We report evidence from controlled experiments that long straight thorns deter herbivory by browsers. Cut branches of three woody species that had their thorns removed suffered significantly greater herbivory by a tethered goat than did paired intact branches. Branches on living Acacia seyal plants that had their thorns removed suffered significantly greater herbivory by a wild population of free-ranging giraffes than did intact branches on the same plants. These differences in herbivory resulted in long term losses of branch length in clipped as opposed to control branches. In addition, branches within reach of giraffes produced longer thorns and a greater density of thorns than did higher branches. These results imply that increased thorn length is an induced defense.

We explore here the occurrence of aggressive ants in an apparently symbiotic relationship with the savanna tree Acacia drepanolobium and their effects on giraffe herbivory on the Athi-Kapiti Plains, Kenya. Trees taller than 1.3 m were more likely to be occupied by aggressive ants in the genus Crematogaster than were shorter trees. Ants wereconcentrated on shoot tips, the plant parts preferred by giraffes. Trees with relatively more foliage had more swarming ants than did trees with less foliage. The feeding behavior of individual freeranging giraffes on Acacia drepanolobium was studied. Giraffe calves exhibited a strong sensitivity to Crematogaster ants inhabiting A. drepanolobium, feeding for significantly shorter periods on trees with a greater number of aggressive ants. Older giraffes were apparently less sensitive to ants, and did not feed for shorter periods on trees with fuller foliage, despite significantly greater ant activity on these plants. The thorns of A. drepanolobium are significantly shorter than are the thorns of A. seyal, a species without symbiotic ants, a pattern that may indicate a trade-off between ants and thorns as defenses.

Sea turtle egg mortality, egg predation, and small organisms associated with turtle nests were studied at Playa Ostional, Costa Rica. Sites with concentrated sea turtle nesting were compared with solitary nesting sites as a function of place and time based on ANOVA, Akaike's Information Criterion, and Bayesian analyses. Results indicate that sea turtle egg mortality was significantly associated (P < 0.005) with flowing water that erodes or saturates nesting sites, and with overlapped nesting in which sea turtles disturb each other's nests. Sarcophagid and calliphorid fly larvae (Bayesian prior = 1.19; posterior = 2.27), fungi (prior = 1.14; posterior = 1.92), mites (prior = 0.51; posterior = 1.15), and several other types of small organisms increased in number after turtle egg laying (N = 303 nests; 34,451 turtle eggs). During peak sea turtle nesting periods, visitation to nesting sites by poachers and vertebrate predators was high, and relative number of nests disturbed by these predators was low (P < 0.02). In multimodel analysis, the three most parsimonious models were: (1) turtle egg mortality and distance from mean high tide; (2) turtle egg predation and distance from mean high tide; and (3) turtle egg mortality and nesting density, with Akaike weights of 0.224, 0.203, and 0.153 respectively. Intensive sea turtle nesting might result in upwelling and turnover of nesting debris and nest organisms, and may influence biotic community structure of sandy beach ecosystems.Abstract in Portuguese is available at http://www.blackwell-synergy.com/loi/btp.

Some plants respond to browsing with compensatory regrowth of plant tissues and with increased thorn growth. Associations between browsers and their preferred forage were examined through wandering quarter vegetation sampling and observational studies in an effort to understand how some plants respond to browsing by large African herbivores. Acacia seyal (n ¼ 2680) A. drepanolobium (n ¼ 1850), and Balanites glabra (n ¼ 960) were three species of frequently browsed indigenous plants examined on Game Ranching Ltd. in Kenya. There were several statistically significant associations revealed. Individual A. seyal exposed to intensive browser utilization were observed to lose shoot tips, produce long thorns, and have relatively few flowers and fruits. Browser utilization was associated with increased lateral branching in A. drepanolobium and with an increased occurrence of short, thickened spines in B. glabra. Thorns, spines and flowers were measurable indicators of relative browser utilization, and may be useful features to monitor in the management of large African mammals and their prickly forage