环纹榕传粉榕小蜂的传粉模式

榕小蜂的传粉结构、 传粉行为以及寄主榕树的花药胚珠比是判断榕-蜂互惠系统传粉模式的重要依据。本研究于2010年8月至2011年6月对位于云南西双版纳地区的试验样树环纹榕Ficus annulata进行观察, 对出榕果的环纹榕传粉榕小蜂Deilagaon annulatae进行电镜扫描和室内显微镜下对其进行行为观察。电镜扫描显示： 环纹榕传粉榕小蜂D. annulatae位于胸部的花粉筐消失, 具一片可粘附花粉的毛区, 提示其属于被动传粉的种类。传粉行为观察发现, 该蜂没有主动采集花粉的行为, 显然存在的花粉刷已丧失了主动收集花粉的功能。寄主植物环纹榕F. annulata属于典型的自动散粉让榕小蜂沾附花粉的榕树种类。环纹榕传粉榕小蜂D. annulatae是西双版纳热带地区已知58种传粉榕小蜂中唯一体色为黄色的种类, 该蜂偏爱在低温的夜晚出蜂。除了传粉榕小蜂, 一种金小蜂Lipothymus sp.也在雌花期进入榕果内繁殖, 并且其数量显著高于环纹榕传粉榕小蜂D. annulatae (P<0.0001)。自然单果的繁殖中, 环纹榕传粉榕小蜂的数量显著高于种子数量, 呈现出榕-蜂互惠系统中罕见的传粉榕小蜂主导的局面。综合榕-蜂的传粉特征显示, 环纹榕F. annulata及其传粉榕小蜂D. annulatae互惠系统是被动传粉的模式。     

榕树-榕小蜂互惠合作系统中的非对称性博弈

榕树及其传粉榕小蜂是自然界中目前所知道的关系最为紧密的互利共生系统之一。随着研究的深入, 越来越多的证据发现榕树-传粉榕小蜂之间互惠合作的过程中存在着复杂的竞争和对抗关系, 例如榕树与传粉榕小蜂之间对公共资源的竞争、传粉欺骗与宿主对传粉者的惩罚、榕树与传粉小蜂之间的“军备竞赛”等。在相互竞争或者对抗关系中, 双方表现出非对称性相互作用。其非对称性关系主要表现出如下3个特征: (1)收益不对称, 即榕树(宿主)与传粉榕小蜂(共生体)之间在资源利用等方面的实力不对称; (2)榕树与传粉榕小蜂之间的信息不对称; (3)进化速率不对称。这些非对称的相互作用可能导致种群的波动、榕树与传粉榕小蜂相互适应和进化策略的变化。因此, 理解榕树与传粉榕小蜂之间的非对称交互作用有助于理解为什么合作和冲突在互利共生关系中经常能同时存在, 也将有助于解释榕树-传粉榕小蜂种间相互关系和物种的多样性。     

榕树–榕小蜂协同进化中的非对称性相互作用及其集合种群效应

在协同进化研究中,传统观点认为物种之间的相互作用是对称性的,在进化过程中将形成一个稳定的均衡状态或进化稳定策略。然而,近来的观测和实验数据表明,物种间的协同进化可能存在非对称性相互作用,而且这种非对称性可能造成集合种群效应(metapopulation effect)或形成非均衡状态(例如混沌,chaos)。本文利用"榕树–榕小蜂"这一经典的协同进化模式系统来介绍协同进化过程中的非对称性相互作用,以及这种非对称性如何产生集合种群效应。在榕–蜂共生系统中,栖身于榕果中的榕小蜂除了传粉小蜂之外,还有一些"投机"的非传粉小蜂。非传粉小蜂比传粉小蜂具有更强的竞争力,而榕树则可惩罚不合作的非传粉蜂,同时奖励合作的传粉小蜂,从而形成了复杂的非对称性种间相互作用。在某个斑块生境中,非传粉小蜂通过竞争作用排斥传粉小蜂,然而随着非传粉小蜂在蜂群中的比例不断升高,榕树惩罚作用(如落果等)常常会导致整个非传粉蜂群的数量急剧下降,甚至局域性灭绝。随后,其他斑块的传粉小蜂则迁移过来填补空白。然而,随着传粉小蜂种群的数量增长,该斑块又会集聚更多的投机性非传粉小蜂,进而诱发榕树再次进行惩罚。这样的非对称性相互作用导致了非传粉小蜂、传粉小蜂以及榕树种群的集合效应,其种群大小呈现周期性的"此消彼长"式的循环。     

