臺灣狐蝠與八重山狐蝠之遺傳結構

Abstract

次族群之間反覆發生的拓殖及滅絕事件，形塑了一個物種的地理分布及遺傳結構，並反映了其生活史的特徵。當這樣關聯族群(metapopulation)之動態發生於人為干擾的環境並橫跨多個國界時，其各族群的成長趨勢及生存威脅往往不盡相同，故具備跨越國界的整合性保育思維及視野，便顯得至關重要，尤其是對於分布廣泛並多棲息於脆弱島嶼生態系的受脅物種，例如舊世界的大型果蝠──主要為狐蝠屬(Pteropus spp.)的物種成員。臺灣狐蝠 (Pteropus dasymallus formosus) 為臺灣特有亞種，原分布於綠島，惟因大量獵捕及棲地劣化而一度瀕臨滅絕。近年來，於臺灣境內的龜山島及本島的花蓮市區發現新建立之居留族群，數量穩定上升。本研究中，第一部分的目的即旨在建立及優化非侵入式 (食渣為主) 的採樣方法，以提升調查效率並減少對該保育類物種的干擾，亦評估該方法學應用於長期族群監測的可行性。隨後，我們使用微衛星及粒線體 (D-loop) 基因座，檢驗臺灣狐蝠與鄰近亞種—八重山狐蝠 (P. d. yayeyamae) 的遺傳多樣性、遺傳結構及近代的族群數量變遷，以釐清兩亞種的基因交流及分化狀況。 結果顯示，非侵入式樣本為一穩定且可靠的DNA監測來源，尤其葉渣和排遺的個體鑑定成功率可達至少五成，有效提升履及不易或罕見個體出沒之樣點調查效率。遺傳分析揭示了亞種之間顯著存在遺傳分化，惟臺灣與八重山族群仍為亞種成對比較中，最為接近的一對組合，其中以與那國島為最獨立的遺傳組成、龜山島則展現了次遺傳結構並擁有最多樣的祖系來源。檢視臺灣境內的基因交流，並無個體在三處居留族群中被重複鑑定，然而遺傳結構及單倍型網絡圖仍間接顯示了基因流的存在。族群動態分析並無觀察到亞種或島嶼存在瓶頸效應，而八重山特別可見族群擴張的跡象。考量臺灣族群中，明顯存在來自八重山島嶼不對等的基因流及潛在移入者，顯示現今臺灣主要族群的來源，很可能來自八重山亞種的近代移入個體。有鑑於臺灣及八重山的頻繁基因交流，目前於島嶼間觀測到的顯著分化應是來自近代的奠基者效應 (founder effect) ，而非長時間的演化隔離，故除了遺傳及族群動態明顯獨立的與那國島應視為獨立單元外，臺灣及八重山境內的其餘諸島，皆應整合為單一的保育單位進行監測管理。綜合而言，考量該物種的廣泛分布、跨洋及跨島播遷的高度潛力，保育策略應同時兼顧島嶼棲地及海岸廊道的串聯。

Recurring colonization and extinction events among subpopulations shape a species’ geographical range and genetic structure, and also reflect its life-history traits. A cross-border perspective is critical for conservation when metapopulation dynamics are influenced by human-altered environments or extend across national borders, where populations face differing trends and levels of threat—particularly for wide-ranging, threatened taxa inhabiting fragile insular ecosystems, such as large old world fruit bats, namely, flying foxes (primarily Pteropus spp.). The Formosan flying fox (Pteropus dasymallus formosus) is the endemic subspecies in Taiwan and one of the five subspecies of the Ryukyu flying fox. The population was once abundant on Lyudao (Green Island) but has nearly gone extinct due to intensive hunting and severe habitat loss. In contrast, the recently discovered populations on Gueishan Island (Turtle Island) and in Hualien City have been steadily increasing. In the first part of this study, to maximize monitoring efficiency while minimizing disturbance to this protected species, we developed and validated an optimized methodology using non-invasive samples, primarily fruit pellets, and evaluated its utility for long-term monitoring. We then employed microsatellite loci and a mitochondrial marker (D-loop) to examine the genetic diversity, genetic structure, and recent demographic changes of P. d. formosus across different populations, as well as its genetic differentiation from P. d. yayeyamae. Non-invasive samples proved to be a reliable and stable source of DNA, with a notably high individual identification success rate (nearly 50%) from leaf pellets and feces. This substantially increased sampling success, particularly at sites with rare sightings or limited accessibility. Significant differentiation was observed among subspecies, with the closest pairwise relationship found between the populations in Taiwan and Yaeyama. Notably, the population on Yonaguni-jima exhibited the most distinct genetic composition, in contrast to the Gueishan Island population, which displayed substructure and possessed the greatest lineage diversity. No individual was recorded in more than one of the three populations in Taiwan; however, the genetic structure and haplotype network still provided indirect evidence of gene flow. No genetic bottleneck was detected among subspecies and islands; however, a notable sign of population expansion revealed in Yaeyema. The presence of asymmetric gene flow and putative migrants from neighboring islands suggests that the newly discovered populations in Taiwan most likely originated from Yaeyama over past few generations. Considering the recurrent gene flow between Taiwan and Yaeyama, the significant genetic differentiation among the islands is probably the result of founder effects rather than long-term evolutionary isolation. Therefore, we suggest that most of the islands between Taiwan and Yaeyama should be treated as a single management unit, with the exception of the Yonaguni-jima, which exhibits clear genetic and demographic independence. Given the species' broad distribution and high potential for oceanic dispersal among islands, conservation strategies should prioritize the protection of insular and coastal corridor habitats.