THE risk of predation has pervasive lethal and nonlethal

Copeia 2013, No. 1, 159 164 Hitting the Ground Running: Environmentally Cued Hatching in a Lizard J. Sean Doody 1,2 and Phillip Paull 1 Evidence is accumulating for the ability of animal embryos to hatch early in response to the immediate threat of egg predation. However, early hatching in response to predation is known from only amphibians, fish, and invertebrates. Herein we present the first quantitative evidence for induced early hatching in a reptile. In two laboratory experiments, delicate skink (Lampropholis delicata) embryos responded to a surrogate predator cue vibrations by hatching,3 days earlier than their spontaneously hatching clutchmates. Early hatching embryos were significantly smaller and left more residual yolk in their eggs, however, suggesting a cost to hatching early. Assuming our vibrations were interpreted as an increase in predation risk, skink embryos can thus forego some yolk absorption and growth when threatened by imminent predation. Simulated predation experiments in the field induced hatching in both nest sites (horizontal rock crevices) and in eggs displaced from nest sites. The hatching process was explosive: early hatching embryos hatched in seconds and sprinted from the egg an average of,40 cm as they hatched. Our results are unusual in demonstrating early hatching in a terrestrial animal with a simple life cycle, and likely extending predation-induced early hatching to reptiles. Early hatching may be widespread in oviparous vertebrates. THE risk of predation has pervasive lethal and nonlethal effects on prey behavior (Abrams, 1984; Lima, 1998). Perhaps the most conspicuous and predictable antipredator response is fleeing, because of the survival benefit of urgent evasion (Edmunds, 1974; Lima and Bednekoff, 1999). Thus, when faced with immediate danger animals often forego beneficial behaviors such as foraging, drinking, mating, thermoregulating, and sleeping (e.g., Hertz et al., 1982; Lima and Dill, 1990). Reducing immediate predation risk is commonly achieved by a flight response: moving rapidly away from the predator (Edmunds, 1974; Lima and Dill, 1990). Embryos have traditionally been considered to be particularly vulnerable to predation because they are imprisoned within the egg until hatching. This vulnerability would be especially relevant in animals without parental care. However, recent research has discovered that embryos can evade imminent predation by escaping their eggs and moving into different habitats (Warkentin, 1995, 2011a). Environmentally cued hatching (ECH), whereby embryos exhibit environmentally induced plasticity in their timing of hatching, is a little-known but fast emerging topic (reviewed in Warkentin and Caldwell, 2009; Warkentin, 2011a). Types of ECH can be simplified into early hatching, delayed hatching, and synchronous hatching, and can be in response to biotic and abiotic risks. For example, hatchingcompetent embryos can delay hatching until environmental conditions are more favorable for emergence, and synchronous hatching can facilitate mass emergence to dilute predation risk (Carr and Hirth, 1961; Martin, 1999). Early hatching occurs when embryos reaching a hatchingcompetent stage of development hatch in response to an acute risk (e.g., eggs under attack) or a chronically elevated risk (e.g., embryos exposed to chemical predator cues; reviewed in Warkentin and Caldwell, 2009). Although this risk can be abiotic (e.g., flood mortality) embryos can hatch early in response to predation. In the best-known examples, attacking snakes and wasps induce Red-eyed Treefrog (Agalychnis callidryas) embryos to hatch hours to days earlier than embryos not under attack (Warkentin, 1995, 2000). Hatching embryos freefall from their leaf-attached eggs into the pond below to begin their larval life, where they face different predators than do their eggs. Early hatching in response to predator attack has previously been demonstrated mainly in frogs (reviewed in Warkentin and Caldwell, 2009). How might this scenario play out in a terrestrial animal with a simple life cycle, where an egg predator can double as a predator of hatchlings? Can terrestrial embryos such as those of reptiles hatch early in response to predation (Doody, 2011), and