比较钝叶榕的三种传粉者及其传粉效率

钝叶榕 (Ficus curtipes)是雌雄同株,它除了依赖钝叶榕传粉榕小蜂Eupristina sp.传粉外,另外两种进入果内繁殖的杨氏榕树金小蜂Diaziella yangi 和Lipothymus sp.金小蜂也能有效地为它传粉,这3种小蜂同时产卵于雌花的子房内,在榕果内繁殖后代.通过控制性放蜂试验,比较研究钝叶榕3种传粉者的传粉效率,结果表明:自然状态下,3种小蜂在绝大多数榕果里只各进1头.在控制性放蜂试验中,3种小蜂的传粉效率均随着放入雌蜂数量的增加而增加,两种金小蜂的传粉效率有时比钝叶榕传粉榕小蜂的传粉效率还高.当钝叶榕传粉榕小蜂分别与两种金小蜂同时放入榕果内传粉时,其生产的种子数量居于或者是接近两种小蜂单独传粉时形成的种子数量,传粉效率没有显著增加.在比较3种小蜂单种分别放1头和2头的传粉效率时,增加单果放蜂数量,钝叶榕传粉榕小蜂和Lipothymus sp. 的平均传粉效率降低,但杨氏榕树金小蜂的平均传粉效率是增加的.对3种不同属传粉小蜂传粉效率的比较,可为研究榕-蜂互惠系统的互惠的起源提供依据.     

西双版纳聚果榕隐头果内小蜂群落结构及种间关系

聚果榕Ficus racemosa Linn.是雌雄同株榕树,它是西双版纳热带雨林生态系统中的一个常见种群。聚果榕必须依靠聚果榕小蜂Ceratosolen fusciceps Mayr传粉才能获得有性繁殖,而聚果榕小蜂又必须依靠聚果榕隐头果内短柱花繁衍后代,两者间形成了种间专一的互惠共生体系。同时,在其隐头果内还存在一个复杂的非传粉小蜂功能群,它们主要是榕树种子和传粉榕小蜂的寄生者。在云南省西双版纳自治州勐腊县勐仑镇选取了5个样地,对聚果榕单果内小蜂群落组成和种间相互关系进行研究。在不同时间段采集聚果榕单果242个,共收集小蜂366660头。聚果榕隐头果内有6种小蜂,隶属小蜂总科Chalcidoidae中的榕小蜂科Agaonidae、长尾小蜂科Caliimomidae、金小蜂科Pteromalidae,其中榕小蜂科中的C．fusciceps是聚果榕唯一的传粉者。金小蜂科中的Apocryta westwoodi Grandi和Apocryta sp.两个种是榕小蜂的寄生者,它们的寄生是传粉榕小蜂的种群数量减少因素之一。长尾小蜂科中Platyneura agraensis Joseph,Platyneura mayri Rasplus和Platyneura testacea Motschulsky3个种是寄生榕树种子或与榕小蜂争夺食物(瘿花)资源的小蜂类群,它们的出现与发生,致使聚果榕正在发育成种子的长柱小花形成瘿花,同时一部分种类把卵寄生在已被榕小蜂产卵的短柱小花子房中与榕小蜂争夺食物资源,致使榕小蜂食物资源的匮乏而死亡,对传粉小蜂种群有明显的影响。传粉小蜂从花托口钻入隐头果内,在隐头果内的长柱小花传粉和短柱小花子房中产卵,5种非传粉小蜂从隐头果外部把产卵器刺穿果肉把卵产在小花子房上。通过对5块样地隐头果内小蜂群落的综合分析发现,传粉榕小蜂为优势种群,而Apocryta sp.种的数量最少。在传粉及非传粉小蜂自然群体中各种类性比明显具有偏雌现象。小蜂群落表现异常的是聚果榕孤立株的样地,该样地传粉小蜂个体数量明显下降,非传粉小蜂个体数量则增加,各小蜂种群之间竞争激烈,出现C．fusciceps和P．mayri两个优势种。     

