The smooth softshell Apalone mutica is a North American trionychid turtle typically found in moderate to fast moving rivers and streams. Such habitats contrast with more resource rich environments (e.g., marshes and swamps) inhabited by other turtle species for which parental investment and egg and hatchling components have been studied. We measured components of eggs, embryos, and hatchlings and determined levels of pre-ovulatory parental investment in Apalone mutica from Arkansas. Eggs averaged 7.34 g wet mass and 28.8% non-polar lipids (NPL) by dry mass. Triacylglycerol, an energy storage lipid, was the major lipid class component of eggs, averaging 82% of total egg lipids. Levels of structural (mostly polar) lipid classes were generally similar to those found in eggs of other turtle species, although proportions of phophatidylethanolamine were higher and increased with egg size. Incubation time averaged 55 ± SE 3.4 d. At 18 and 36 d of development, embryo dry mass represented 11% and 34%, respectively, of embryo dry mass at hatching. Thus, approximately two-thirds of the total increase in embryo dry mass occurred during the final one-third (18 d) of development. Hatchlings averaged 6.60 g wet mass and 25.9% NPL by dry mass. The index of parental investment in care (hatchling NPL/egg NPL) averaged 75.4%, the highest such index reported among turtles. High levels of pre-ovulatory, parental investment in trophic care of neonates appear to be a factor important to hatchling survival during early life stages such as dispersal from the nest, movement to aquatic habitats, and the period prior to which hatchlings attain a positive energy balance in relatively low resource environments.

Deadly toxins and resistance to them are an evolutionary enigma. Selection for increased resistance does not occur if predators do not survive encounters with toxic prey. Similarly, deadly toxins are of no advantage to individual prey if it dies delivering the toxins. For individual selection to drive the coevolutionary arms race between resistant predators and lethal prey, the survivorship of individual predators must covary with their resistance. The extreme toxicity of the rough skinned newt Taricha granulosa appears to have coevolved with resistance in its predator, the common garter snake Thamnophis sirtalis, yet the mechanism by which individual selection can operate has been unclear in this and other lethal prey-predator systems. We show that individual snakes assess their own resistance relative to newt toxicity and reject prey too toxic to consume. Rejected newts all survived attacks and attempted ingestion by snakes that sometimes lasted over 50 min. Behavioral moderation of toxin exposure by snakes provides the association between individual resistance and fitness necessary for coevolution of lethal toxins and resistance to occur.

We studied kin discrimination in the larvae of the four-toed salamander Hemidactylium scutatum. The results of a spatial affinity assay supported the hypothesis that larvae can recognize kin and do so whether they have been raised alone or with siblings. Larvae were attracted to siblings and did not avoid non-siblings. Two additional experiments tested the hypothesis that kin discrimination is a function of predator density. The consistent predator avoidance pattern of larval H. scutatum was to lower activity level. In the experiment characterized by a higher predator density, sibling groups of larvae exposed to a predator lowered their amount of swimming more than larvae in mixed sibling groups. These results are consistent with the hypothesis of context-dependent kin discrimination and suggest that individuals can balance costs and benefits of behaviors as a function of ecological and social factors. This study is the first to demonstrate kin recognition in the larval stage of any species in the family Plethodontidae.

The technique of counting growth rings to estimate age of turtles is widespread in the scientific literature. Review articles to date have provided primarily lists of authors who have found the technique useful or not, but these reviews have failed to evaluate properly how well the technique actually works. In an attempt to examine how well the published literature supports a biologically meaningful relationship between age and number of growth rings, we surveyed 145 scientific papers that have used counts of rings on scutes to estimate age of individual turtles.