if so how might they escape the predator as a hatchling? We quantified environmentally cued hatching in the delicate skink, Lampropholis delicata, a lizard with a simple life cycle and terrestrial eggs. We tested the hypothesis that delicate skink embryos can hatch early in response to predators by using standardized vibrations in the laboratory as a crude surrogate for vibrations emitted by predators. We also investigated a cost of early hatching by testing the hypothesis that early hatching embryos hatch at a smaller body size than spontaneously hatching embryos (Vonesh, 2000; Kusch and Chivers, 2004). To quantify the explosive nature of hatching in a semi-natural context, we used three simulated predation field experiments to measure the influence of a predator on (1) the propensity of embryos to hatch and flee the nest site; (2) the hatching sprint distance; and (3) the propensity to hatch when the eggs are pushed from the nest site. We discuss our contribution within the context of the emerging literature on environmentally cued hatching and its proximate and adaptive mechanisms in animals. MATERIALS AND METHODS Study species and study site. The delicate skink, Lampropholis delicata, is a small lizard species native to eastern Australia (Cogger, 2000). It is a terrestrial species that mainly inhabits leaf litter (Taylor and Fox, 1991). In the Sydney area it lays 2 6 parchment-shelled eggs from November to February (Joss and Minard, 1985; Forsman and Shine, 1995; Thompson et al., 2001) under or in cover objects such as leaves, logs, rocks, grass, crevices, or man-made objects (e.g., Shea and Sadlier, 2000; Cheetham et al., 2011). Communal nesting in L. delicata is relatively common (Shea and Sadlier, 1 School of Biological Sciences, Monash University, Clayton, Victoria 3800, Australia; E-mail: (PP) phillippaull1@gmail.com. 2 Present address: Department of Ecology and Evolutionary Biology, 569 Dabney Hall, University of Tennessee, Knoxville, Tennessee 37996-1610; E-mail: jseandoody@gmail.com. Send reprint requests to this address. Submitted: 20 September 2011. Accepted: 17 September 2012. Associate Editor: J. F. Schaefer. F 2013 by the American Society of Ichthyologists and Herpetologists DOI: 10.1643/CE-12-111 Copeia cope-13-01-21.3d 20/12/12 21:42:32 159 Cust # CE-12-111

160 Copeia 2013, No. 1 2000; Cheetham et al., 2011). Mothers abandon the eggs after laying, and the eggs incubate for 4 8 weeks until hatching, depending on temperature (Downes and Shine, 1999). Hatching generally occurs in late summer (Wapstra and Wapstra, 1986). The study site was a section of Lane Cove River National Park (LCRNP) in northern Sydney, New South Wales, in southeastern Australia (33u479140S, 151u089340E; elev. 52 m). The climate is humid temperate, and the site is mainly closed wet sclerophyl (eucalypt) forest surrounded by urban areas. The park encompasses the narrow steep-sided valley around the Lane Cove River. Numerous sandstone escarpments ranging from ground level to a few meters high occur throughout the forest and park, and delicate skinks lay their eggs in narrow horizontal crevices in those escarpments (Cheetham et al., 2011). Laboratory experiments quantifying early hatching. Two laboratory experiments were conducted: one using field collected eggs and a second using eggs laid by gravid females in the laboratory. For experiment 1, 61 skink eggs from 20 clutches were collected from a single communal nest on 12 December 2010 at LCRNP. The exact lay dates for the eggs were unknown, but were between 28 November (nest site was empty) and 12 December (egg collection date). Although these eggs were in one crevice, sibling eggs were attached to one another, allowing us to know clutch as a factor(a further 26 single eggs were not taken). Eggs were placed in moist vermiculite in a plastic box and transported to the laboratory at Monash University on 16 December. In the laboratory eggs were split into two groups (incubators): a treatment group to be vibrated and a control group (not vibrated). Eggs were carefully separated from their clutchmates by hand and individually placed on top of moist vermiculite (2:1 vermiculite:water ratio) in small plastic dressing containers (Decor 35 ml). Containers were placed into one of the two incubators (R-com 50 