Please use this identifier to cite or link to this item:
標題: 蘭嶼紅頭溪水生及半水生昆蟲多樣性與昆蟲多樣性研究之評論
Aquatic and Semiaquatic Insect Diversity of Hontou Stream in Lanyu with Critiques on Insect Diversity Researches
作者: 范姜俊承
Fanjiang, Chun-Cheng
關鍵字: 生物多樣性;biodiversity;水生昆蟲;多樣性指數;對數序列;對數常態分布;科級生物指標;評論;aquatic insect;diversity index;logarithmic series;lognormal distribution;rarefaction;Critiques
出版社: 昆蟲學系
本研究於2003年在蘭嶼紅頭溪進行水生及半水生昆蟲多樣性調查。調查結果利用Shannon-Wiener index、Simpson’s index、logarithmic series、lognormal distribution、rarefaction分析並評論之。紅頭溪依棲地類型分為森林區、草原區、半水泥區、水泥區及河口區。共採得水生昆蟲7目27科47形態種,以森林區物種數最多;半水生昆蟲僅有闊水黽科、微水黽科各1種,兩者皆分布於森林區、草原區及半水泥區。水生昆蟲資料分析結果顯示Shannon-Wiener指數方面以草原區為最高,Simpson指數則以森林區具有較高之多樣性。除半水泥區外,其餘各區皆符合logarithmic series,其中以森林區有最高的多樣性指數α,然而五區中僅半水泥區符合lognormal distribution。Rarefaction分析結果以森林區為最高,主要由於森林區有最多的物種數。由於各分析方法所得之結果不一致,造成解讀昆蟲多樣性之困擾或各唱其調。各棲地之科級生物指標(FBI)介於4.49至6.86之間;除水泥區之各季節均相似外,各棲地FBI皆隨季節而變化,其變化原因主要為水生昆蟲之族群變動。多樣性分析目前尚有幾個問題:(1) 一般未能依據正確的取樣方法估計各物種之相對豐度(pi);(2)物種調查通常無法涵蓋所有的物種數(S);(3)不同採集方法所採得各物種之機率並不相等;(4)昆蟲的族群密度與齡期結構隨季節而變動,且不同齡期之昆蟲被採得的機率不同。因為多數的昆蟲屬於r-strategists,其pi並非固定,因此昆蟲之族群動態的差異會影響pi的正確性。這些問題顯示,僅依靠簡單的指數或機率統計並無法正確的分析、描述與解釋昆蟲多樣性。因此,昆蟲多樣性的研究除基礎之物種調查外,應重視以生態學的方法來研究影響物種組成及相對豐度變動的原因,才能進而對生物多樣性的研究與永續維持有具體的貢獻。

The semiaquatic and aquatic insect fauna of Hontou Stream in Lanyu were surveyed in 2003. The data were analyzed by using Shannon-Wiener index, Simpson's index, logarithmic series, lognormal distribution and rarefaction. Moreover, we criticized the methods of analysis. Five habitats were defined along Hontou Stream based on environmental characteristics: forest, grassland, cemented dike, cemented aqueduct and estuary areas. We collected 22,057 individuals of aquatic insects belonging to 7 orders, 27 family and 47 morphological species. The forest area has the richest number of species among all habitats. Two species of semiaquatic insects, each belong to Veliidae and Hebridae, were found in forest, grassland and cemented dike areas. In the result of the analysis base on aquatic insects, the highest Shannon-Wiener index was obtained for grassland area, while the highest Simpson's index was obtained for the forest area. All habitats fit logarithmic series except cemented dike area and the forest area has the highest α index. Only the data from the cemented dike area fitted the lognormal distribution by using chi-square test. Rarefaction curve for forest area was the highest due to its richness in the number of species. The fact that various analyzing methods lead to different results brings problems in interpreting the data of insect biodiversity which may come to very different conclusions. The family biotic index (FBI) for different habitats ranged from 4.49 to 6.86. The values of FBI changed seasonally in all habitats except cemented aqueduct. The fluctuation of FBI was mainly influenced by the population dynamics of aquatic insect in the stream. Some major problems of biodiversity studies should be aware: (1) In most biodiversity surveys, it was often failed to use a proper sampling method to estimate the relative abundance of each species ( ); (2) The total number of species (S) is often hard to be obtained in a survey; (3) The probability of catching each species is not equal, which is also affected by different collecting method; (4) Insects population density and age structure fluctuate with seasons and the probabilities of catching insects in different developmental stages are also variable. Because most insects are r-strategists, there is no fixed pi for each species. Therefore, it is inappropriate to calculate any diversity index based on pi, and simple indices and probabilistic statistics cannot analyse, describe, and interpret insect diversity meaningfully. In conjunction with the basic taxonomic survey, we propose to study the change of species composition and fluctuation of relative abundances based on ecological methods and food chain structure. Only then could a solid contribution to the biodiversity research and its sustainability be expected.
Appears in Collections:昆蟲學系

Show full item record
TAIR Related Article

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.