Nest success, feeding behavior and anti-predator strategies, were studied to determine if bellbirds (Anthornis melanura) are limited by predation and if pest control can relieve the pressures of predation so that food supply may become a limiting factor for bellbirds. The nest success of bellbirds at the Rotoiti Nature Recovery Project (RNRP), Nelson Lakes, New Zealand (2002-2004 breeding seasons) was compared with data from other bellbird nest success studies in New Zealand. Two years of nest monitoring resulted in a low overall nest success at the RNRP (39 %), an area low in predator density, compared with 16.4 - 67.2 % found for other sites. One-way anova comparisons between the nest success of bellbirds in areas grouped by predator density (absent, low and high) were not significantly different. Regression analyses of nest success versus predator density (stoat or rat tracking indicies) were also not significant. A test of change in bellbird numbers (numbers in February/numbers in November in the same breeding season) versus rat density (tracking index in November) showed no significant trend at the RNRP. A test of density dependence (change in bellbird numbers versus numbers in November) was also not significant. The percent of time bellbirds spent feeding was significantly higher at the RNRP compared to Mt. Misery (Nelson Lakes, New Zealand) during the breeding season only (RNRP = 34 % and Mt. Misery = 23 %). The percent of time bellbirds spent foraging for food was also significantly higher at the RNRP compared to Mt. Misery during the breeding seasons (RNRP = 47 % and Mt. Misery = 36 %). This suggests that predator control at the RNRP has allowed bellbird numbers to increase so that nesting success may be partly limited by food supply during the breeding season. However, despite higher densities of bellbird and other bird species at the RNRP, there was no difference in the percent of time spent feeding (RNRP = 36.7 % and Mt. Misery = 36.3 %) or foraging (RNRP = 49 % and Mt. Misery = 51 %) between the RNRP and Mt. Misery during winter months The nest height of bellbirds was not significantly related to the nest outcome (excluding nests that failed due to non-predation causes) at either the RNRP or Kowhai Bush. Nest predation occurred at a range of heights, which suggests introduced predators are not specialists at any nest height. Bellbirds at the RNRP 111 visited their nests more frequently and stayed for shorter periods at the nest compared with bellbirds at the Cheeseman Valley, consistent with the hypothesis bellbirds alter their behavior to avoid disclosing their nest position. The results of this study, although not significant, do indicate that bellbird nest success is limited by high densities of predators. One-way anova analysis on nest success during the feeding stage was nearly significant (P = 0.096). Regression analysis trend lines show an increased negative effect on nest success at higher predator densities. Few replicates and large variation in nesting success within treatments affect the power of the results. Bellbirds may persist in areas with high predator densities due to anti-predator strategies when visiting their nests. With pest control, ecosystem interactions may be revitalized so that limiting factors other than predation become important for bellbird nest success.

Nest success, feeding behavior and anti-predator strategies, were studied to determine if bellbirds (Anthornis melanura) are limited by predation and if pest control can relieve the pressures of predation so that food supply may become a limiting factor for bellbirds. The nest success of bellbirds at the Rotoiti Nature Recovery Project (RNRP), Nelson Lakes, New Zealand (2002-2004 breeding seasons) was compared with data from other bellbird nest success studies in New Zealand. Two years of nest monitoring resulted in a low overall nest success at the RNRP (39 %), an area low in predator density, compared with 16.4 - 67.2 % found for other sites. One-way anova comparisons between the nest success of bellbirds in areas grouped by predator density (absent, low and high) were not significantly different. Regression analyses of nest success versus predator density (stoat or rat tracking indicies) were also not significant. A test of change in bellbird numbers (numbers in February/numbers in November in the same breeding season) versus rat density (tracking index in November) showed no significant trend at the RNRP. A test of density dependence (change in bellbird numbers versus numbers in November) was also not significant. The percent of time bellbirds spent feeding was significantly higher at the RNRP compared to Mt. Misery (Nelson Lakes, New Zealand) during the breeding season only (RNRP = 34 % and Mt. Misery = 23 %). The percent of time bellbirds spent foraging for food was also significantly higher at the RNRP compared to Mt. Misery during the breeding seasons (RNRP = 47 % and Mt. Misery = 36 %). This suggests that predator control at the RNRP has allowed bellbird numbers to increase so that nesting success may be partly limited by food supply during the breeding season. However, despite higher densities of bellbird and other bird species at the RNRP, there was no difference in the percent of time spent feeding (RNRP = 36.7 % and Mt. Misery = 36.3 %) or foraging (RNRP = 49 % and Mt. Misery = 51 %) between the RNRP and Mt. Misery during winter months The nest height of bellbirds was not significantly related to the nest outcome (excluding nests that failed due to non-predation causes) at either the RNRP or Kowhai Bush. Nest predation occurred at a range of heights, which suggests introduced predators are not specialists at any nest height. Bellbirds at the RNRP 111 visited their nests more frequently and stayed for shorter periods at the nest compared with bellbirds at the Cheeseman Valley, consistent with the hypothesis bellbirds alter their behavior to avoid disclosing their nest position. The results of this study, although not significant, do indicate that bellbird nest success is limited by high densities of predators. One-way anova analysis on nest success during the feeding stage was nearly significant (P = 0.096). Regression analysis trend lines show an increased negative effect on nest success at higher predator densities. Few replicates and large variation in nesting success within treatments affect the power of the results. Bellbirds may persist in areas with high predator densities due to anti-predator strategies when visiting their nests. With pest control, ecosystem interactions may be revitalized so that limiting factors other than predation become important for bellbird nest success.