digital incubator, P&T Poultry, Cornwall, UK) both at a constant temperature of 27uC and constant humidity of 80%. Both the incubators and containers had transparent lids, allowing us to determine if eggs had hatched. Eggs were left incubating and undisturbed in a constant temperature room at 20uC. On 5 January (age,4 weeks), the estimated earliest possible hatching date based on incubation data from the literature, containers from the treatment incubator were removed and placed on a laboratory orbital shaker (Bioline, Edwards Instrument Company, Narellan, New South Wales, Australia) daily, for one minute, checked for hatching, and returned to the incubator. Hatching was also monitored daily starting 5 January in the control incubator. Care was taken not to disturb eggs in the control incubator, as inadvertent vibrations could cause hatching and compromise the experiment. When an egg hatched in the control incubator, the incubator lid was very carefully and lightly removed, the individual container carefully removed, and the incubator lid carefully replaced. This action was practiced prior to the experiment to ensure rapid monitoring while minimizing disturbance. Hatching age was thus determined for each egg as hatching date minus the estimated lay date for the eggs (4 December). Hatchlings were measured (snout vent length to the nearest 0.5 mm) within 24 hrs of hatching. For experiment 2, we captured 100 gravid females by hand from LCRNP during 25 27 November and transported them to the laboratory. In the laboratory ten mothers each were housed in ten circular polyethylene tubs (diameter: 1100 mm 3 height: 420 mm) in a semi-constant temperature room (25 30uC). A centrally mounted heat lamp (40 W) created a thermal gradient to allow behavioral thermoregulation and tubs were lined with newspaper to absorb excreta. Each tub included ten artificial nest sites, each of which was a section of hollow cement paver (146 mm L 3 36 mm H 3 37 mm W; Riviera Australis), cut laterally to allow easy inspection of the nest with minimal disruption. Each nest site was capped at its outermost end with a moistened sponge which sat in a petri dish of water; this was refilled with 10 ml of water daily to maintain nest humidity above 80%. Lizards were fed crickets and watered ad libitum. Thirty eggs from 15 clutches were subsequently taken from the tubs for the experiment. Eggs were not always collected from the tubs immediately after laying because the setup was designed for an unrelated experiment. In experiment 2, it was not always possible to know clutch because some sibling eggs became separated, and single eggs had to be included to ensure sufficient sample size. Eggs were collected prior to expected date of hatching competence and transported to small dressing containers in two incubators as in experiment 1. In experiment 2, however, eggs were incubated at a constant temperature of 25uC. The vibration protocols were the same as in experiment 1. In this experiment, age at hatching was determined as the hatching date minus the actual lay date. Hatchlings were measured as in experiment 1. Hatchlings from both experiments were released at the collection site. Simulated predation field experiments. These three experiments simulated predation by manipulating eggs in nature. Eggs were selected based on the number of eggs in the crevice, and on our ability to manipulate them without disturbing other eggs. Although we did not know the lay dates of the eggs, the late experimental date (23 February) relative to the general timing of laying suggested that embryos were hatching competent. In experiments 2 and 3, eggshells were examined for residual yolk immediately following hatching. In the first simulated predation (SP) experiment we quantified the flight response of embryos to simulated predation in the nest crevice. Specifically, we wanted to know what proportion of embryos would hatch explosively in response to simulated predation. By hatching explosively we mean within seconds and moving quickly from the egg. Each of 19 eggs in a communal nest within a single crevice 1.4 m above ground was gently probed with the blunt end of a bamboo cooking skewer until the egg hatched. We were careful not to disturb the other eggs in the crevice. The narrow crevice necessitated