The diets of sympatric rodents partially define their realized niches. Identifying items in stomachs of introduced rodents helps determine rodents' trophic positions and species most at risk of consumption. In the Hawaiian Islands, which lacked rodents prior to human arrival, three rodents (Rattus rattus or black rat, R. exulans or Pacific rat, Mus musculus or house mouse) commonly coexist in native habitats where they consume a wide range of plants and animals. These three rodent species were trapped in montane forest for 2.5 years; their stomach contents were analyzed to determine short-term diets (n = 12-95 indiv. per species), and isotopic fractions of d 15 N and d 13C in their bone collagen were analyzed to further estimate their trophic positions (n = 11-20 indiv. per species). For all three species, [75 % of individuals had plants and [90 % had arthropods in their stomachs, and significant differences in mean relative abundances were found for food items in stomachs among all three rodents. Rodents may be dispersing some native and non-native seeds, including the highly invasive Clidemia hirta. Most identifiable arthropods in rodent stomachs were non-native, and no stomachs contained birds, snails, or lizards. The d 15 N and d 13 C signatures were consistent with trophic feeding differences revealed from stomach contents. Dietary niche differentiation by coexisting rodent species is evident in this forest, with Pacific rats being intermediate between the mostly carnivorous house mouse and the mostly herbivorous black rat; such findings can help forecast rodent impacts and direct management efforts in ecosystems where these invasive animals coexist.

Large-scale rodent control reduces pre-and postdispersal seed predation of the endangered Hawaiian lobeliad, Cyanea superba subsp. superba (Campanulaceae) Abstract Large-scale rodent control can help to manage endangered species that are vulnerable to invasive rodent consumption. A 26 ha rodent snap-trap grid was installed in montane forest on Oahu Island, Hawaii, in order to protect endangered snails and plants.RichardTo assess the effectiveness of this trapping operation in reducing fruit consumption and seed predation of the endangered Hawaiian lobeliad, Cyanea superba subsp. superba, pre-and post-dispersal C. superba fruit consumption were monitored for 36 plants at the site with rodent control (Kahanahaiki) and 42 plants at an adjacent site without rodent control (Pahole). Over 47 % of all monitored fruit were eaten on the plants at Pahole compared to 4 % at Kahanahaiki. Images captured using motion-sensing cameras suggest that black rats (Rattus rattus) were the only pre-dispersal fruit consumers. To quantify post-dispersal fruit consumption, and to identify the culprit frugivore(s), mature fruit were placed in tracking tunnels positioned on the forest floor and checked daily. At Pahole, all of the fruit were consumed by rats compared to 29 % at Kahanahaiki. Lastly, to determine if rodents from the sites were predators or dispersers of C. superba seed, fruit were fed to captive black rats and house mice (Mus musculus). Black rats consumed entire fruit, killing all the seed, while mice did little damage to the fruit and seed. Therefore, large-scale rat trapping can directly benefit the reproduction of C. superba subsp. superba. Controlling black rats at restoration sites appears integral to the successful restoration of this endangered plant species.