Salmon, Michael

Artificial lighting disrupts sea turtle hatchling orientation from the nest to the sea. I studied how a light-induced landward crawl affects the ability of hatchlings to later crawl to the sea, and swim offshore from a dark beach. A brief (2 min) landward crawl had no effect on orientation, as long as waves (used as an orientation cue while swimming) were present. In the absence of waves (a flat calm sea), landward-crawling hatchlings failed to swim offshore while those crawling seaward were well oriented. A longer (2 h) landward crawl impaired the ability of hatchlings to crawl to the sea. These results demonstrate that previous exposure to artificial lighting compromises subsequent orientation, both on land and in the sea. On the basis of my results, I suggest several changes to current management practices, currently used when releasing misoriented turtles in the wild.

Pole-mounted street lighting on coastal roadways is often visible in adjacent areas. At roadways near sea turtle nesting beaches, these lights can disrupt the nocturnal orientation of hatchlings as they crawl from the nest to the sea. Our objective was to determine if an alternative lighting system (light-emitting diodes, embedded in the roadway pavement) prevented orientation disruption of loggerhead hatchlings. Hatchlings at the beach oriented normally when the embedded lights were on, or when all lighting was switched off. However, turtles showed poor orientation when exposed to pole-mounted street lighting. Light measurements revealed that street lighting was present at the beach, whereas embedded lighting was absent. I conclude that embedded lighting systems restrict light scatter, leaving adjacent habitats dark, and therefore protect the turtles from artificial lighting allowing for normal seafinding.

This study's objective was to determine if the transfer of a crawling direction to a magnetic compass in loggerhead hatchling sea turtles ( Caretta caretta L.) was facilitated by how long the turtle crawled (an "endogenous timing" component). I first determined how long it took hatchlings to crawl from their nest to the ocean. Two types of experiments were then carried out. In the first, crawling time varied. In the second, both crawling time and direction varied. I found that at most beaches hatchlings crawled to the ocean in less than 5 min. My experiments showed that if crawls are too short (1 min), or too long (5 min), vector transfer is weakened compared to a 2 min crawl. I also found that a period of non-directional crawling interfered with the ability of a 2 min crawl to promote calibration. These results confirm that efficient transfer of a crawling vector, maintained by visual compass, to a swimming vector, maintained by a magnetic compass, depends upon an endogenous timing program in hatchlings. The temporal properties of that program are, in turn, apparently shaped by where their mothers place nests on the beach.

This study's objectives were to determine if coastal (shallow-water, oceanic reef) aggregations of juvenile green turtles (Chelonia mydas L.) in Palm Beach, Florida occupied distinct home ranges and how these home ranges compared in size and resource availability with those studied elsewhere. Six immature green turtles were captured, measured, and subjected to esophageal lavage to determine diet. Each turtle was returned to its initial capture site within 24 h with an ultrasonic transmitter used to track movements. All turtles were <65 cm SCL, had ingested similar macroalgae, and occupied markedly small home ranges (mean = 2.38 +/- 1.78 km 2), largely restricted to the reef itself. Diving and feeding activity peaked during the day; at night, activity was minimal. The food and sleeping site resource distribution at this specific location coincides with the turtles' home range size and shape, with considerable overlap of core areas.

Previous qualitative assessment indicated that signature whistles of temporarily captured, free-ranging dolphins remain stable over periods of 2--12 years. This study reports on the stability of signature whistle parameters in wild female Atlantic spotted dolphins in the Bahamas over five or more years and between changes in age class. Signature whistles from seven female dolphins were pooled into blocks of 'early' and 'late' years for the time assessment. Signature whistles from five females were pooled by age class for the second analysis. Duration, minimum frequency, maximum frequency, and change in frequency were measured from spectrograms of whistles for statistical analysis. No significant changes were found in any of the signature whistle parameters either between early and late periods of time or with a change in age class.

