The effect of multiple stressors on Kaua'i's forest birds: demography, health, and trophic cascades

Abstract

Island ecosystems harbor unparalleled biodiversity, providing habitat for a disproportionate share of the world's species relative to their area. Nevertheless, the vulnerability of islands to disturbances that result in negative consequences renders them among the most endangered environments. Bird populations on islands are particularly susceptible and often strongly affected by invasive species, habitat loss, climate change, and disease, all of which has resulted in widespread decline and extinction. In the Hawaiian Islands, native forest bird populations confront numerous threats, with the majority of remaining native bird populations classified as threatened, endangered, or teetering on the brink of extinction. To sustain and recover island bird populations, it is essential to understand the direct and indirect impacts of anthropogenic changes on forest bird populations, and to what degree each driver of change influences bird health, demographics, and habitat quality. The goals of this project were to quantify how various stressors influence the body condition and sex ratios of Kaua'i's forest birds, and to assess if primary and secondary forest regeneration processes that affect forest bird resource and habitat availability are influenced by invasive rodent control. To address these questions, I collected and analyzed three years of forest bird banding data in the Alaka'i Wilderness Preserve on Kaua'i's Island (Hawai'i, U.S.A.) from 2021 to 2023. Birds were banded in and outside of areas that have been subject to intensive rodent control for the past 5-9 years. I focused on four focal species of birds, three native and one non-native, and collected data to assess the body condition and sex of these captured birds. Mixed effects models were used to determine whether body condition was significantly associated with rodent control intensity on a local and landscape scale, malarial infection status, mean temperature during the peak mosquito breeding season, frequency of extreme rain events, canopy density, and topographic wetness index (TWI). Similarly, I used general linear models to test if rodent control intensity, malarial infection status, or date of capture was associated with a higher or lower likelihood of encountering a female. I found that body condition was negatively associated with positive malarial status for all but one of the focal species, and higher average temperatures and TWI were negatively associated with body condition for all species. None of the predictor variables evaluated were strong predictors of the sex of captured birds. To evaluate the effects of rodent control on forest regeneration processes that sustain bird habitat in the Alaka'i Wilderness Preserve, I examined whether rodent control was associated with changes in seed rain, fruiting density, and seedling emergence. Seed traps were established in February of 2023 and were checked periodically over a six-month period to obtain seed rain data from within and outside of areas of rodent control. Seeds collected from traps were identified to the lowest possible taxonomic level. During each visit to collect seed trap contents, the number of fruits on all fruiting plants within five meters of each seed rain trap was estimated. These estimates were used to calculate relative fruiting densities in treatment and reference plots. To examine seedling emergence, emergence plots in treatment and reference plots were cleared of all vegetation in July of 2022. I used Mixed Effects Models to compare seed rain, fruiting density, and seedling emergence between sites with and without rodent control and found no differences in seed rain, seedling emergence, or fruiting density within and outside of rodent control. My findings suggest that variables that are related to disease transmission may be more influential in predicting the health of forest birds than variables related to predation risk or food and resource availability. These findings also suggest that rodent control, at least at current levels of intensity, may not play a significant role in forest regeneration in Kaua'i's 'Alaka'i Plateau. Further research is needed to understand the effectiveness of rodent control in this system, but reducing the prevalence of avian malaria and mitigating climate change will be critical to alleviating sub-lethal effects on Kaua'i's forest birds. This study contributes to our understanding of the sometimes complex and synergistic effects of invasive predators, disease, forest structure, and climate on forest birds and the dynamics of insular ecosystems. My findings and approach could have implications for conservation and restoration in the many areas globally where rat invasion, climate change, and disease may interact to pose similar threats.