長腳捷蟻的食物選擇與食物傳遞行為

Abstract

不同螞蟻的不同覓食行為，往往代表著不同的覓食策略，使螞蟻能適應不同的環境，進而能廣泛分布在不同的地區。而一般常以最佳覓食理論 (Optimal foraging theory) 來描述覓食策略，亦即以蟻巢（或群落）為單位，假設選汰偏好使蟻巢在單位時間內獲得最大的淨利。本文從「花費」和「利益」為出發點，以長腳捷蟻 (Anoplolepis gracilipes) 為材料，分別探討食物以及覓食距離對於螞蟻覓食行為之影響。此外，也針對其傳遞食物的行為進行深入的探討及模擬。 首先探討食物大小的選擇，由搜尋時間、處理時間的結果顯示：當供給三種不同重量的果蠅作為食物時，長腳捷蟻偏好重量居中之果蠅，顯示這是「花費」和「利益」權衡之結果。不論是何種果蠅，均只要一隻螞蟻即可以搬動，所以在提供果蠅為食物的部份，並沒有觀察到合作搬運以及分泌足跡費洛蒙 (trail pheromone) 的行為。相對於果蠅，在以蟋蟀為食物的試驗部分，長腳捷蟻表現出絕然不同的覓食策略：當偵查蟻 (scout) 發現到有大型食物時，他們會快速回巢募集同伴，同時還會分泌足跡費洛蒙，以群體合作的方式 (Group foraging) 將食物搬運回巢。 接著探討不同覓食距離處理之效應，當覓食距離較遠時，螞蟻需要經過較久的時間，才會發現該食物區；而螞蟻出現在食物區中之高峰是在食物區中之食物幾乎均已被搬完時，因此長腳捷蟻之訊息傳遞似乎有延遲的現象；此外，距離較近的覓食區，會吸引較多的的覓食者進入。若同時提供兩種不同距離的覓食區供長腳捷蟻選擇，亦偏好優先利用較近的食物區，在較近食物區不但有較多的螞蟻進入且搬運食物回巢的速度均顯著較快。結果符合最佳覓食理論的預測。 當螞蟻所遭遇食物的大小是單隻螞蟻個體即可搬運，則在螞蟻將食物搬運回巢的過程中，還會出現傳遞食物的行為，這樣的行為稱為任務分工 (Task partitioning)，而所有參與搬運者形成傳遞鍊 (Transport chain)。傳遞食物有直接傳遞 ( Bucket brigade) 和間接傳遞兩種，兩者皆可能同時發生於同一個食物搬運回巢的過程中，而當食物傳遞發生時，若兩隻螞蟻有直接接觸往往會發生拉扯（Tug）的行為。當提供果蠅作為長腳捷蟻的食物時，野外群落的研究結果顯示，在蟻巢入口附近有高達38%的食物發生拉扯行為。在實驗室更進一步探討食物傳遞行為。結果顯示在食物傳遞鏈中工蟻的體型會由小到大排列且搬運食物的速度會依序增加，以上證據支持任務分工的合作假說。然而傳遞鏈花費較個體長的時間才能將食物搬運回巢，且在很短的覓食距離中，傳遞食物的次數最高可達5次，而且拉扯的行為一直到巢口還會發生，此外，不論食物是單獨被搬運回巢或是以傳遞的方式搬運回巢，實際被搬運回蟻巢的距離都遠大於覓食區直線距離的2倍以上，因此，以上證據並不支持任務分工的合作假說。由於長腳捷蟻為多后型的螞蟻種類，因此推測發生拉扯的螞蟻可能來自不同的蟻后所致，其目的是為了搶奪食物給親屬，此即偏袒假說，不過這個假說仍需更進一步的驗證。 最後以收穫蟻為例，針對食物傳遞行為進行模擬分析，分別探討螞蟻體重、食物重量以及搬運食物回巢的速度等因子間之相互關係及作用。模擬結果發現所搬運的食物重量越大，搬運速度會隨之下降，但獲利卻隨之增加，因此假設選汰偏好搬運速度最大化並不合理。不同體型大小的螞蟻個體各有其最適合搬運的食物大小。較小的食物由較小的螞蟻帶回巢，較大的食物由較大的螞蟻搬回巢，如此對整個蟻巢的獲利才是最佳。而影響食物被搬運回巢的策略，最主要受到搬運速度的影響。當搬運速度夠快時，食物是由單隻螞蟻個體獨自搬運回巢，然而當搬運速度不夠快，容易被其他螞蟻個體阻擋時，食物傳遞鍊就會形成。而影響搬運速度的主要因子則是螞蟻個體大小以及所搬運的食物重量。