钝叶榕榕果内榕小蜂的产卵顺序及其群落结构

通过对钝叶榕榕小蜂行为的观察以及榕果内各类小花的统计,研究了钝叶榕12种榕小蜂的产卵行为和群落结构.结果表明：钝叶榕中除了传粉榕小蜂Eupristina sp.进入果腔产卵以外,还有2种非传粉榕小蜂（杨氏榕树金小蜂和Lipothymus sp.）与传粉榕小蜂在同一时期进入果腔产卵,其他9种非传粉榕小蜂（Walkerella sp.、Micranisa sp.、Sycophilomorpha sp.、Philotrypesis sp.、Sycosapter sp.、Sycobia sp.、Ficomila sp.、Ormyrus sp.和Sycophila sp.）在果外产卵;在钝叶榕榕小蜂群落中,传粉榕小蜂占整个群落总数的62.11%,是该群落的优势种,杨氏榕树金小蜂和Lipothymus sp.分别占整个群落总数的27.19%和4.71%,其他9种非传粉榕小蜂占5.99%.钝叶榕中的非传粉榕小蜂通过各自产卵时序和幼虫食性分化的繁殖策略来分配榕果中的资源,以实现自身繁殖.非传粉榕小蜂与传粉榕小蜂的数量变化呈显著负相关,但非传粉榕小蜂与榕果内的种子没有相关性.     

歪叶榕非传粉小蜂的繁殖策略及其对榕-蜂共生系统的影响

2004年8月至2005年8月在西双版纳热带植物园内,通过广泛收集歪叶榕榕小蜂标本、非传粉小蜂产卵行为学观察和阻止传粉者入果等实验方法,研究了我国西双版纳热带雨林下的一种榕树——歪叶榕Ficus cyrtophylla的榕小蜂群落组成结构、非传粉小蜂的繁殖策略以及它们对榕-蜂共生系统的影响。结果表明,歪叶榕中除了具有唯一传粉榕小蜂Blastophag sp.以外,还具有3种非传粉小蜂Platyneura sp.、Philotrypesis sp.和Sycoscapter sp.。在歪叶榕榕小蜂群落中,传粉榕小蜂占整个群落总数的92.21%,是群落的最主要组成者;主要的非传粉小蜂是Sycoscaptersp.,占5.78%; 其次是Philotrypesissp.,占1.84%,而Platyneurasp.仅占群落总数的0.17%。歪叶榕中的非传粉小蜂通过各自产卵时间和食性分化的策略来利用榕果中的资源繁殖后代。非传粉小蜂寄生使传粉榕小蜂的总数和其雌蜂数量都显著地降低,但是对传粉小蜂雄蜂数量没有显著影响,从而导致传粉榕小蜂的雄性性比显著地增加。这说明非传粉小蜂在选择寄居宿主时具有明显的倾向性,而且更多地将卵产于含有雌性传粉小蜂的瘿花之中。因此,非传粉小蜂通过减少雌性传粉小蜂的数量而降低了榕树的雄性适合度,从而在一定程度上对榕 蜂共生系统的稳定存在和发展产生了负面影响。     