the use of a skewer. Although we did not know the lay dates of the eggs, the late experimental date (23 February) relative to the general timing of laying suggested that embryos were hatching competent. Hatchlings fleeing from the crevice were monitored for 30 seconds in the leaf litter below. Hatchlings fleeing deeper into the crevice were not pursued. In the second SP experiment, we quantified the explosive nature of hatching by simulating egg predation and measuring the hatching sprint distance, or the distance that hatchlings sprinted from the egg immediately upon hatching (in the same motion). Thirty eggs from a single communal nest were carefully removed from a crevice with Copeia cope-13-01-21.3d 20/12/12 21:42:33 160 Cust # CE-12-111

Doody and Paull Environmentally cued hatching in lizards 161 a skewer one at a time, without disturbing adjacent eggs. Eggs in this nest were of unknown age, but were expected to be hatching-competent due to the late experiment date (23 February) relative to the general timing of laying. Each egg was quickly placed on a large flat rock, which was,1 m from the nest site, and softly rolled back and forth in place with an index finger until embryonic fluids discharged from pipping slits in the eggshell. To mimic the biting action of a predator the egg was then immediately and repeatedly pinched lightly at a rate of once every two seconds with the index finger and thumb until the hatchling burst sprinting from the egg. The distance of the initial sprint from the eggs was then measured with a ruler. In many cases the eggs hatched while in hand before they could be placed on the rock, and these individuals were excluded from the calculations of mean sprint distance. In the third SP experiment, we examined the propensity to hatch in response to (simulated) predator disturbance when the eggs are pushed from the nest site. In this experiment we carefully rolled 42 individual eggs one at a time from a crevice containing a communal nest, being careful not to disturb adjacent eggs. Eggs in this nest were also of unknown age, but were expected to be hatchingcompetent due to the late experiment date (23 February) relative to the general timing of hatching. Eggs were allowed to freefall 2.1 m to the leaf litter on the ground below. Eggs that did not hatch immediately were pinched as above for up to 30 seconds. Eggs were scored as hatching immediately upon hitting the ground (eggshells breached within 10 seconds), hatching upon being pinched for up to 30 seconds, or not hatching. Data analysis. In hatching experiment 1, where clutch was known but lay date unknown, differences in incubation period and SVL between treatments and controls were analyzed using paired t-tests. Clutches in which fewer than one treatment and one control egg did not survive to hatching were removed from the analysis, and means were calculated for clutches with multiple eggs in a treatment group. In hatching experiment 2, lay date was known but eggs were not always identifiable to clutch because some of the sibling eggs came apart (sibling eggs are normally stuck together). Thus, SVL was analyzed as a function of age at hatching and treatment using linear mixed-effects modeling fit by maximum likelihood, with nest of origin as the random variable. The difference in age at hatching between treatments was analyzed using a randomized block ANOVA with clutch as the blocking factor. Type III sum of squares was used to account for imbalance. Hatching success was the number of eggs hatching divided by the total number of eggs, excluding clutches in which eggs hatched (experiment 1). We did not calculate hatching success in the second experiment because some eggs were desiccating at the beginning of the experiment after being laid, as part of an unrelated experiment. All statistical tests were conducted using R statistical and graphical environment (R Development Core Team, 2009), using the statistical criterion of 0.05. Assumptions were checked for all tests prior to analysis. RESULTS Laboratory experiments quantifying early hatching. Hatching success was 80% in the vibrated eggs and 71% in the control eggs in hatching experiment 1. Embryos stimulated with Fig. 1. Early hatching in the delicate skink, Lampropholis delicata, in response to a vibration