台灣入侵紅火蟻不同社會型之探討

Abstract

入侵紅火蟻 (Solenopsis invicta Buren) 於2003 年10月在台灣被發現，牠的群落 (colony) 具有兩種社會型：單蟻后 (monogyne, M) 與多蟻后 (polygyne, P)；而不同的社會型在生態、生理、行為甚至遺傳方面都有相當顯著的差異。為了對其在台灣本土的擴散以及生態地位所扮演的角色深入了解，甚至探討其入侵源，本研究重點為調查這兩種社會型的入侵紅火蟻在台灣的分佈情形，並比較分析其表皮化學物質的差異。調查的地區主要為危害較嚴重的縣市，包括台北縣、桃園縣和嘉義縣，採集群落數分別為 130、66、37個，共計233群落；我們先22個群落內具有生殖能力的蟻后數目來決定其社會型，並與 Gp-9 這個分子指標相互映證，顯示分子指標的可信度很高；且根據分子指標判定其餘所有群落的社會型。結果顯示在這三個縣市兩種社會型的分佈狀況呈現鑲嵌混合分佈的形式，然而多蟻后社會型的群落在大部分的地區佔了相當高的比例，特別是桃園縣及嘉義縣，以嘉義縣為例，在37個檢測的群落中只有4個單蟻后社會型的群落；而在台北縣的林口鄉則顯示出單蟻后社會型居多的相反情形，在86個檢測群落中只有9個多蟻后社會型的群落，這種現象顯示出事實上兩種社會型並不是以均勻的比例共存於每一地區，而是會以其中一種社會型會占有壓倒性的多數。同時利用氣相層析儀 (gas chromatography) 與氣相層析質譜儀 (gas chromatography-mass spectrometry) 對不同社會型的入侵紅火蟻的表皮化學物質作分析，分別比較蟻后和工蟻兩種階級在不同社會型間的差異；結果顯示在烷類所構成碳氫化合物質 (hydrocarbons) 與主要由生物鹼 (alkaloids) 所構成的非碳氫化合物質均發現在兩種不同的社會型之間，在量上具有統計上的顯著差異；另外在質上我們也發現單蟻后社會型的工蟻比多蟻后社會型的工蟻更多了一種生物鹼的化學物質；進一步使用主成份分析法 (principal component analysis) 和判別分析法 (discriminant analysis) 對於這些在兩種社會型之間有顯著差異的表皮化學物質進行排序 (ordination) 分析，雖然結果並不能顯著分群 (clustering)，但是都顯示來自於單蟻后社會型的蟻后群聚較緊密，而來自多蟻后社會型的蟻后則明顯地較鬆散。由於這些化學物質的合成受到遺傳影響，這種現象顯示兩個涵義，第一個是單蟻后社會型的工蟻對其他蟻后有強烈的排他性而造成的選汰壓力導致其遺傳結構上較為均質，另一個涵義顯示單蟻后社會型的遺傳組成可藉由雄蟻將其基因帶入多蟻后社會型的遺傳組成，但反之則不然；因此二者之間無法分群，且後者的遺傳變異較前者為大。至於牠們用以辨認的究竟是哪幾種成分，尚待進一步的研究。

Solenopsis invicta Buren, the red imported fire ant (RIFA), was first found in Taiwan in October 2003. It has two social forms, monogyne in which a colony has a single queen, and polygyne in which a colony has multiple egg-laying queens. There are significant differences in ecology, physiology, genetics, and social behaviors between monogynous and polygynous colonies. The first part of this study is to focus on the distribution of the two social forms for realizing the dispersal and occupied niche in Taiwan, and the second part is to compare cuticular chemicals of the two social forms in the fire ant. The investigated areas were seriously infested counties, including Taipei, Taoyuan, and Chiayi Counties; the sampled colonies were 130, 66, and 37, respectively, that amounted to 233 colonies in total. Initially, the social forms of 22 colonies were determined by the number of egg-laying queens, and then confirmed by a molecular marker using specific primers of a gene Gp-9. The accuracy of molecular marker examined by using these two methods was reliable in Taiwanese colonies, and thus the other collected colonies were determined by this Gp-9 marker. Our result showed that the two social forms distributed sympatrically, nevertheless they are not evenly distributed. In most areas, polygynous colonies were more than monogynous, particularly in Taoyuan and Chiayi Counties, for instance only four out of 37 colonies were monogynous in Chiayi County; but in a few areas, monogynous colonies outnumbered polygynous, for example, Linkou Township of Taipei County had 77 monogynous colonies out of 86. Meanwhile, the cuticular chemicals of fire ant from both monogyne and polygyne were analyzed by gas chromatography and gas chromatography-mass spectrometry, and both castes of queens and workers were compared. Some hydrocarbons (i.e., mostly alkanes) and non-hydrocarbons (i.e., alkaloids) were significantly different in quantities between the two social forms, moreover an alkaloid-like substance was only found in monogynous workers. The two social forms did not cluster separately by principal component analysis nor by discriminant analysis of the chemicals with significantly different quantities; nevertheless both of those analyses displayed the monogynous queens were closer to each other than polygynous. For these cuticular chemicals in the fire ant have genetic basis, we suggest that lower genetic variation in monogynous queens was probably caused by strong selection pressure of monogynous workers to the conspecific but different colonies. Besides, it is consistent with the asymmetrical gene flow from monogynous to polygynous populations through males in monogyne. .Therefore, the cuticular chemicals can not be cluster well between monogynous and polygynous ants, and the variance of chemicals of the latter is higher than those of the former. However, the chemicals involved in discrimination behaviors need to be further confirmed in future.