榕果挥发物对传粉榕小蜂的吸引作用

榕树 /榕小蜂专一性共生系统的维持 ,与榕树开花期释放的特殊的挥发性化合物以及榕小蜂对其寄主榕树的化学识别和定位紧密相关。研究选取了西双版纳地区常见的 3种榕树 ,即对叶榕 Ficus hispida、木瓜榕 F.auriculata和鸡嗉子榕F.semicordata的榕果作为实验材料 ,利用野外诱捕实验、室内生物检测实验检测传粉榕小蜂 Hymenoptera:ChalcidoidaeAgaonidae对 12种信息化合物及榕果的二氯甲烷浸提物的趋向性反应 ,研究不同榕属植物的传粉榕小蜂对相同的信息化合物的反应差异 ,以及传粉榕小蜂受不同发育时期榕果浸提物吸引的显著性程度。诱捕实验中对叶榕小蜂 Ceratosolen solmsimarchali对香叶醇的趋向性反应显著 ,大果榕小蜂 C.emarginatus对接受期榕果浸提物和芳樟醇都有明显的趋向性反应 ,而对间花期榕果浸提物则无显著反应。嗅觉仪生物检测实验中 ,鸡嗉果榕小蜂 C.gravelyi对香叶醇和松油醇都表现出显著的趋向性反应。结果表明 ,对叶榕、鸡嗉子榕传粉榕小蜂对 12种信息化合物的反应存在一定的差异 ,木瓜榕传粉榕小蜂对香叶醇和木瓜榕接受期榕果浸提物的趋向性反应比间花期榕果强得多     

同种榕小蜂在木瓜榕两种地理型上的繁殖及传粉特征

榕树与其传粉榕小蜂的共生关系常被认为是专一性的,但该系统中有些榕小蜂可在不同种榕树或者同种榕树的不同亚种、变种和地理型上产卵和传粉。探讨榕小蜂在不同寄主中的繁殖和传粉特征,有利于理解非专性榕蜂系统形成的过程及稳定机制。本研究中,作者分别对比分析了传粉榕小蜂Ceratosolen emarginatus在木瓜榕(Ficus auriculata)的两种地理型auriculata-form和oligodon-form上的产卵和传粉特征。结果显示,进蜂量为1,2,3只时,相同寄主上的榕小蜂后代和种子数量均随进蜂量的增加而增加,且平均单只繁殖雌蜂的后代及种子数量均无差异。这可能是由于进蜂量较低时,两寄主可被利用的繁殖资源较充足,榕小蜂间不存在干扰竞争,可最大化地利用雌花资源。另一方面,进蜂量相同时,同一寄主上产生的种子数量明显多于榕小蜂后代数量,说明榕树的繁殖利益更占优势。榕小蜂在auriculata-form上产生的后代总量和平均单只雌蜂后代数量与oligodon-form均无差异,但后者产生的种子数量明显多于前者,说明当繁殖资源充足时寄主不影响榕小蜂的繁殖,然而寄主差异影响种子产生,即auriculata-form和oligodon-form的繁殖能力已发生分化。     

聚果榕两种非传粉小蜂产卵与果实脱落的关系

在聚果榕与其传粉榕小蜂Ceratosolen fusciceps组成的互利共生系统中,与传粉榕小蜂共存的还有榕树的寄生性小蜂Platyneura testacea和Platyneura mayri.这些寄生性小蜂由于不能给榕树带来任何收益而只是利用榕树的种子或与传粉小蜂竞争植物的雌花资源,因而可能导致榕树与榕小蜂之间合作系统的崩溃.植物果实的脱落机制普遍被认为是维持系统稳定的关键因素之一.然而,定量实验和野外观测发现P. mayri产卵并不会引起果实脱落,只有P. testacea产卵会使果实大量脱落.通过对3株样树进行比较发现:当产卵时的P. testacea数量越多时,它产生的瘿花数就越多,榕果发生脱落的比例越高.P. testacea的过度产卵是导致榕果选择性脱落的主要原因.结果表明:脱落机制并不能完全阻止非传粉小蜂的寄生,榕树只能选择性地脱落掉先于传粉小蜂产卵的榕果.这也同时表明维持榕树与榕小蜂互利共存的机制不仅仅只有榕果的脱落机制,可能还存在其它未被发现的机制.     

Can fine-scale post-pollination variation of fig volatile compounds explain some steps of the temporal succession of fig wasps associated with Ficus racemosa?

