Space Use by a Neotropical Water Strider (Hemiptera: Gerridae): Sex and Age-class' Differences

Differential movement patterns and microhabitat distributions of age classes have recently been repotted in several water surface-inhabiting insects Uamieson and Scudder 1974, Vepsalainen and Jarvinen 1974, Wilson et al. 1978, Birch et al. 1979). Juvenile age classes may be displaced because of aggression by adults (Wilson et al. 1978) or because of selection for different microhabitats (Vepsalainen and Jarvinen 1974). Differential use of space by adult males and females may be the consequence of other behaviors, such as selection of mates or oviposition sites. This note describes differences in movement patterns and home range size by sex and age classes of a neotropical water strider, Potamobates tridentatus Esaki (Hemiptera: Gerridae). Although the intricate locomotory adaptations enabling water striders to skate on the water surface are understood (Milne and Milne 1978), ecological aspeCts of their movements remain poorly known. As with most insects, we know how water striders move, but not where they go, or how sexes and age classes differ in space use. We present evidence that distincr behavioral patterns are the consequence of differences in body size and ecological requisites, most importantly food and mates. From 12 to 23 August, 1978, we studied water strider movement patterns in a small stream, the Quebrada Camaronal, at Sirena, Parque Nacional del Corcovado, Osa Peninsula, Costa Rica (8°23'N, 83°20'W), about one kilometer upstream from the Pacific Ocean. The vegetation surrounding the study site is predominantly secondary growth within Tropical Wet Forest (Holdridge 1967), with a discontinuous canopy and no emergent vegetation in the stream. In the mid-rainy season the stream in the study site varied in width from one to four meters and in depth from 5 to 60 cill. Current speed averaged 0.13 m'sec-1 and water temperature 25.5°C. We captured 116 P. tridentatus individuals from the first 20 m of the study site and individually marked early instars, fifth instars, adult males, and adult females by painting the mesonota with small spots of enamel paint. We noticed no change in the behavior of water striders after marking. Third and fourth instars were indistinguishable, but fifth instars could be recognized by dorsal pattern, larger size, and the presence of wing buds. The sex of subadults could not be determined. Adult males and females differed in the form of their conspicuous genitalia. We measured middle femur lengths (chosen as a morphological characrer because of its direcr relationship to stride length) on several individual third-fourth instars, fifth instars, and adult females and males; lengths averaged, respectively, 5.7 mm, 6.0 rom, 9.0 rom, and 8.9 mm. Mter water striders were marked and left undisturbed for 24 h, we censused the study area in the momings (0700 h) and/or afternoons (1600 h) on 13 August and from 16 to 23 August. We recorded the location of all individually marked water striders to the nearest 5 meters along a 115 m stream transect during each of 14 censuses. To determine space use on a finer scale, we studied movements while individuals foraged actively during the morning beneath a 2.5 m by 2.5 m frame suspended above the stream 20 m upstream from the lower end of the study site. The frame was divided by a string grid into 100 quadtats, each 625 cm2. We conducted focal animal sampling for 25 five minute periods, noting at five second intervals the coordinates of the quadtat in which we observed marked individuals. If a focal animal did not remain under the grid for the entire five minute period, that sample was considered an incomplete observation and the data were not included in subsequent analyses. We quantified four aspects of P. tridentatus movement patterns for each age and sex class: home range, activiry center, and between-census movement distance, for which we used census data; and local feeding area diversity, for which we used five minute observation sample data. Home range was defined as the maximum distance between all observed locations of an individual; it is therefore a linear measure rather than a traditional measure of area. Activity center was calculated as the mean location (position along transecr) of all resightings. We computed home ranges and activity centers only for those animals that were resighted on at least half of the 14 censuses. Between-census movement distances measured the distance moved between consecutive censuses. To determine local feeding area diversity (the tendency to forage widely or remain in a restricted area), we used the Shannon-Wiener information index, H' = -}; Pi in p; (Pielou 1974), where Pi was the proporrion of time spent in each quadtat during each five minute sampling period. During the censuses, focal animal sampling, and miscellaneous observations between census periods, we noted feeding behavior and the identities of copulating individuals. Unless otherwise indicated, all statistical tests are twotailed and are desctibed in Sokal and Rohlf (1969). The proportions of adult male, adult female, and fifth instar water striders that were resighted on at least half of the censuses were high-o.71 (N = 35), 0.82 (N = 22), and 0.78 (N = 9), respectively-and not significantly