarack (Larix laricina), red maple (Acer rubrum), white oak (Quercus alba), and northern red oak (0. rubra). The island has no roads or human dwellings, and has been bordered by a provincial government fish sanctuary for the past 10 yr, and is therefore relatively undisturbed. Turtles were captured most often by hand and sometimes by dip net while walking and c noeing the study areas. Searches began in late April and continued until November, except in 1995 when field work concluded after nesting (1 July). For each turtle captured, we recorded midline carapace and plastron lengths, and carapace and plastron widths (to 0.1 mm) using Summit calipers, and weight to 1 g using a Pesola spring scale. Some gravid females (N = 5 in 1994, N = 10 in 1995) were fitted with radiotransmitters (165 MHz) (Holohil Systems Inc., Woodlawn, Ontario) and radiolocated and observed at least once every evening and morning during the nesting season until they oviposited. The mass of each transmitter was equivalent to approximately 4% of body mass. The turtles were radiolocated using a TRX-1000S or a TRX2000S receiver (Wildlife Materials Inc., Carbondale, Illinois) and a threeor four-element folding, aluminum Yagi antenna. Depending on topog aphy, signals could be received at distances of up to approximately 600 m. Du ing 1992 to 1995, females caught in June were palpated for shelled eggs. Between 6-17 June, we also x-rayed females in 1994 (N = 17) and 1995 (N = 20) to determine the number of oviducal eggs (Gibbons and Greene, 1979). Turtles were x-rayed at 200 mA and 45 kV peak for 1/15 sec at a local veterinary clinic. Females wer returned to their site of capture within 4 h of x-ray. Eggs from 11 natural nests were measured during 1994 to 1995. The eggs of five nests in 253 This content downloaded from 157.55.39.55 on Tue, 23 Aug 2016 04:23:02 UTC All use subject to http://about.jstor.org/terms J. D. LITZGUS AND R. J. BROOKS TABLE 1. Proportion of female spotted turtles determined to be gravid in June from 1992 to 1995 in Georgian Bay, Ontario. Reproductive condition of females was assessed by palpation and observed oviposition in all 4 yr, and by x-rays in 1994 and 1995. No. females No. females Year captured gravid % gravid
20
Justin D. Congdon, Arthur E. Dunham, Richard C. van Loben Sels · Conservation Biology · 1993 · 650 citations
Life Sciences, Biology, Ecology +17