八色鳥(Pitta nympha)的雛鳥食性、巢位選擇及繁殖成功率

Abstract

Tropical and subtropical Asia is currently confronting particularly severe forest conversion pressure. The vulnerable Fairy Pitta (Pitta nympha) is an understudied passerine which breeds in the low-elevation forests of eastern Asia and probably winters on Borneo. Because of severe habitat loss, suitable habitat within its breeding and wintering ranges as well as on its migration routes might not be sufficient to maintain its population. In addition, there are no published intensive studies on any migrating pitta species. Thus, there is an urgent need to establish knowledge of all aspects of its ecology and biology and then to propose suitable management measures to conserve this threatened species. In this dissertation, I firstly propose an effective method using playback to census this elusive species during its early breeding season. Followed by studies on its nestling diet, nest site selection and nesting success in west-central Taiwan. The effectiveness of using playback of recorded calls to census the presence and abundance of the Fairy Pitta was evaluated from late April to May 2001. The response of Fairy Pittas to playback was rapid; with 94% occurring in the first 5 min. Playback substantially increases the likelihood of detecting Fairy Pittas. Both the number of stations at which pittas were detected and the number of pittas detected per station increased with the use of playback. In addition, both the number of stations that detected pittas and the number of pittas detected per station were significantly higher in early than mid- and late May, but did not differ at different times of the day (early morning, noon, and afternoon). Our results indicate that playback considerably increased detectability, and 5 min of playback at a station is sufficient. However, it is important to census the Fairy Pitta population during the pre- to early-nesting period, and this should be carried out at least twice to ensure a high certainty of determining the presence or absence of Fairy Pittas at a station. The nestling diet of Fairy Pitta was studied by videotaping eight broods from 2000 to 2002. Adults usually brought 1-3 items at each feeding visit to nestlings, and the prey sizes were usually 2-10 cm in length. The total number of food provisionings to provide food for nestlings was estimated to be around 900-1000 based on a brood observed for the entire nestling period. Earthworms were the most important food item for all broods and occurred in more than 73% of the adult’s feeding visits. However, the occurrences of earthworms as food decreased with nestling age in a brood which was observed during drought conditions, possibly due to a decline in earthworm abundances during this dry period. Nest site selection and nesting success of Fairy Pitta were studied in a hilly area predominantly planted with bamboo (Dentrocalamus latiflorus) over three breeding seasons from 2004 to 2006. I and field assistants tracked success of all nests, quantified nest outcomes and determined possible causes of nest failures. Using binary logistic regression, I compared nesting habitats between each nest site and paired random location (n = 65 pairs), and between successful (i.e., a nest that successfully fledged at least one offspring, n = 32) and depredated nests (n = 32). Akaike’s Information Criterion (AICc) and model averaging were used to make inference about the weighted support for the importance of individual habitat variables through comparison of sets of competing models. Fairy Pitta chose to nest in areas with increased canopy cover (relative importance, RI = 0.99) and a diversity of trees (RI = 0.87) coupled with decreased shrub density immediately around the nest (RI = 0.75) and moderate small shrub density near the nest (RI = 0.79) and situated at a moderate distance from the nearest tree (RI = 0.57). Nests built in places that were closer to or farther form the base of the nearest tree (RI = 0.74), with higher overhead cover (RI = 0.63), with a steeper downward slope (RI = 0.48) and with greater difficulty to access (RI = 0.35) were more successful. Thus, of the five habitat characteristics that were associated with nest site selection, only one was related to nesting success with an inverse relationship. Although most habitat characteristics preferred did not positively relate to nest success, Fairy Pitta rarely located their nests in areas without these characteristics. Also, although bamboo density was not included in the best nest site selection models, it did significantly negatively correlate with the diversity of trees. Therefore, I recommend that managers protect remaining broadleaf forests and restore similar habitat in areas of rough terrain, usually found along valleys, to enlarge areas of suitable habitat.