cue in the laboratory. Age at hatching in experiment 1 was estimated based on a median date between the earliest and latest possible lay dates (see Materials and Methods). Bars are means 6 1 SD. ***P, 0.0001. vibrations in the laboratory hatched significantly earlier than did control embryos in both experiments (Fig. 1; Table 1). The mean difference in hatching time between treatment and control eggs was 3.4 days (n 5 15 clutches) for experiment 1, and 3.4 days (n 5 15) in experiment 2 (Fig. 1). Age at hatching (incubation period) was significantly greater in hatching experiment 2 than in hatching experiment 1 (Table 1; F 1,66 5 626.32, P, 0.001), and was presumably related to incubation temperature. In both hatching experiments, hatchling body size was significantly smaller in early hatching lizards, compared to spontaneously hatching lizards (Table 1). In experiment 2, there was no significant effect of age at hatching (t 28 5 1.44, P 5 0.161) or an age at hatching 3 treatment interaction (t 28 5 0.687, P 5 0.498) on SVL, so they were removed from the model. Hatchling body size did not differ between experiments (F 1,58 5 0.00, P 5 0.979). A small but unmeasured amount (up to 5 mm long and 2 mm wide) of residual yolk remained in the eggs in early hatching lizards. In spontaneously hatching lizards, residual yolk was absent or was only the size of a pin-head (approximately 1 mm in diameter). Both early hatching and spontaneously hatching skinks were alert, fast-moving, and responded to touch by sprinting short distances. Simulated predation field experiments. In the first simulated predation experiment, 16 of 19 (84%) embryos hatched rapidly (within ten seconds of probing) and immediately launched themselves from the nest crevice to the ground 1.4 m below, in response to their eggs being probed gently with the blunt end of the skewer. The remaining three embryos also hatched rapidly but moved deeper within the nesting crevice. Fourteen of the 16 hatchlings ran under leaves upon hitting the ground, while two remaining hatchlings hesitated, moving under leaves after our close approach. In the second simulated predation experiment, embryos removed from the nest and stimulated to hatch early sprinted 41.6 6 22.76 SD cm directly from the egg in one motion (range 5 5 89 cm; n 5 30). However, some embryos had difficulty running due to being caught in sticky Copeia cope-13-01-21.3d 20/12/12 21:42:34 161 Cust # CE-12-111

162 Copeia 2013, No. 1 Table 1. Early Hatching and Its Trade-off with Hatchling Body Size in the Delicate Skink, from Two Laboratory Experiments. Data are means ±1 SD. Lay dates used to calculate age at hatching in experiment 1 are medians (see Materials and Methods). Significance is from paired t-tests or singlefactor ANOVA. Attribute Treatment (vibrations) Control (no vibrations) Significance Experiment 1 Age at hatching (days) 37.9 6 2.43 41.3 6 2.82 T 14 5 7.35, P, 0.0001 Snout vent length (mm) 15.3 6 0.89 16.0 6 1.07 T 11 5 2.50, P 5 0.029 Experiment 2 Age at hatching (days) 43.0 6 1.33 46.4 6 1.51 F 1,32 5 32.61, P, 0.0001 Snout vent length (mm) 15.3 6 0.13 16.1 6 0.13 T 30 5 4.42, P 5 0.0001 embryonic fluids and membranes. Lizards generally ran in the direction that they emerged toward, and those sprinting further into the leaves hid under them. In several cases eggs hatched in hand before they could be tested, despite the sprint surface being,1 m from the nest site. A small but unmeasured amount of residual yolk remained in all eggs after hatching. In the third simulated predation experiment, 36 of 42 eggs (86%) allowed to freefall from a nest crevice began the hatching process as they hit the ground (in leaf litter). In 12 of these, hatching and emergence from the egg occurred in one motion as the eggs landed. All six of the eggs that did not hatch upon hitting the ground did so within 30 seconds after being gently pinched between the fingers. All hatchlings disappeared in the leaf litter as they hatched. As in the above experiment, a small but unmeasured amount of residual yolk remained in all eggs. DISCUSSION Our experiments were the first to experimentally quantify early hatching in a reptile in response to mechanical stimulation (Doody, 2011), and our results likely reflect antipredator behavior. Although there