Volatile organic compounds (VOCs) emitted by flowers play an essential role in mediating the attraction of pollinators. However, they also attract other species exploiting resources associated with flowers. For instance, VOCs emitted by figs play a major role in encounters between Ficus spp., their mutualistic pollinating wasps, and all the members of the community of non-pollinating fig wasps (NPFWs) that exploit the mutualistic interaction. Because pollinators might be in limited supply for a tree bearing many inflorescences, the plant might maximize its individual reproductive success by reducing the attractiveness of inflorescences once they are pollinated, so that pollinators orient only towards the tree's unpollinated figs. Changes in VOCs emission that bring this about could represent an important cue for NPFWs that exploit particular stages of fig development. In this study, by monitoring precisely the presence of fig-associated wasps on figs of F. racemosa, a common widespread fig species, we demonstrated that 4–5 days and 15 days following pollination represent two critical transitional steps in the succession of different wasp species. Then, focusing on the first one of these transitional steps, by investigating the composition of fig VOCs at receptivity and from 1 to 5 days following pollination, we detected progressive quantitative and qualitative variation of floral scent following pollination. These changes are significant at 5 days following pollination. The qualitative changes are mainly due to an increase in the relative proportions of two monoterpenes (α-pinene and limonene). These variations of the floral VOCs following pollination could explain why pollinating wasps stop visiting figs very shortly after the first pollinators enter receptive figs. They also possibly explain the succession of non-pollinating wasps on the figs following pollination.     

Evolution of a complex coevolved trait: active pollination in a genus of fig wasps

Only three insect lineages have evolved complex active pollination behaviour and only fig wasps (Agaonidae) have also reverted from active to passive pollination. Previously, it was assumed that there was a single origin of active pollination in fig wasps, followed by one independent loss in each of five genera. We show here that there have been three to six changes in pollination behaviour within just one genus (Pleistodontes). The results suggest multiple gains of active pollination in fig wasps, but are sensitive to assumptions about the relative costs of gaining and losing this complex behaviour. In addition, previous comparative studies at higher taxonomic levels have reported correlated evolution between active pollination in wasps and low anther/ovule ratios in figs. We report that changes in pollination behaviour between congeneric species correlate perfectly with changes in anther/ovule ratios in the host figs, showing no phylogenetic inertia in coadaptation at the species level.     

Foundress re-emergence and fig permeability in fig tree-wasp mutualisms

Some female pollinating fig wasps (foundresses) re-emerge from figs after oviposition/pollination. We investigated why this occurs in the mutualism between the gynodioecious Ficus montana and Liporrhopalum tentacularis. Re-emergence increased with foundress density in figs and some foundresses oviposited in two male figs, indicating that they re-emerge because of oviposition site limitation. Re-emergence was independent of fig diameter, indicating that permeability is not because of fig age at entry. Rather, as some foundresses also pollinate two female figs we suggest permeability is selected for because it increases pollinator production and/or efficiency (although, potentially opposing these hypotheses, we also found between-tree differences in permeability in male figs). In addition, we show that re-emergence is much more common than previously suspected, and more common from gynodioecious than monoecious fig species. We argue that our findings in F. montana could explain this pattern of incidence.     

Shift to mutualism in parasitic lineages of the fig/fig wasp interaction

The interaction between Ficus and their pollinating wasps (Chalcidoidea, Agaonidae) represents a striking example of mutualism. Figs also host numerous non-pollinating wasps belonging to other chalcidoid families. We show that six species of Ficus that are passively pollinated by the agaonid genus Waterstoniella also host specific wasps belonging to the chalcidoid genera Diaziella (Sycoecinae) and Lipothymus (Otitesellinae). Both belong to lineages that are considered as parasites of the fig/fig wasp mutualism. We show that these wasps are efficient pollinators of their hosts. Pollen counts on wasps of a species of Diaziella hosted by Ficus paracamptophylla show that Diaziella sp. transports more pollen than the associated pollinator when emerging from its natal fig. Further, the number of pollinated flowers in receptive figs is best explained by the number of Diaziella plus the number of Waterstoniella that had entered it. Figs that were colonised by Diaziella always produced seeds: Diaziella does not overexploit its host. Similarly, figs of Ficus consociata that were colonised solely by a species of Lipothymus produced as many seeds as figs that were colonised only by the legitimate pollinator Waterstoniella malayana . Diaziella sp. and Lipothymus sp. seem to pollinate their host fig as efficiently as do the associated agaonid wasps. Previous studies, on actively pollinated Ficus species, have found that internally ovipositing non-agaonid wasps are parasites of such Ficus species. Hence, mode of pollination of the legitimate pollinator conditions the outcome of the interaction between internally ovipositing parasites and their host.     

