物种间相互作用主要包括对双方有利的互利共生、仅对一方有利的敌对关系以及对双方有害的竞争作用，不同的种间相互作用会导致作用双方不对等的收益与代价，从而使得相对应的物种枝系形成不同的协同多样化模式。在榕树-榕小蜂系统中，榕树与传粉榕小蜂为互利共生关系、榕树与造瘿榕小蜂为敌对植食关系、传粉榕小蜂与造瘿榕小蜂为竞争关系、拟寄生榕小蜂与传粉和造瘿榕小蜂为敌对寄生关系。榕小蜂体内寄生着沃尔巴克氏体（Wolbachia），能够对榕小蜂的生殖进行调控，但也可能存在互利共生的作用。榕小蜂体内还有很多其他的内共生细菌，可能也与榕小蜂具有互利共生作用。因此，在该系统中同时涵盖了多种不同收益/代价的种间关系。
首先，本文以西双版纳地区26种榕树的26种传粉榕小蜂、33种造瘿榕小蜂和81种拟寄生榕小蜂为研究材料，利用榕树的ITS、ETS、G3pdh和trnL-F基因，利用榕小蜂的18S rRNA、28S rRNA和COI基因，重建榕树和榕小蜂的系统发育树，并分别利用基于距离矩阵的PACo方法和基于拓扑结构的Jane软件对每两者之间进行协同谱系分析。结果显示，互利共生的榕树与传粉榕小蜂之间谱系关系显著一致，具有更多的协同成种事件；敌对植食关系的榕树与造瘿榕小蜂之间谱系关系不一致，具有更多的宿主转移事件；竞争关系的传粉榕小蜂与造瘿榕小蜂之间谱系关系显著一致，具有较多的协同成种事件和丢失事件；拟寄生榕小蜂与宿主榕小蜂直接的敌对寄生关系及与榕树间接的敌对寄生关系间均有显著一致的谱系关系，且具有较多的协同成种事件和较多的宿主转移事件。上述结果表明相互有利的个体能够推动二者形成紧密的谱系关系并促进协同成种。相互有害的个体能够推动二者相互关系的断裂，产生丢失事件；但当双方竞争的资源是生活史必需的资源且高度一致时，反而会出现较多的协同成种。对一方有害的敌对关系，由于一种榕树同时遭受多种造瘿榕小蜂或一种榕小蜂同时遭受多种拟寄生榕小蜂，反而表现出造瘿榕小蜂或拟寄生榕小蜂较多的宿主转移事件，这种宿主转移的强度可能是由造瘿榕小蜂内部或拟寄生榕小蜂内部竞争的强弱和可利用资源的多少而决定的。另外，拟寄生榕小蜂与宿主榕小蜂之间还发生较多的协同成种，表明拟寄生榕小蜂对宿主小蜂的依赖性。
其次，本文以西双版纳地区13种雌雄同株榕树和13种雌雄异株榕树的28种传粉榕小蜂、45种造瘿榕小蜂和84种拟寄生榕小蜂为实验材料，利用Wolbachia的16S rRNA和5个分型管家基因，调查了榕小蜂感染Wolbachia的情况，重建Wolbachia系统发育树，与榕小蜂和榕树进行协同系统发育分析。结果显示，榕树的性系统会影响榕小蜂Wolbachia感染率，雌雄异株榕树的榕小蜂Wolbachia感染率显著高于雌雄同株榕树榕小蜂的感染率，不同性系统榕树的榕小蜂感染的Wolbachia菌株具有一定的差别；Wolbachia与雌雄异株榕树之间具有显著一致的谱系关系，这可能是因为Wolbachia在雌雄异株榕树中受到强烈的正选择作用，可能会帮助榕小蜂特别是传粉榕小蜂识别能够繁殖后代的雄树，从而达到种群的稳定维持，并且雌雄异株榕树中较高的Wolbachia感染率，可能是促进传粉榕小蜂物种形成的一个因素。不同榕小蜂类型也会影响榕小蜂的Wolbachia感染率，传粉与造瘿榕小蜂的Wolbachia感染率都显著高于拟寄生榕小蜂的感染率，这可能是因为Wolbachia在植食性榕小蜂中受到强烈的正选择作用，传粉榕小蜂和造瘿榕小蜂感染的Wolbachia可能会帮助二者抵御拟寄生榕小蜂的寄生。另外，我们发现同一种榕树的榕小蜂之间没有相似的Wolbachia菌株，拟寄生榕小蜂与其宿主榕小蜂之间也没有相似的Wolbachia菌株，表明西双版纳地区榕小蜂感染的Wolbachia很少通过近距离接触或者通过食物链进行水平传递。Wolbachia与各类型榕小蜂间具有显著一致的谱系关系，表明Wolbachia的传递方式主要还是垂直传递，并且二者之间存在长期的协同进化关系。除此之外，Wolbachia与拟寄生榕小蜂之间具有较多的协同成种和较少的宿主转移事件，可能是由于拟寄生榕小蜂抗性作用的严格选择，使得二者寄生关系更为稳定。
最后，本文以巴拿马地区10种常见榕树的10种传粉榕小蜂、3种造瘿榕小蜂和1种拟寄生榕小蜂为实验材料，利用Sanger测序和metabarcoding测序获得16S rRNA序列信息，得到榕小蜂内共生细菌的群落结构和谱系关系。结果显示大多数传粉榕小蜂内共生细菌中主要为肠杆菌科（Enterobacteriaceae）和Pantoea属的细菌，非传粉榕小蜂内共生细菌主要为Wolbachia属的细菌，两种类型榕小蜂内共生细菌的群落差异显著。主动传粉的Pegoscapus属榕小蜂和被动传粉的Tetrapus属榕小蜂的内共生细菌的群落无明显差异。肠杆菌科和Pantoea属的细菌可能与榕小蜂具有互利共生的关系，从而促进榕树榕小蜂系统的长期稳定，Wolbachia属的细菌可能能够帮助非传粉榕小蜂逃避榕树的惩罚机制。