were limitations to our findings, our field and laboratory data offer compelling evidence for early hatching in response to mechanical stimulation. Delicate skink embryos responded to our surrogate predator cue vibrations by hatching,3 days earlier than their spontaneously hatching clutchmates (Fig. 1). Moreover, we demonstrated that hatching in delicate skinks is explosive: hatchlings sprint directly from the egg, and this flight response likely reflects antipredator behavior. Lastly, we identified a potential cost to hatching early: early hatching skinks were significantly smaller and left behind larger residual yolks in their eggs than spontaneously hatching skinks. Lizard embryos can thus forego the final days of yolk absorption and growth when faced with imminent predation. The shorter incubation period in experiment 1 reflected warmer incubation temperature (27uC) compared to experiment 2 (25uC) in this species with temperature-dependent developmental rate (Downes and Shine, 1999). The degree of early hatching was very similar between the two experiments (3.4 days in both), and our split clutch design ensured that our result was biologically real. The most likely source of error in these types of experiments is accidentally stimulating hatching in the control incubator through inadvertent vibrations (removing and replacing the incubator lid when retrieving hatchlings). Extreme care was used with our control incubators, however. Although early hatching in reponse to fungal attack has been quantified (Moreira and Barata, 2005), early hatching in response to predation risk is currently known in frogs and fish (Warkentin and Caldwell, 2009; Smith and Fortune, 2009; Warkentin, 2011a). However, early hatching in response to predation has been suspected to occur in reptiles. For example, eggs of the lizard Plica plica hatched immediately when disturbed, suggesting that their rapid escape may be an adaptation for eluding predators (Vitt, 1991). Alternatively, movement of one egg while hatching may trigger hatching in other eggs, facilitating synchronous hatching (Vitt, 1991). Although empirical support is rare, synchronous hatching could dilute predation, satiate predators, or allow escape of the nest before hatching smells attract predators to unhatched eggs (Carr and Hirth, 1961; Vitt, 1991). Our data support an antipredator role for early hatching in skinks, but do not allow us to distinguish between hypotheses involving individual hatching vs. synchronous hatching (but see below for discussion of vibrations). Warkentin (2005) showed that embryos of the frog Agalychnis saltator, a species which lays its eggs on leaves overhanging water bodies, will respond to vibration cues from predators by hatching and falling into the water below. Terrestrial skink embryos differ from embryos of Agalychnis in that skink hatchlings face potential predation by the egg predator (e.g., an egg-eating snake would no doubt consume a hatching embryo or neonate). As a result, not only do hatching-competent embryos need to hatch rapidly in response to predator attack, but hatchlings also need to move quickly from the immediate nest site. Our three simulated predator experiments provided support for both of these behaviors in delicate skinks. Simulated predation caused skinks to hatch rapidly and sprint from the next crevice onto the ground below. Moreover, eggs rolling out of a nest crevice hatched immediately upon hitting the ground. Thus, like Agalychnis, skink hatchlings possess a free-falling antipredator escape mechanism. Eggs are commonly displaced from crevices containing communal nests at LCRNP (Cheetham et al., 2011). Although some of these eggs are displaced by conspecific mothers, it seems probable that these oval-shaped eggs could also be inadvertently displaced during attacks by egg predators. The cue for early hatching in our experiments was apparently vibrations. Vibrations are being increasingly revealed as a hatching stimulus in early hatching frog embryos. Only one research group has examined how embryos recognize vibrations as a hatching cue: embryos of the frog Agalychnis callidryas are able to distinguish between vibrations made by predators (snakes) and benign Copeia cope-13-01-21.3d 20/12/12 21:42:36 162 Cust # CE-12-111