Of the 145 papers surveyed, the authors of 44 papers, which comprised 49 case studies, presented data testing the use of growth ring counts for a population of turtles. Of these 49 case studies, 6 reported that the use of the technique was reliable for aging their turtle species past sexual maturity, 15 reported its use to be reliable for aging turtles to young adult (i.e., sexual maturity), 8 reported its use to be reliable for aging juvenile turtles, 2 found it to be reliable with no age limit given, and 8 reported its use unreliable as a method for aging their turtles. The remainder of the case studies presented data that were difficult for us to interpret as reliable or unreliable. Although 22 papers addressed the pattern of growth ring deposition, only four case studies had sufficient data to indicate that a consistent number of rings was added each year. In this paper, we illustrate how the widespread use of this technique ultimately has led to its acceptance without the rigor of tests of its validity or accuracy. We conclude that (a) studies attempting to calibrate the relationship between growth rings and age are few, (b) a majority of the papers that we surveyed referenced other papers that did not themselves include a test justifying growth ring counts as an estimate of turtle age, (c) aging turtles from counts of growth rings might be feasible in some types of studies, for some species at some locations, but only after calibrating the relationship between ring counts and age for each circumstance, and (d) there is currently no justification for generalizing the use of growth rings to estimate turtle age for many species.

Using microsatellite analysis, we investigated the reproductive success and genetic structure of a translocated colony of the Bahamian iguana (Cyclura cychlura inornata) 10-yr post introduction. We investigated the reproductive success of the founder males to determine if all were contributing equally to the descendant gene pool. We compared the genetic diversity in the founder population with that in the incipient, translocated population to determine if the number of translocated individuals was sufficient to retain the genetic diversity derived from the source population. In 1999, the estimated population size for the translocated colony was between 75 and 90 individuals. Blood was taken from 35 iguanas (16 males, 18 females, and 1 juvenile) of mixed age, including the four original translocated males and three of the four original translocated females. Of the eight polymorphic microsatellite loci in C. cychlura, only five were polymorphic in C. c. inornata. With two exceptions, the low average allelic diversity and heterozygosity among these loci (2.4 and 0.45, respectively) hindered accurate determination of parentage among the founders. Nevertheless, the data indicated that the potential parental contributions appear equal for the founder males and that, at 10-yr post introduction, the small amount of genetic diversity at the amplified loci in the founding population was retained. The low level of genetic variation appears to have no negative short-term effects on the translocated or other populations of C. cychlura inhabiting the Exuma Island chain, making them excellent candidates for translocation and repatriation programs if physiological and environmental requirements are met.

We conducted a chromosome banding analysis (heterochromatin, nucleolus organizer regions, DAPI fluorescence, distamycin A/mithramycin fluorescence) of six phylogenetically basal dendrobatid frog species (Colostethus chalcopis, C. leopardalis, Mannophryne herminae, M. neblina, M. olmonae, M. trinitatis). With the exception of C. chalcopis (2n = 22), all examined species had a chromosome complement of 2n = 24 chromosomes. The C- and Q-band analyses showed that constitutive heterochromatin is present at the centromeres of all species, with Q⁻-regions occurring at the positions of the nucleolus organizer regions (NORs) of C. leopardalis, M. olmonae, and M. trinitatis. The C-band polymorphisms were detected in M. herminae and M. neblina on chromosome No. 6 and in M. herminae on chromosome No. 7. Silver-staining and distamycin A/mithramycin fluorescence resolved a single pair of NORs in each species. The DAPI fluorescence revealed pericentromeric bands on chromosome No. 1 and 5 in M. trinitatis and on chromosome No. 1 and 4 in C. chalcopis. Chromosome data clearly allow a distinction between M. olmonae and northern Trinidadian M. trinitatis, laying to rest arguments that M. olmonae may not be a good species. In conjunction with chromosome information found in the literature, our data confirm that chromosome complements of phylogenetically basal dendrobatids can be quite variable, with both 2n = 24 and 2n = 22 complements present. Discrepancies in the karyotypes of M. trinitatis from near Caracas, Venezuela, and from the Northern Range of Trinidad indicate the possibility of cryptic species within M. trinitatis. The notion that a reduction in chromosome number by up to three chromosome pairs has taken place among Dendrobatidae from a putative ancestral 2n = 24 karyotype is shown to be congruent with current molecular hypotheses of dendrobatid relationships.