Complex interactions on fig trees: ants capturing parasitic wasps as possible indirect mutualists of the fig–fig wasp interaction

Like other mutualisms, pollination mutualisms attract parasites, as well as opportunistic and specialist predators of the pollinators and parasites. These associated species influence the evolutionary dynamics of pairwise mutualisms. Predatory ants are frequent associates of pollination mutualisms, but their effects on the complex interactions between plants, pollinators and parasites have not yet been clearly established, even in the case of the well-described obligate interaction between figs and fig wasps. We attempted to quantify such effects for ants associated with three fig species, two dioecious ( Ficus condensa [Bruneï], F . carica [France]) and one monoecious ( F . racemosa [India]). In all these cases, ant presence on a fig tree strongly reduced the number of parasitic wasps on the figs. Experimental exclusion of ants resulted in an increase in the number of non-pollinating fig wasps on F . condensa and F . racemosa . Experimental ant supplementation led to a decrease in the number of non-pollinating fig wasps on F . carica . Moreover, on F . condensa , the level of reduction of the number of parasitic wasps depended on the number and identity of the ants. On F . carica , non-pollinating fig wasps even avoided trees occupied by the dominant predatory ant. The consistency of the effect of ants in these three cases, representing a geographically, ecologically, and taxonomically broad sample of figs, argues for the generality of the effect we observed. Because reduction of parasitism benefits the pollinator, ants may be considered as indirect mutualists of plants and pollinators in the network of complex interactions supported by fig trees.     

Between‐species facilitation by male fig wasps in shared figs

1. Facilitation is recorded from diverse plant–insect interactions, including pollination and herbivory. 2. The significance of facilitation resulting from the behavior of males of multiple fig wasp species inside figs was investigated. Female fig wasps emerge from natal figs via exit holes dug by males, especially male pollinators. When no males are present, the females struggle to escape and may die. 3. Ficus microcarpa L. is a widely‐established invasive fig tree from Southeast Asia. Its pollinator is absent in South Africa, so the tree cannot reproduce, but two Asian non‐pollinating fig wasps (NPFW) Walkerella microcarpae and Odontofroggatia galili occupy its figs. Abundance patterns of the two NPFW and the proportion of male‐free figs in South Africa, Spain (where the pollinator is introduced), and in China, where the native fig wasp community is diverse, were compared to determine the consequences of reduced species richness for insect survival. 4. Female fig wasps in male‐free figs were found to be trapped, and small clutch sizes contributed to the absence of males in both species. The presence of pollinators in Spain allowed most NPFW to develop in figs containing males. Far more male‐free figs were present in South Africa, elevating mortality rates among female NPFW. Facilitation of female release by males of other NPFW species nonetheless benefitted the rarer species. 5. Selection pressures in South Africa currently favour greater aggregation of NPFW offspring and/or less female biased sex ratios.     

