Restoring island birds and seed dispersal in New Zealand's fenced mainland island sanctuaries
Date
2018
Authors
Bombaci, Sara Petrita, author
Pejchar, Liba, advisor
Reed, Sarah, committee member
Savidge, Julie, committee member
Smith, Melinda, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Island ecosystems are global biodiversity hotspots, but many island species face population declines and extinction. These losses are mainly driven by invasive mammals that consume or compete with native animals and degrade their habitats. The decline of island animal populations may also impact ecosystem processes that depend on them, e.g. seed dispersal, pollination, and nutrient cycling. The island nation of New Zealand has pioneered a unique solution – fenced mainland island sanctuaries – which exclude invasive mammals from natural habitats and provide opportunities to restore native birds and other wildlife. Yet, critics question whether sanctuaries, which are costly and require continuous maintenance, effectively conserve birds and ecosystems, given minimal research on sanctuary project outcomes. I assessed if sanctuaries are an effective conservation tool for restoring birds and seed dispersal in New Zealand. I compared bird population densities and bird-mediated seed dispersal in three fenced sanctuary sites to three paired reference sites (with minimal mammal control). From January-April 2016 and 2017, I set seed traps to measure dispersed-seed abundance, conducted focal tree observations to determine foraging rates for six tree species, and used distance sampling-based point counts to survey birds at randomly placed sampling locations within each site. I supplemented my bird surveys with a paired acoustic sampling method that uses acoustic recorders to increase survey sample size when estimating population densities with distance sampling. I tested the effectiveness of paired acoustic sampling for monitoring New Zealand forest bird populations by assessing whether density estimates from acoustic data were biased relative to densities estimated from human point counts, and by assessing whether the paired acoustic sampling method corrects bias from acoustic data, when present. Thus, the objectives of this dissertation were to assess 1) whether audio data could be used to estimate population densities for New Zealand forest birds (Chapter 1), 2) whether fenced mainland island sanctuaries increase the density of native or introduced bird species relative to unprotected areas in New Zealand (Chapter 2), and 3) whether sanctuaries enhance bird-mediated seed dispersal (foraging rates and dispersed-seed abundance) relative to unprotected areas (Chapter 3). I found that acoustic recorders underestimated bird population densities for four bird species but incorporating statistical offsets from the paired sampling method in generalized linear mixed models corrected the bias for all four species. Across both years, I found 0.27 to 9.00 more birds/ha on average for nine of twelve native bird species (including seven frugivores) in sanctuaries compared to unprotected sites, and no difference in mean population densities for three introduced bird species (two frugivores) and three biogeographically recent native species (one frugivore). Mean foraging rates and dispersed seed counts were also higher (0.1-0.6 more fruits consumed/observation period; 2-22 more seeds dispersed/plot) in sanctuary sites for several native tree species. Frugivore density, or both frugivore density and fruit abundance were significant positive predictors of foraging rates and dispersed seed counts for most tree species. Finally, native bird densities were correlated with foraging rates and dispersed seed counts for most tree species, but introduced bird densities were rarely correlated with foraging or dispersed seed counts. Thus, higher densities of native frugivorous birds in sanctuary sites appears to have resulted in increased fruit removal and dispersed seed counts for several native tree species. My study is one of the first to demonstrate that fenced mainland island sanctuaries, which require a substantial investment of conservation funds, are meeting ecological objectives. Furthermore, I demonstrate that paired acoustic sampling can be used to produce unbiased population density estimates from acoustic data, relative to human point counts, which holds great promise for increasing the scope and efficiency of bird population monitoring in New Zealand. I show that sanctuaries increase bird population densities for several native bird species and have no effect on introduced and biogeographically recent native species. These findings support predictions made over a decade ago on the potential 'winners' and 'losers' of mammal eradication and offer evidence that fenced sanctuaries effectively conserve New Zealand's native bird populations. By increasing bird densities and seed dispersal, fenced mainland island sanctuaries could be a viable tool for restoring other island and mainland ecosystems under threat from invasive mammals.
Description
Rights Access
Subject
bird conservation
island biodiversity
restoration
invasive mammal control
bioacoustics
plant-animal interaction