珊瑚適應策略中整合種內變異性之研究

Abstract

種內變異性 (Intraspecific variability) 是天擇和物種適應的必要條件，因此它是物種生存的關鍵要素。目前，前人提出了四種珊瑚的適應策略 (Adaptive strategy, i.e., competitive, weedy, stress-tolerant and generalist)， 但是這些策略都是利用珊瑚的性狀的平均值所定義的， 而完全沒有考慮珊瑚的性狀會隨著環境變化而改變。本研究旨在評估不同珊瑚種之間的種內變異性以及它們的來源，希望藉此檢驗將種內變異性整合進珊瑚的適應策略的可行性。在本研究中，我們選擇了四個對比的環境（台灣北部的淺水區， 綠島的淺水、深水區和潮池）和兩個對比的季節（夏天和冬天）進行採樣， 共採集了七種珊瑚，並且測量這些物種的八個和能量獲取和分配有關的性狀(corallite width, tissue biomass, Symbiodium density, cnidocyte density, chlorophyll b and c contents, host and zooxanthellae protein contents)。接著，我用高爾矩陣計算了這些樣本的性狀相似性並且將結果用PCoA呈現在一個性狀表現空間（Performance trait space）。在這個空間內，每個物種的性狀表現都被呈現在一個具有一定位置和大小的物種表現生態棲位 (Species performance niche)。一方面， 我用物種表現生態棲位的中心點來評估它們是否佔據同一個位置，並且用PERMANOVA來檢驗。另一方面， 我用性狀表現豐度 (Performance trait richness) 和性狀表現散佈 (Performance trait dispersion) 來代表這些物種的種內變異性。接著，我把種內變異性的來源分成了兩部分，環境條件間可變性和環境條件內可變性 (Between and within environmental condition variabilities)。最後，根據樣本的性狀，我用k均值聚類把樣本分成了幾個大類。我總共獲得了235個珊瑚的樣本，在性狀表現空間內，結果顯示所有的物種都具有獨特的位置和大小。在這些物種中，性狀表現豐度和性狀表現分佈都表明Stylophora pistillata 和Psammocora profundacella分別具有最大和最小的種內變異性，而其他的種則處於中間。 在環境條件間可變性的方面，我發現了很多性狀在不同環境間具有顯著統計差異，並且不同物種的性狀可變性（flexibility）也不一樣。而在環境條件內可變性的方面，我也發現Psammocora profundacella 和 Porites lutea 比Stylophora pistillata、 Cyphastrea sp.、Galaxea fascicularis 和Isopora palifera 具有較低的可變性。最後，k均值聚類把來自七個種的樣本分了四大類，我建議將每一類稱作戰術 (Tactic)。我之後又將戰術加入了對廣生種 (Generalist) 和狹生種 (Specialist) 的定義中。這種簡單的將珊瑚的適應性策略一分為二的框架考慮了種內變異性，可能可以更真實地反映自然界中的珊瑚適應性策略。最後，本研究為朝珊瑚的種內變異性整合進它們適應性策略的定義邁出了第一步。

Intraspecific variability offers a ground for natural selection to operate and species to adapt. It represents a key factor for species survival. In corals, four adaptive strategies (i.e., competitive, weedy, stress-tolerant and generalist) were previously defined on the base of average trait values, which overlooked intraspecific variability in their response to different environmental conditions. Here, I assessed the intraspecific variability of coral species, explored its sources and examined the potential of integrating this variability into the definition of coral adaptive strategies. Eight physiological traits characterizing energy acquisition and allocation (i.e., corallite width, tissue biomass, Symbiodium density, cnidocyte density, chlorophyll b and c contents, host and zooxanthellae protein contents) were examined in seven coral species sampled in three contrasted habitats (i.e., shallow and deep of Green island, shallow of North of Taiwan and tidal pool) at two seasons (i.e., summer and winter). Individuals were compared for their trait similarity using Gower distance and visualized into a performance trait space using a Principal Coordinates Analysis (PCoA). Each species was characterized by the position and size of the species performance niche delineating all individuals of that species. The divergence of species performance niche was further examined by the position of their centroids and tested by permutational multivariate analysis of variance(PERMANOVA). The relative proportion of the species performance niche to the overall performance niche (i.e., species trait richness) and dispersion (i.e., species trait dispersion) were used to represent the intraspecific variability of that species. The sources of the intraspecific variability were then decomposed into between (BV) and within (WV) environmental condition variabilities. Eventually, k-means clustering was used to partition the individuals into k clusters based on their trait values. In total, we collected 235 samples and the result showed all the species exhibited different sizes of species performance niche and occupied different positions in our performance trait space. Among these species, both species trait richness and species trait dispersion showed Stylophora pistillata and Psammocora profundacella possessed the largest and smallest intraspecific variabilities, respectively, while other species resided in between. For BV, I detected extensive significant changes of traits across environmental conditions and species showed different traits flexibilities. For WV, Psammocora profundacella and Porites lutea showed significantly lower values than Stylophora pistillata, Cyphastrea sp., Galaxea fascicularis, and Isopora palifera. Lastly, all the individuals from the seven species were partitioned into four k-means clusters, which I proposed as tactics. I further add tactics into the definition of generalist and specialist, which may be a better framework of defining adaptive strategies with the consideration of intraspecific variability. Overall, this study constitutes a first step toward an integration of intraspecific variability into a definition of coral adaptive strategies.