Foraging behavior is crucial for an ant species to adapt to its environment. Ants may adopt different foraging strategies to cope with environmental change, e.g., available food size and food-patch distance. Therefore the objective of this study was to explore the effect of food size and foraging distance on the retrieval behavior of yellow crazy ant, Anoplolepis gracilipes in the field and in laboratory. Besides, we also investigate an interesting retrieval strategy, sequential load transfer behavior, and explore its adaptive value with modeling. To explore the retrieval behavior and food selection of A. gracilipes, three species of fruit flies that differed in size (Drosophila virilis > D. albomicans > D. melanogaster in weight) were given to workers for choosing in two experimental designs. Both experiments showed that the workers prefer the fly of medium-sized. Choosing medium-sized food could be a result of trade-off between the cost and benefit of carrying different sizes of food. We also provided crickets (Gryllus bimaculatus) to ants as a large-sized food supply and fruit fly as small-sized food supply. We compare the foraging behaviors observed from the two kinds of food patches. The results showed that when workers encountered a large-sized food, i.e., cricket, they quickly returned back to nest and laid trail to recruit nestmates. Under this condition, foraging workers could be divided into two types, the scouts and the recruits. On the contrary, when the workers encountered small-sized food, all the workers foraged solitarily and their foraging routes were completely different. Therefore, foraging strategy of the yellow crazy ant is very plastic. They can adjust their foraging behavior to different types and sizes of food quickly. In the food-patch distance experiment in the field, the short-distance patch was discovered and utilized much more quickly than the far-distance patch. When ants were given the choice between two food patches with equal quality but with different foraging distance. The results are consistent with the prediction of the optimal foraging theory, i.e., the near patch was utilized much more quickly than the far patch in various distance-ratio treatments and this preference is enhanced by the large distance ratio. Ant flows and retrieval speed in the near patch were also significantly higher than those in the far patch. Many social insects show sequential cooperation in the foraging process, in which workers form a transport chain and transfer a load from one to another, this is known as task partitioning. In this study, sequential load transport in the ant, A. gracilipes, was analyzed, and the assumptions and predictions of the bucket brigades (BBs) hypothesis were tested. Our results suggested that in the vicinity of the nest entrance, 38% of foods were retrieved with tug behavior. Worker size and the retrieval speed increased with each transfer, and the foraging efficiency improved along the transport chain to the nest. These results support the BBs hypothesis of sequential cooperation. However, the retrieval time for the transport chain was significantly longer than that of individual foraging. Moreover, distances of both the transport chain and individual foraging were significantly longer than that of the shortest linear route, and the distance covered by larger workers was shorter. Thus the loaded worker probably showed avoidance behavior in order to avert being robbed. Food transfer might not be a cooperative strategy in A. gracilipes. Nepotism hypothesis that states load robbing for close kin is first proposed to account for the maintenance of a non-cooperative sequential load transport in a colony. Finally, based on the data of seed-harvesting ant, we run the simulation of load transfer and test the effects of worker size, load size and retrieval speed on load transfer. Results showed that speed decreases abruptly with increasing loading ratio of a worker. The optimal load for a worker increases with increasing body size and there is an optimal load for an ant of specific size. Compared the speed ratio for a worker in a transfer chain to that in a single foraging, results showed that longer distance and higher loading mass carried by the worker favor the evolution of transfer chain.