以網路型式量化食物網中物種的獨特性與功能多樣性

Abstract

在氣候劇烈變遷時代，生物多樣性的喪失儼然是一個迫切關注的議題，而如何量化生態系中物種的重要性和多樣性是生態研究中的兩個緊迫問題。由於物種的獵補關係而形成食物網，因此引發我們想從拓撲網路學角度量化物種重要性和生物多樣性。傳統上，物種網路的重要性即考慮了物種對整個食物網的影響以及其所處位置的中心性；近來，另一個物種重要性的評斷概念也逐漸被重視，那就是物種獨特性。在本論文中，我們提出了一個直覺的物種獨特性測量方法，以矩陣方式量化了食物網中物種彼此間的相互影響，進而計算物種間的距離與獨特性。本論文所產生的結果與過往研究方法所產生的結果近乎相同；然而避免重複使用計算的資訊、所耗費的計算時間也更少。此外，本論文也提供了一個基本框架，可用於接受不同的網路特質和距離度量以量化物種的獨特性。 生態系統的生物多樣性可以通過多種方式量化，其中之一即為「功能多樣性」，其可量化生態系統中物種特徵的異質性，此處「功能多樣性」包括物種特徵的豐富性,多樣性與差異性。本論文中，提出了三種不同類型且基於網路基礎的「功能多樣性」測量方式。第一種是基於食物網的物種相互作用結構、第二種是基於各種拓撲中心性指數、第三種類型是基於物種在食物網中的獵食關係。此處也研究了基於網路的功能多樣性與食物網的網路屬性之間的關係，以及它們與傳統的基於生態特徵的功能多樣性指數的關係。本論文分析表明，連接性稀疏的食物網和具有高度模塊化結構的食物網往往具有高度基於網絡的功能多樣性。此外，基於網絡的功能多樣性指數與其傳統的基於特徵的對應物之間的適度相關性表明，我們的方法提供了生態系統功能多樣性的補充視圖。

The loss of biodiversity is a major concern in the era of global warming and climate change. How to quantify species’ importance and biodiversity of an ecosystem are two pressing issues in ecological research. Since species interact trophically forming a food web, it is nature to quantify species importance and biodiversity from a network perspective. Traditionally, the network perspective of species importance considers the effect of a species on the whole food and the centrality of species’ network position. Recently the concept of species uniqueness has been suggested as an alternative view on species importance. In this study, we propose a simple species uniqueness measurement. Our approach quantifies the effects between species, which constitute the interaction structure of a food web. Rows of such an interaction matrix are compared to compute distances between species, which are then used to calculate uniqueness values of species in a food web. Our approach produces results almost identical to that from a previous approach; however, ours requires less information and therefore requires much shorter computation time. Our approach also provides a basic framework for quantifying species uniqueness using different network-related information and distance measures. Biodiversity of an ecosystem can be quantified in various ways. One of them, functional diversity, quantifies the heterogeneity in species traits in an ecosystem. Since species’ network positions in a food web reflect their functional roles, we argue functional diversity of an ecosystem can also be measured from a network perspective. In this study, we propose three different types of network-based functional diversity measurement. The first type is based on the interaction structure of a food web, and the functional diversity of a food web is the average dissimilarity between species’ interaction profiles. The second type is based on various centrality indices. Here, different centrality indices are applied to quantify the network position of species; and the functional diversity of a food web is quantified by several properties of species distribution in a multi-dimensional centrality trait space. The third type is based on the trophic role of species in a food web. Functional diversity here includes average trophic role dissimilarity between species, the number of trophic role groups, and how evenly species are partitioned into different trophic roles. Furthermore, we investigate the relationship between network-based functional diversity and several network properties of a food web, as well as their relationship with conventional trait-based functional diversity indices. Our analysis suggests that sparsely connected food webs and those with highly modular structures tend to have high network-based functional diversity. Also, the moderate correlation between network-based functional diversity indices and their conventional trait-based counterparts suggests that our approach provides a complementary view of an ecosystem’s functional diversity.