In the ocean, lighting varies with habitat; the eye's spectral sensitivity must vary with visual ecology. Green turtles are the only sea turtle whose spectral sensitivity has been studied. Loggerheads and leatherbacks see visible light between 340 and 700 nm. However, the wavelengths detected with the greatest sensitivity by both species are those best transmitted at the specific depths where food, mates and predators are likely to be encountered. Both species have trichromatic vision, but the species differ in the concentration and peak sensitivity of each visual pigment resulting in either a broadly tuned (loggerhead) or finely tuned (leatherback) spectral sensitivity. Spectral sensitivity of leatherbacks overlaps both bioluminescence of prey, and light available in clear, deep, oceanic waters.

Hatchling sea turtles use visual cues to orient to the water. Streetlights placed on coastal roadways can attract the turtles inland. Filters were designed to be used with coastal roadway lighting to eliminate the more harmful wavelengths of light. I tested the General Electric 2422 and NLW filters in a laboratory setting with hatchling loggerhead and green turtles. Both species of turtles were attracted to the amber filtered lighting in arena experiments. Loggerhead hatchlings were used in T-maze experiments where they were given a choice between amber filtered and unfiltered lighting. The turtles preferred the unfiltered lighting to the filtered lighting, even when it was 100 to 1000 times dimmer. I conclude that amber filtered lighting does afford some protection to sea turtles, although it must be used in conjunction with other light management techniques to prevent the disruption of hatchling turtle orientation.

Leatherback sea turtles (Dermochelys coriacea) feed exclusively on gelatinous prey. Hatchlings are solitary and must possess a predisposition to respond to prey. In laboratory experiments, I studied the responses of nineteen leatherback hatchlings to visual (jellyfish model and shapes: circle, square, diamond) and chemical (homogenates of three prey) stimuli presented alone or as paired (visual + chemical) treatments once daily. When presented alone visual stimuli resembling jellyfish outlines elicited stronger feeding responses (changes in locomotion and orientation) than those not resembling jellyfish. Chemical stimuli alone induced a rheotaxis, but responses evoked by some homogenates were stronger than responses to others. Paired stimuli evoked stronger orientation and more consistent increases in swimming (flipper stroke) rate, indicating additive effects. Results suggest that both stimuli elicit food searching behavior and when they begin to forage, hatchlings already possess predispositions to respond to an adaptive array of prey shapes and odors.

Under normal conditions, hatchling sea turtles crawl toward the ocean but streetlights placed on coastal roadways can attract the turtles toward land. Two light filters were designed to exclude the shorter light wavelengths most attractive to turtles. I did laboratory tests to determine if green turtle and loggerhead hatchlings oriented normally ("seaward") in the presence of filtered lighting. Light passed through either filter (#2422 acrylic and NLW) attracted the turtles unless coastal cues (an elevated horizon) were strong or background (full moon) illumination was present. Green turtles and loggerheads responded differently to the same filters, indicating that neither filter provided adequate protection for both species. I conclude that these filters fail to protect the turtles. Conventional forms of light control (shielding and/or lowering light fixtures, decreasing wattage, or turning off problem lights) remain the best way to shield turtles from the harmful effects of artificial lighting.

The purpose of this study was to determine if filtered street lighting affected the nesting behavior of loggerhead sea turtles ( Caretta caretta L.). My study site was a nesting beach at Carlin Park, Jupiter, Florida. During the 1999 nesting season, portions of the beach were either kept dark or were illuminated by four 70 W high-pressure sodium (HPS) streetlights, each housed in a cut-off fixture covered by an acrylic (model #2422) filter. These filters excluded all light wavelengths below 540 nm. The excluded wavelengths repel nesting females. Daily counts of nesting and non-nesting crawls were made. Data from the 1999 nesting season were compared to historical records of nesting at the site between 1990 and 1998. I found no evidence that filtered lights affected nesting densities, or the ratio of successful to unsuccessful crawls. These results suggest that at Carlin Park, the nesting behavior of loggerhead females is unaffected by exposure of the beach to filtered street lighting.