印度-西太平洋帶魚之類緣地理及系群結構

Abstract

帶魚科(Trichiuridae)的物種具高經濟價值，是許多漁業的捕撈目標。 帶魚是兇猛的掠食魚類，在生態系統上扮演重要的角色。 本科魚類特化的外型，使本科的分類學統整相對比較不容易，過去關於本科內屬間的亞科的歸屬關係、帶魚屬內物種的分類狀況及最常見之日本帶魚系群結構都存在著一些爭議。 本論文利用傳統型態測量、幾何型態測量及粒線體DNA來作為工具，以協助釐清帶魚科之系統分類、確認帶魚屬的物種分類狀態與演化關係及釐清西北太平洋地區日本帶魚(T. japonicus)之族群結構。 利用粒線體DNA分析本科內9屬的物種結果顯示，隱足帶魚亞科(Aphanopodinae)及帶魚亞科(Trichiurinae)為單源系關係，但鱗足帶魚亞科(Lepidopodinae)則否。 若將共同有尾鰭消失及臀鰭退化隱於皮下特徵之小帶魚屬(Eupleurogrammus) 及狹顱帶魚屬(Tentoriceps)歸類到帶魚亞科當中，則更符合單源系原則。 另外棘背帶魚屬(Assurger)則與深海帶魚屬呈現相對小的間隙(gap)，因此建議將棘背帶魚屬合併到深海帶魚屬當中，棘背帶魚學名回歸最初命名--Assurger anzac Alexander, 1917。 沙帶魚屬(Lepturacanthus)亦與帶魚屬呈現緊密關係，且沙帶魚屬為旁係關係，因此亦建議合併到帶魚屬當中。 由外部型態分析及粒線體DNA分析的結果均認為，原先認為西北太平洋區域的白帶魚複合體(T. lepturus complex)，應區分為三個物種：白帶魚(T. lepturus)、日本帶魚(T. japonicus)及南海帶魚(T. nanhaiensis)。 利用傳統型態分析結果三物種在測量形質的比例上皆有顯著差異，但部分有重疊。 利用幾何型態分析抽取出形狀的參數(H)則可以區分出此3物種，貢獻度最大的前兩個變數分別代表可以代表體高及肛前背鰭基底長，利用全長對體高的比例及全長對肛前背鰭基底長的比例可以區分出此3物種。 粒線體DNA重建的類緣關係顯示，此3物種可以區分為3群，彼此間沒有鑲嵌的現象，分子變異數分析也顯示主要的變異均集中在物種之間(ΦCT = 0.967)。 帶魚屬物種類緣地理的分析結果顯示，所有的物種的種化事件大約都在8.5百萬年前中新世(Miocene)晚期發生，地點可能在今日中南半島、蘇門達臘及婆羅洲之間的區域。 日本帶魚族群結構分析顯示， 日本帶魚有高的單倍基因型多樣性(0.97 - 0.99)，但核酸多樣性不高(0.005 - 0.009)，顯示可能族群曾歷經瓶頸效應，之後族群再度擴張。 分子變異數分析(AMOVA)認為將此區域族群劃分為泛-東海群(包含黃海、東海、及台灣海峽)及南海群時，分佈於群間的變異為最大(ΦCT = 0.165)，此結果顯示台灣海峽為日本帶魚的屏障，阻隔了東海及南海的基因交流。 然而從類緣關係分析發現東海及南海的分支當中有鑲嵌的現象，顯示過去南海仍透過台灣海峽與東海交換部分個體。 今日東海及南海之間的族群，主要因台灣海峽的季節流場變化及帶魚幼魚跨越海區界線的能力，所以無法有效交流。 根據溯祖理論推估，西北太平洋地區的日本帶魚族群分化，應始於更新世中期最大冰期事件，之後當冰層消融海水上升之後族群開始擴張而形成今日的族群。 本研究建議以兩個系群的管理模式，即東海系群與南海系群，來進行資源管理。

Abstract The trichiurid fishes commonly called hairtails are important resources for various types of fisheries. Hairtails, being situated on the top of trophic pyramid, also play important roles in the demersal eco-systems. Due to highly specialized external appearance, the taxonomic status of hairtails is ambiguous as compared to the other fishes ranging from subfamily down to population levels. In this study, we used traditional measurements, geometric morphometrics and mitochondrial DNA as tool to reveal the phylogenetic relationships among genera, the species status of the genus Trichiurus and the population structure of T. japonicus in the western North Pacific. The phylogenetic trees among 9 genera showed that both Aphanopodinae and Trichiurinae are monophyletic groups, but not for Lepidopodinae. We suggest putting Eupleurogrammus and Tentoriceps to Trichiurinae, because they all shared with apomorph of reduced anal and caudal fins. In addition, Assurger and Evoxymetopon showed so closed relationship that they should combine into a complete genus. Similarly, due to closed relationship between Lepturacanthus and Trichiurus and paraphyly of Lepturacanthus, we also suggest that they belong to a same genus. The results from morphological and mitochondrial analyses showed that there are three valid species contained in the 'T. lepturus' complex -- T. lepturus, T. japonicus and T. nanhaiensis. Traditional measurements showed that the ratio of measurements pairs were significant but overlapping. However, the shear (shape component, H) had fine resolution power to discriminate three species without overlapping. The first and second contributed of variables for the shear were represented body depth and preanal dorsal fin based length. Therefore, three species can be separated by using these two measurements compared to their total length. The phylogenetic relationships showed three distinct groups with no outliers. AMOVA showed that the major component of the variances concentrated among species (ΦCT = 0.967). The analyses of phylogeography of Trichiurus showed that all speciation events occurred around 8.5 million years ago in tropic waters among Indo-China, Sumatra and Borneo during the late Miocene. The high haplotype diversities (0.97-0.99) with moderate nucleotide diversities (0.005 - 0.009) might result from historical bottleneck and subsequently population expansion. Populations in the area were sub-structured into two groups of the SCS and pan-ECS (ECS+TS+YS), confirmed by AMOVA (ΦCT = 0.165). These results indicate that the TS served as a barrier, which interrupts mixture between populations in the ECS and SCS. However, intermittent gene flow were also traceable in the phylogenetic analyses, indicating that the SCS gained a small number of migrants from the TS. Limited larval dispersal ability across marine boundaries and monsoon-influenced flow patterns in the TS well explain a non-panmictic structuring. Coalescent theory estimated that the populations were subdivided during the middle Pleistocene glacial maxima, and expanded when the ice sheets retreated. Two management stocks are suggested for conservation purposes; i.e., the ECS (including the TS) and the SCS, to strengthen current fishery regulatory programs.