Doody and Paull Environmentally cued hatching in lizards 163 sources (rainfall, wind), and can assess different components of vibrations (Warkentin, 2005; Warkentin et al., 2006, 2007; Caldwell et al., 2009). Vibrations can also play a role in other contexts: vibrations made by sibling embryos may cause early or synchronous hatching in pig-nosed turtles, Carettochelys insculpta, within a flooding context (Doody et al., 2012). Our vibrations via a laboratory shaker and handling, although coarse, sufficiently mimicked vibrations made by predators, although it is possible that we instead mimicked sibling vibrations. Playback experiments might clarify whether early hatching in delicate skinks was in response to the direct threat of predation vs. an indirect pre-emptive mechanism to facilitate synchronous hatching. The full suite of predators of delicate skink eggs is not known, but we found shed skins of the Golden Crowned Snake (Cacophis squamolosus) in skink nest crevices. Golden Crowned snakes feed mainly on skinks but also consume reptile eggs (Shine, 1980). Other egg predators include monitor lizards, bandicoots, ants, and centipedes (Carter, 1992; S. Doody, pers. obs.). Early hatching embryos are likely to incur some cost to hatching a few days before their spontaneously hatching clutchmates, otherwise they would not spend a few extra days in the egg, when undisturbed. In Agalychnis, larger more-developed hatchlings are more likely to survive exposure to a variety of predators (Warkentin, 1995). Although we did not measure survival, early hatching delicate skinks were longer than those hatching spontaneously. Early hatching amphibians are generally smaller and/ or less-developed (Warkentin, 2011b). For example, embryos of Reed Frogs (Hyperolius cinnamomeoventris) hatch at a smaller size from clutches infested with predatory fly larvae (Vonesh, 2000), and early hatching embryos of Fathead Minnows (Pimephales promelas) induced by predatory crayfish cues were also smaller (Kusch and Chivers, 2004). The greater amount of residual yolk left behind in early hatching skinks in early hatching skinks represents energy that could have been either been metabolized or internalized for later use. Future research should quantify differences in residual yolk between early and spontaneously hatching embryos, as well as the impact of this difference on hatchling growth rate, body size, and neonate survival. ACKNOWLEDGMENTS We thank C. Jolley and N. Pezaro for assistance in the field, R. Fowler, B. Lees, E. Love, and R. San Martin for assistance in the laboratory, and N. Pezaro and K. Warkentin for useful discussions. The research was funded in part by Monash University (startup funding for J. S. Doody). LITERATURE CITED Abrams, P. A. 1984. Foraging time optimization and interactions in food webs. American Naturalist 124:80 96. Caldwell, M. S., J. G. McDaniel, and K. M. Warkentin. 2009. Frequency information in the vibration-cued escape hatching of red-eyed treefrogs. Journal of Experimental Biology 212:566 575. Carr, A. F., and H. Hirth. 1961. Social facilitation in green turtle siblings. Animal Behaviour 9:68 70. Carter, D. B. 1992. Reproductive ecology of the lace monitor Varanus varius in southeastern Australia. Unpubl. Ph.D. thesis, Australian National University, Canberra. Cheetham, E., J. S. Doody, B. Stewart, and P. Harlow. 2011. Embryonic mortality as a cost of communal nesting in the delicate skink. Journal of Zoology, London 283:234 242. Cogger, H. 2000. Reptiles and Amphibians of Australia. Reed New Holland, Sydney. Doody, J. S. 2011. Environmentally cued hatching in reptiles. Integrative and Comparative Biology 51:49 61. Doody, J. S., B. Stewart, C. Camacho, and K. Christian. 2012. Good vibrations: sibling embryos expedite hatching in a turtle. Animal Behaviour 83:645 651. Downes, S. J., and R. Shine. 1999. Do incubation-induced changes in a lizard s phenotype influence its vulnerability to predators? Oecologia 120:9 18. Edmunds, M. 1974. Defense in Animals: A Survey of Antipredator Defenses. Longman, Harlow. Forsman, A., and R. Shine. 1995. Parallel geographic variation in body shape and reproductive life history within the Australian scincid lizard Lampropholis delicata. Functional Ecology 9:818 828. Hertz, P. E., R. B. Huey, and E. Nevo. 1982. Fight versus flight body temperature influences defensive responses of lizards. Animal Behaviour 30:676 679. Joss, J. M. P., and J. A. Minard. 