Widespread anthropogenic invasions have prompted concerns that naturalized organisms could threaten biodiversity. In particular, invasive weeds can negatively affect native biota through a variety of means, including disrupting mutualisms. This thesis was designed to observe and test dispersal mutualisms in a native forest during autumn when the majority of plant species are fruiting. In this thesis I examined whether the invasive plant barberry (Berberis glaucocarpa) was influencing the behaviour of a native frugivore bellbird (Anthornis melanura) and a range of dispersal related services in a native forest, Kowhai Bush near Kaikoura. To test these 18 banded bellbirds were followed through autumn 2011. These observe bellbirds were split between control and test bird. Barberry fruit was removed from the test bird territories. I recorded whether bellbirds changed their territory sizes, foraging and daily behaviours. During 52 hours of observations, bellbirds were never observed feeding on barberry fruit. No significant changes to bellbird behaviour or territories were observed after the removal of barberry fruit. Bellbird diet overall was dominated by invertebrates (83% of foraging observations), with smaller contributions from fruit (16%, nearly all on Coprosma robusta), nectar and honeydew. Since bellbirds did not eat barberry fruit, removal of this weed is unlikely to negatively affect bellbirds during autumn. Which other bird species were dispersing barberry was recorded. I recorded 242 hours of videotape footage on 24 fruiting plants. A total of 101 foraging events were recorded of 4 different bird species: silvereyes (Zosterops lateralis) 42 visits, blackbirds (Turdus merula) 27 visits, song thrush (Turdus philomelos) 29, and starlings (Sturnus vulgaris) 3 visits. The species differed in the mean length of time they spent in plants, so the overall contribution to barberry fruit removal was 32.6% silvereyes, 24.3% blackbirds, 42.9% song thrush and 0.1% starlings. To find out the relative contribution of exotic and native birds to dispersal of fruits in Kowhai Bush, I mist-netted 221 birds of 10 species and identified any seeds in the 183 faeces they deposited. A total of 21 plant species were observed fruiting in Kowhai Bush during this time. A total of 11 different plant species were identified from 1092 seeds. Birds were further observed feeding on 3 other plant species which were not observed in faecal samples. This left 7 plants with unobserved dispersal vectors. There were likely four main dispersers, bellbirds, silvereyes, song thrush and blackbirds and five minor, brown creeper (Mohoua novaeseelandiae), tui (Prosthemadera novaeseelandiae), fantails (Rhipidura fuliginosa), dunnock (Prunella modularis) and starlings. However there was considerable variability between these bird species dispersal abilities. Introduced birds’ song thrush and blackbirds were observed dispersing naturalized plant seeds at higher than expected rates in comparison to native frugivores bellbirds and silvereyes. I also measured the gape sizes on mist netted birds and on samples of fruit from Kowhai Bush. Both silvereyes and bellbirds were found to be eating fruit larger than their gape, but despite this two native (Hedycarya arborea and Ripogonum scandens) and three exotic plants (Vitis vinifera, Taxus baccata and Crataegus monogyna) had large fruit that were probably mainly dispersed by song thrush and blackbirds. Hence, introduced birds were important seed dispersers for large fleshy fruited seeds in Kowhai Bush. Demonstrating that interactions among native and exotic flesh fruited plants and frugivores is important within forest communities.

Bellbirds (Anthornis melanura) are one of two endemic honeyeaters in New Zealand. They are still common in many parts of the country, but also rare and absent in some areas where they were abundant before. Together with tui (Prosthemadera novaeseelandia) bellbirds are responsible for avian pollination of several native plants in New Zealand. Habitat fragmentation and geographical constraints prevent bellbirds from re-occupying formerly inhabited areas. After translocations birds often disperse; some just disappear, some are thought to fly back to their source site and probably many die. Supplementary feeding provides energy rich food, which helps to ensure birds are well-nourished until they can discover natural food sources at the release sites. Conspecific song playback as an attractant for many bird species can be used as public information about high habitat value and additionally lure birds to good food sources and nest sites. Artificial feeding stations were set up at 18 locations in Kennedy's Bush, Port Hills, Canterbury, New Zealand. Feeders filled with sugar water were available for five to six consecutive days and during observation hours the effect of song playback was tested. Variation in bird numbers resulting from experimental treatments was recorded using five-minute bird counts prior to treatments and additional counts every minute of observation time. Bellbird numbers varied strongly with habitat types, showing highest numbers for dense and tall bush areas and close to creeks and lowest for open areas besides pasture and in post-flowering flax fields. Bellbird numbers increased when song was broadcast, with the strongest effect in areas where bellbird numbers were moderate or low, but not in areas where bellbirds were absent. Feeding stations were not observed to be used by bellbirds. The strong response in bellbird numbers to conspecific song playback proved the attractiveness of song for bellbirds. Nevertheless, its function as a lure for bellbirds to artificial feeding stations was not shown. This contrasting result indicates that the use of bellbird song as an enticement for bellbirds has to be investigated further.