Taiwan’s Breeding Bird Survey reveals very few declining species

Abstract

The global target is to achieve Nature-Positive by 2030. However, tools and indicators for biodiversity conservation are overwhelmingly concentrated in Europe, with no relevant indicators previously available for East Asia. Taiwan, with its rich biodiversity and nationwide citizen science projects, offers ideal conditions for developing such indicators. This study used long-term monitoring data from the Taiwan Breeding Bird Survey, a nationwide systematic citizen science project, to determine population trends of 107 breeding bird species in Taiwan between 2011 and 2019. From these data, three multi-species indicators were established: the Taiwan Forest Bird Indicator, the Tawian Farmland Bird Indicator, and the Taiwan Introduced Bird Indicator. The results were published in 2023 in Ecological Indicators. Among the 107 breeding species, 17 showed significant increases, 88 remained stable, and 2 experienced significant declines. Six species were close to the significance threshold and require close monitoring. Both the Taiwan Forest Bird Indicator and the Taiwan Farmland Bird Indicator showed steady growth, indicating stable environmental conditions for these groups. In contrast, seven introduced species increased markedly, highlighting their potential threat. The methodology developed in this study can be applied to other taxonomic groups, ecosystems, and conservation issues, and has already been adopted in eight domestic and international projects. It can also serve as a standard national framework for evaluating conservation effectiveness — continuously accumulating records through citizen science, updating trends and indicators, and reviewing and adjusting conservation strategies. A diverse suite of indicators can contribute to the creation of a national biodiversity conservation dashboard, enabling progress towards the 2030 nature-positive growth target to be assessed.

Methodology and Implementation Steps

計算各穿越線各物種的平均個體數，以此數值建立各物種的時空矩陣。為避免調查樣點及調查時間本身的變異影響，族群變化趨勢以以平滑階層模型進行分析，同時也減少極端值對分析造成的影響。平滑階層模式是目前用來估算族群區是最理想的方法，唯樣區必須以分層逢機取樣或逢機取樣來決定，相較之下皆優於卜瓦松分布的廣義線性模型和廣義加成模型。廣義線性混合模型已廣泛用於估算各生物類群的族群趨勢，但是這樣的分析方式容易受到不同調查旅次本身狀況的影響，導致族群趨勢的信賴度降低，平滑階層模型則能克服這個問題。本計畫將以第一次調查為起始時間，指標值設定為100。MCMC運算將以三條起始值不同的運算鍊進行 20,000 次的分析。R-hat 值將用來確認收斂狀況。趨勢結果將呈現中位數、2.5百分位及97.5百分位。複合物種指標方面，將選定的鳥種的族群變化趨勢，以幾何平均方式建立指標。族群變動值將從 7,500個MCMC孤寂值中重複隨機取 10,000 次，用來估算複合物種指標的信賴區間。同樣以第一次調查為起始時間，指標值設定為100。趨勢結果將呈現中位數、2.5百分位及97.5百分位。如果2.5百分位及97.5百分位之間不包含起始值100者視為顯著，包含則視為不顯著。

Innovation and Cross-Disciplinary Collaboration

1. 使臺灣成為東亞首次建立複合物種指標的國家。
2. 臺灣首次以公民科學資料建立指標，提供新的建置標準作業程序，大幅提升公民科學資料的應用價值。
3. 可應用於各生物類群、生態系及相關保育議題，目前已有7項國內外計畫效法應用。
4. 可作為國家保育成效的標準檢視流程，持續以公民科學累積紀錄、更新趨勢及指標、檢視調整保育成效與策略。多元的指標可建構國家生物多樣性保育儀表板，於2030年檢視是否實現自然正成長。

Expected Results and Contributions

1. 強調數量普遍的麻雀和白頭翁、棲息於草生地的粉紅鸚嘴、農地的掠食者大卷尾和棕背伯勞、棲息於溪流的鉛色水鶇等鳥種的族群接近「顯著減少」的門檻，彰顯過去未曾受關注的保育鳥種。
2. 強調外來鳥種擴張會是未來台灣重要的鳥類保育議題，需加強外來種管理。
3. 「2024臺灣鳥類紅皮書名錄」即以鳥類族群變化趨勢為依據，評估各鳥種的受脅程度，再進一步評估是否列為法定保育類野生動物。
4. 「族群快速下降」之野生動物，明列為慈心基金會「綠色保育標章」的保育對象。
5. 將鳥類族群趨勢及指標更新至2021年，併於「2024臺灣國家鳥類報告：摘要報告(中英文版)」報導臺灣繁殖鳥類的族群的最新變化趨勢。