Teleost fishes typically first encounter the environment as free-swimming embryos or larvae. Larvae are morphologically distinct from adults, and major anatomical structures are unformed. Thus, larvae undergo a series of dramatic morphological changes until they reach adult morphology (but are reproductively immature) and are considered juveniles. Free-swimming embryos and larvae are able to perform a C-start, an effective escape response that is used evade predators. However, escape response performance improves during early development: as young fish grow, they swim faster (length-specific maximum velocity increases) and perform the escape more rapidly (time to complete the behavior decreases). These improvements cease when fish become juveniles, although absolute swimming velocity (m s(-1)) continues to increase. We use studies of escape behavior and ontogeny in California halibut (Paralichthys californicus), rainbow trout (Oncorhynchus mykiss), and razorback suckers (Xyrauchen texanus) to test the hypothesis that specific morphological changes improve escape performance. We suggest that formation of the caudal fin improves energy transfer to the water and therefore increases thrust production and swimming velocity. In addition, changes to the axial skeleton during the larval period produce increased axial stiffness, which in turn allows the production of a more rapid and effective escape response. Because escape performance improves as adult morphology develops, fish that enter the environment in an advanced stage of development (i.e., those with direct development) should have a greater ability to evade predators than do fish that enter the environment at an early stage of development (i.e., those with indirect development).

Blood chemistry is commonly used to explore the dynamics of health in captive animals in biomedical facilities. Yet, studies of blood serum chemistry in teleost fish are almost nonexistent (1). Such information would be beneficial in determining the baseline health and physiology of captive animals. Physiological assays would allow pre-screening of experimental subjects and decrease the loss of time, effort, and animals by indicating which fish are most suited to survive the rigors of surgery, experimentation, and long-term maintenance. In mariculture, early indication of disease or infection would allow therapeutic measures to be. taken to alleviate adverse conditions and resultant death or disease. The oyster toadfish, Opsanus tuu, is a valuable vertebrate model for biomedical research (2, 3), but little is known about its blood chemistry. This study's objective is to establish some initial baseline values for blood parameters that may be important in understanding the health of the captive toadfish. Male and female toadfish of various sizes (200-1200 g; 18-33 cm-standard length) were selected from the holding tanks in the Marine Resources Center (MRC) of the Marine Biological Laboratory. Selection was based on the alertness (erect fin position), integument appearance, and regular respiration of the fish-the same characteristics investigators currently use to choose' toadfish for experiments. Blood was collected in July 1997 (n = 20) and June 1998 (n = 40), between 0800 and 1000 h, from recently captured fish that had been acclimated for l-2 weeks in MRC holding tanks. These fish were taken from the general population in the holding tanks, and therefore their health status was unknown. Consequently, any fish that died within 45 days of the initial blood sampling (less than 10%) were excluded from the study. Within 1.5 min after the fish were anesthetized in a 0.001% solution of MS-222 (tricaine methanesulfonate), blood was drawn by caudal puncture. A 3ml heparinized syringe and a 25gauge needle were used to extract l-3 ml of mixed arterio-venous blood. Each fish was then weighed, measured, sexed, and tagged with an Avid radio tag. Only one blood sample was taken from each fish. The percentage of the blood volume composed of red blood cells (hematocrit) was determined by transferring a portion of each blood sample directly from the syringe into capillary tubes, which were centrifuged and then read using a Monoject Scientific Critocaps micro-hematocrit capillary tube reader. The remaining blood

Poor recruitment has generated the hypothesis that the endangered razorback sucker Xyrauchen texanus is particularly vulnerable to predation early in its life history. We compared the escape responses of razorback suckers, which are adapted to the historically warm waters of the Colorado River, with those of rainbow trout Oncorhynchus mykiss, an introduced coldwater species, throughout early development at water temperatures of 12°C and 18°C. We quantified escape performance using maximum velocity, acceleration, and time to maximum velocity and acceleration. Both species showed complete temperature compensation for escape performance; individuals reared at 12°C performed as well as those reared at 18°C. Performance was also similar between species, although two variables exhibited a species × size interaction. Small razorback suckers were faster (greater maximum velocity and acceleration) than small rainbow trout, while large larvae performed similarly. We also determined that razorback sucker larval escape performance falls within the range reported for other fishes. Therefore, we conclude that razorback suckers do not have “poor” escape performance and that temperature does not directly cause decreased performance. However, a cold temperature reduces growth rates and delays razorback suckers' attainment of a “predator‐proof” size. Small larvae are also more likely to perform uncoordinated, ineffective escape responses. Hence, razorback sucker performance is indirectly diminished by temperature.