传粉榕小蜂与非传粉小蜂间寄主识别行为的趋同进化

在高度专性传粉的榕树-榕小蜂互惠共生系统中普遍存在着一些非传粉小蜂,它们中的一些种类进入果腔后也能为榕树传粉,且在形态和物候上已与传粉榕小蜂发生了趋同进化。但其寄主识别行为是否也与传粉榕小蜂发生了趋同进化还不得而知。我们在西双版纳选择了钝叶榕(Ficuscurtipes)及其3种进果繁殖小蜂开展了相关的行为实验。3种小蜂中,1种是钝叶榕的专性传粉榕小蜂(Eupristina sp.),另外2种是寄居性非传粉小蜂(杨氏金小蜂Diaziellayangi和Lipothymus sp.),这2种非传粉小蜂进入果腔后也像传粉榕小蜂那样为钝叶榕传粉。我们以钝叶榕不同发育时期的榕果及这3种小蜂为材料,采用Y型嗅觉仪观察了这3种小蜂对各发育时期榕果和信息化学物质6-甲基-5-庚烯-2-醇、6-甲基-5-庚烯-2-酮及这2种化合物的混合物的选择行为。结果表明,当提供雌花期榕果与其他发育时期榕果和空气对照供这3种小蜂选择时,它们均显著地偏向于选择雌花期榕果;当提供雄花期榕果与其他发育时期榕果和空气对照供这3种小蜂选择时,它们均显著地偏向于选择其他发育时期榕果和空气对照,即都会避开雄花期榕果;此外,这3种小蜂均对钝叶榕雌花期果释放的一种主要化合物6-甲基-5-庚烯-2-醇的同一剂量(1μL)表现出显著的偏好。这一结果为传粉榕小蜂与非传粉小蜂间的寄主识别行为趋同进化的假说提供了证据。     

How to be a fig wasp down under: The diversity and structure of an Australian fig wasp community

Endophytic insects and their parasitoids provide valuable models for community ecology. The wasp communities in inflorescences of fig trees have great potential for comparative studies, but we must first describe individual communities. Here, we add to the few detailed studies of such communities by describing the one associated with Ficus rubiginosa in Australia. First, we describe community composition, using two different sampling procedures. Overall, we identified 14 species of non-pollinating fig wasp (NPFW) that fall into two size classes. Small wasps, including pollinators, gallers and their parasitoids, were more abundant than large wasps (both galler and parasitoid species). We show that in figs where wasps emerge naturally, the presence of large wasps may partly explain the low emergence of small wasps. During fig development, large gallers oviposit first, before and around the time of pollination, while parasitoids lay eggs after pollination. We further show that parasitoids in the subfamily Sycoryctinae, which comprise the majority of all individual NPFWs, segregate temporally by laying eggs at different stages of fig development. We discuss our results in terms of species co-existence and community structure and compare our findings to those from fig wasp communities on other continents.     

Ecology of a fig ant–plant

Mutualistic interactions are embedded in networks of interactions that affect the benefits accruing to the mutualistic partners. Figs and their pollinating wasps are engaged in an obligate mutualism in which the fig is dependent on the fig pollinator for pollination services and the pollinator is dependent on fig ovules for brood sites. This mutualism is exploited by non-pollinating fig wasps that utilise the same ovules, but do not provide a pollination service. Most non-pollinating wasps oviposit from outside the inflorescence (syconium), where they are vulnerable to ant predation. Ficus schwarzii is exposed to high densities of non-pollinating wasps, but Philidris sp. ants patrolling the syconia prevent them from ovipositing. Philidris rarely catch wasps, but the fig encourages the patrolling by providing a reward through extra-floral nectaries on the surface of syconia. Moreover, the reward is apparently only produced during the phase when parasitoids are ovipositing. An ant-exclusion experiment demonstrated that, in the absence of ants, syconia were heavily attacked and many aborted as a consequence. Philidris was normally rare on the figs during the receptive phase or at the time of day when wasp offspring are emerging, so predation on pollinators was limited. However, Myrmicaria sp. ants, which only occurred on three trees, preyed substantially on pollinating as well as non-pollinating wasps. F. schwarzii occurs in small clusters of trees and has an exceptionally rapid crop turnover. These factors appear to promote high densities of non-pollinating wasps and, as a consequence, may have led to both a high incidence of ants on trees and increased selective pressure on fig traits that increase the payoffs of the fig–ant interaction for the fig. The fig receives no direct benefit from the reward it provides, but protects pollinating wasps that will disperse its pollen.