Collared Kingfisher - nest predatorGolden White-Eye at the nest - you can barely see his beakBridled White-Eye nest - part of the focal studiesMicronesian Honey-Eater nest - 1st documented sighting of nest found on Saipan The Effects of Nest Predation on the Forest Birds of the Mariana Islands

Project Officer: Dana Winkelman

Project Officer: Gordon Rodda, USGS, Fort Collins Science Center

Principal Investigator: Julie Savidge

Graduate Student: Thalia Sachtleben

Project Start Date: 09/02

Expected Completion Date: 10/01/05

Funding Agency: DOI, USGS, BRD, RWO 72

Introduction: Avian research in the Mariana Islands has been limited. Those studies that have addressed reproductive success have reported low nest success for some species. Predation by introduced rats is thought to be a contributing factor on the island of Saipan. Additionally, non-native vegetation dominates much of Saipan, and it is unknown how predation on nests may vary between native and non-native forest. This study will provide information on the importance of nest predation for an endangered species and 2 relatively common species of birds on Saipan. Additionally, we will examine nest success in native versus non-native forest. A rodent reduction experiment, combined with population estimates of rodents, would elucidate the potential amount of rodent reduction that could be achieved on Saipan, and combined with data on actual predation rates on nests, suggest how reduction might benefit island bird populations. Baseline data on rodent population densities will be very important if brown treesnakes are or become established on Saipan.

Objectives: 1) Identify the predators responsible for predation on nests of target bird species on the island of Saipan; 2) quantify the abundance of rats in native vs. non-native forest; 3) Assess nest success and predation rates on nests of breeding birds in native vs. non-native forest; 4) quantify nest densities in native vs. non-native forest; 5) determine whether certain nest site characteristics make nests more vulnerable to predation; 6) if rodents are found to be important nest predators and if permits can be
obtained, reduce rodents on certain study plots and follow avian nest success.

Progress: The first field season was completed. Four native forest and 4 non-native forest study sites were selected and vegetation characterized. A total of 206 Golden White-eye, Bridled White-eye, and Rufous Fantail nests were monitored. Apparent nest success averaged between 50-67% (daily survival rates are yet to be determined). All nest failures were due to predation. We worked with still and video cameras to try and capture predation events. Many problems were encountered with the still cameras. Video cameras appeared to have more utility and predation on a nest by a native Micronesian Starling was documented. Difficulties were also encountered with the rat trapping efforts (primarily disturbance of traps by nontargets), and it was decided that this portion of the project would be discontinued. The second field season was completed in May 2004, and data analyses are almost complete. 120 Bridled White-eye, 73 Golden White-eye, 66 Rufous Fantail nests, and 2 Nightingale Reed-warbler nests were monitored in 2004. Both Nightingale Reed-warbler nests fledged young. All nest failures were due to predation, except for 3 Golden White-eye and 7 Bridled White-eye nests that failed due to desertion. Daily survival rates of Golden White-eyes and Rufous Fantails were found to be constant across forest types (native and non-native) and years (2003 and 2004), while daily survival rates of Bridled White-eyes were constant across forest types but differed by year. Daily survival rates were similar for species and years with the exception of Bridled White-eyes in 2004. Daily survival rates for Bridled White-eyes in 2004 were lower than those for Bridled White-eyes in 2003, and were lower than daily survival rates for Rufous Fantails and Golden White-eyes in both years. Line transect sampling using distance sampling methodology was used to quantify nest densities in native and non-native habitat. Transects were surveyed once in 2003 and three times in 2004. Differences in nest densities were assessed and found to differ by species, forest type, and survey period. Nest densities ranged from 0 – 115 nests/km2 for Bridled White-eyes, from 0 – 41 nests/km2 for Golden White-eyes, and from 0 – 24 nests/ km2 for Rufous Fantails. Six predation events were recorded through imprints left by predators in wax bait eggs (used in conjunction with the still cameras), and three predation events were recorded with video cameras. Two predations were perpetrated by native Collared Kingfishers, and six by Micronesian Starlings. One predator was an unidentified small bird (suspected to be the parent bird trying to remove the bait egg from the nest). The effects of nest site placement characteristics on nest survival differed by species. No nest site characteristics appeared to influence Rufous Fantail nest survival, while the number of nest support branches positively affected Golden White-eye nest survival, and nest concealment negatively affected Bridled White-eye nest survival.