1985. On the reproductive cycles of Lampropholis guichenoti and L. delicata (Squamata: Scincidae) in the Sydney region. Australian Journal of Zoology 33:699 704. Kusch, R. C., and D. P. Chivers. 2004. The effects of crayfish predation on phenotypic and life-history variation in fathead minnows. Canadian Journal of Zoology 82: 917 921. Lima, S. L. 1998. Stress and decision making under the risk of predation: recent developments from behavioral, reproductive and ecological perspectives. Advances in the Study of Behavior 27:215 290. Lima, S. L., and P. A. Bednekoff. 1999. Temporal variation in danger drives antipredator behavior: the predation risk allocation hypothesis. American Naturalist 153:649 659. Lima, S. L., and L. M. Dill. 1990. Behavioural decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68:619 640. Martin, K. L. M. 1999. Ready and waiting: delayed hatching and extended incubation of anamniotic vertebrate terrestrial eggs. American Zoologist 39:279 288. Moreira, P. L., and M. S. T. Barata. 2005. Egg mortality and early embryo hatching caused by fungal infection of Iberian rock lizard (Lacerta monticola) clutches. Herpetological Journal 15:265 272. R Development Core Team. 2009. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Shea, G. M., and R. A. Sadlier. 2000. A mixed communal nest of skink eggs. Herpetofauna 30:46 47. Shine, R. 1980. Comparative ecology of three Australian snake species of the genus Cacophis (Serpentes: Elapidae). Copeia 1980:831 838. Smith, G. R., and D. T. Fortune. 2009. Hatching plasticity of wood frog (Rana sylvatica) eggs in response to mosquitofish (Gambusia affinis) cues. Herpetological Conservation and Biology 4:43 47. Taylor, J. E., and B. J. Fox. 1991. Recolonization of regenerating open forest by terrestrial lizards following sand mining. Australian Journal of Ecology 16:137 148. Thompson, M. B., B. K. Speake, K. J. Russell, and J. J. McCartney. 2001. Utilization of lipids, protein, ions and energy during embryonic development of Australian Copeia cope-13-01-21.3d 20/12/12 21:42:36 163 Cust # CE-12-111

164 Copeia 2013, No. 1 oviparous skinks in the genus Lampropholis. Comparative Biochemistry and Physiology 129A:313 326. Vitt, L. J. 1991. Ecology and life history of the scansorial arboreal lizard Plica plica (Iguanidae) in Amazonian Brazil. Canadian Journal of Zoology 69:504 511. Vonesh, J. R. 2000. Dipteran predation on the eggs of four Hyperolius frog species in western Uganda. Copeia 2000:560 566. Wapstra, M., and E. Wapstra. 1986. Breeding record of the delicate skink, Leiolopisma delicata, in southern Tasmania. Tasmanian Naturalist 87:5 6. Warkentin, K. M. 1995. Adaptive plasticity in hatching age: a response to predation risk trade-offs. Proceedings of the National Academy of Sciences of the United States of America 92:3507 3510. Warkentin, K. M. 2000. Wasp predation and wasp-induced hatching of red-eyed treefrog eggs. Animal Behaviour 60:503 510. Warkentin, K. M. 2005. How do embryos assess risk? Vibrational cues in predator-induced hatching of red-eyed treefrogs. Animal Behaviour 70:59 71. Warkentin, K. M. 2011a. Environmentally cued hatching across taxa: embryos respond to risk and opportunity. Integrative and Comparative Biology 51:14 25. Warkentin, K. M. 2011b. Plasticity of hatching in amphibians: evolution, trade-offs, cues and mechanisms. Integrative and Comparative Biology 51:111 127. Warkentin, K. M., and M. S. Caldwell. 2009. Assessing risk: embryos, information, and escape hatching, p. 177 200. In: Cognitive Ecology II: The Evolutionary Ecology of Learning, Memory, and Information Use. R. Dukas and J. Ratcliffe (eds.). University of Chicago Press, Chicago. Warkentin, K. M., M. S. Caldwell, and J. G. McDaniel. 2006. Temporal pattern cues in vibrational risk assessment by red-eyed treefrog embryos, Agalychnis callidryas. Journal of Experimental Biology 209:1376 1384. Warkentin, K. M., M. S. Caldwell, T. D. Siok, A. T. D Amato, and J. G. McDaniel. 2007. Flexible information sampling in vibrational risk assessment by red-eyed treefrog embryos. Journal of Experimental Biology 210:614 619. Copeia cope-13-01-21.3d 20/12/12 21:42:36 164 Cust # CE-12-111