運用臺北市大崙尾山步道公民科學資料探討氣候變化對植物物候之影響

面對全球氣候變遷議題，物候學已成為許多生態學家及環境科學家所投入之重要研究領域，然而國內鮮有針對區域尺度下植物群落之長期物候監測，尤其是低海拔都市近郊森林步道，對於植物群落如何適應外在氣候變化反映於物候上，仍所知有限。本研究運用由荒野保護協會志工累積之公民科學資料─大崙尾山步道十年植物物候觀察紀錄，並分別透過群落和物種兩種層級之分析，探討氣候變化對大崙尾山步道植物物候之影響。群落層級研究之物候型態分析結果顯示：大崙尾山步道之開花物種數高峰落在3至5月，果實成熟物種數高峰則落在9至11月。在排除僅有單一物種紀錄的科別後發現兩觀察區多數科別開花期和果熟期較為集中，多數相同分類群的植物有相近繁殖物候型態。稜線區整體平均開花期（五月下旬)早於步道區（六月中旬），果熟期則較晚（八月上旬和七月下旬），花芽期至果熟期平均歷期，稜線區（75天）長於步道區（63天），比較兩區落葉樹種生長物候型態發現，稜線區開始萌芽展葉或落葉月份平均早了步道區約1個月，上述差異成因可能與微環境、物種組成、種內個體差異及物候紀錄時間範圍不同等因素有關。交互相關性分析結果顯示：累積降水量與太陽輻射量之間具有顯著負相關性，推測兩者可能共同影響開花植物群落之繁殖物候。過去一段時間之區域性氣候波動（如：ENSO事件），不僅是驅動大崙尾山當下氣候變化（尤其氣溫）的重要因素，也可能對地區尺度下植物群落繁殖物候造成直接或間接地影響，然而不同尺度氣候因子對兩區繁殖物候指標的影響程度及時滯範圍並不完全一致。物種層級研究結果顯示：大崙尾山瀕危植物─白果雞屎樹（Lasianthus chinensis）族群，受2016年1月霜凍事件影響，該年主要繁殖物候─開花期高峰消失，隨後秋、冬季分別出現兩波異常開花期高峰，然而後者幾乎未結果，推測與蟲媒不同步有關。隨著全球暖化趨勢，未來30年其主要繁殖物候期將可能逐漸提前1至2個月，花芽和開花期高峰平均歷程將略為延長、結果和果熟期則略為縮短，族群整體開花至果實成熟期峰值更將顯著減少，種實來源降低情況下，恐影響族群新個體補充，因應未來暖化趨勢與極端天氣，對其族群存續之負面效益可能大於正面。根據不同的研究問題與目的，本研究透過不同的分析方法，了解氣候變化對大崙尾山步道植群及白果雞屎樹繁殖物候之影響，研究成果有助於拓展我國對於本地低海拔森林植物於物候研究領域之科學認知，亦可提供長期物候觀察紀錄應用於植物適應氣候變化之參考。為進一步落實物種保育，期許未來能持續針對國內受不同威脅程度或保育需求迫切之物種，增加觀察個體與頻率，及早擬定植物物候監測計畫，以利生物多樣性保育工作之推展與實現。

關鍵字

植物物候觀察；公民科學；生物多樣性保育；氣候變化；瀕危物種；白果雞屎樹

並列摘要

Facing the issues of global climate change, “phenology” has become an important research field attracting many ecologists and environmental scientists. However, long-term regional phenology monitoring records of plant communities are quite few in Taiwan, especially for those sub-urban forest trails in low-altitude areas, that limits our understanding towards the plant pheonological shifts reflecting the adaptation to climate change. Based on citizen science, the decadal observation records of plant phenology in the Dalunweishan hiking trail conducted by SOW(society of wilderness) volunteers, this study aims to explore the effects of climate variation on plant phenology at the community-level and the species-level. Based on the long-term statistical analysis results, the peaks of the flowering species was found to occur from March to May, while those of the fruit-ripen species from September to November. Results also showed that most families (excluding one with only a single species) appeared similar flowering or fruit-ripen patterns in the two observation areas. Species belonging to the same taxon tended to display similar phenological patterns. The main flowering time of the plant community in the crest area was earlier than that in the trail area (late May compared to mid-June), and the main fruit-ripen time is later (early August compare to late July). The duration of the reproduction (from flower bud to fruit-ripen) of the plant community in the crest area was longer than that in the trail area (75 days compared to 63 days). Comparing the growth phenology patterns of the deciduous trees in the two areas, results indicated that the species in the crest area began to sprout or defoliate 1 month earlier than those in the trail area.The aforementioned differences between the two areas may be related to the discrepencies associated with microenvironment, species composition, intraspecies differences, and temporal differences of phenological records. Results of the cross-correlation analysis showed that there was a significant negative correlation between precipitation and solar radiation in Dalunwershan, which in turn jointly affected the reproduction phenology of flowering plant communities. Earlier regional climate fluctuations (e.g., ENSO events) were suggested to be important factors driving the current local climate variation (e.g., temperature), as well as directly or indirectly affected the reproduction phenology of the plant communities in the Dalunweishan hiking trail. Yet the extent and time lags to which different climatic factors could affect the reproductive phenolgy in the two regions were not completely consistent. At the species level, based on the results simulated from the system dynamics model, the reproductive phenology of the endangered species of Lasianthus chinensis will be affected under different climate change scenarios for the next 30 years, of which the stages of budding and flowering may be in average advanced by 1 to 2 months, stages of fruiting and ripening may be shifted in average one month ahead, and peaks of each reproductive stage might become less obvious, with slightly prolonged duration of budding and flowering stages, and shortened duration of fruiting and ripening stages. Depending on the research questions and the community- or species-levels, this research employed different analysis methods to explore the effect of climate variaton on phenology of plant community in the Dalunweishan hiking trail and the long-term reproductive risks of Lasianthus chinensis. Resesarch results can strengthen our understanding of the phonology for the local low-elevation forest plant communities, and provide references to the adaptation of the plant communities based on the long-term observation records. To further ascertain the species conservation, it is recommended to increase the number of observation individuals or the recording frequency of phenological observations for species under different threats or with potential for conservation value across various regions in Taiwan, and to develop plant phenological monitoring plan in time to carry and achieve the vision of biodiversity conservation.