综上所述，本论文揭示了榕树-榕小蜂系统中不同种间相互作用的协同多样化模式，并探究比较了不同种间相互作用之间可能的相互影响及系统中一些特征对协同多样化模式的影响。

Pairs of species can interact in manners that are mutually beneficial (mutualism), mutually detrimental (competition), or beneficial to one partner but detrimental to the other (antagonism: predation, herbivory, parasitism). Distinct interactions have contrasting ecological effects that, sustained over generations, can influence micro- and macroevolution. Dissimilar benefits and costs for these interactions should cause contrasting codiversification patterns between interacting clades. In fig microcosm, pairs of fig and pollinating wasp species engage in highly specialized mutualisms, and parasitic galling wasp species act as antagonists of figs. Because both mutualistic pollinating and parasitic galling wasps require the same resource for reproduction, they can compete with each other. Parasitoid wasps rely on other wasp larva for reproductive development, so they act as antagonists of pollinating or galling wasps. Fig wasps are frequently infected with Wolbachia, which can manipulate host reproduction and may also provide fitness benefits. Besides Wolbachia, there are many other endosymbiotic bacteria in wasps and they may have beneficial effects on fig-fig wasp mutualism. Fig microcosm provides a singular opportunity to compare the cophylogenetic relationships associated with multiple interspecific interactions.
Firstly, we collected 26 species of mutualistic pollinating fig wasps, 33 species of parasitic galling wasps, and 81 species of parasitoid wasps associated with 26 fig species in Xishuangbanna, Yunnan. Using ITS, ETS, G3pdh and trnL-F genes for figs, 18S rRNA、28S rRNA and COI for fig wasps, we inferred the phylogenies for all four clades. We then compared the pairs of phylogenies to assess phylogenetic congruence and the relative frequencies of cospeciation, association duplication, switching and loss. The paired phylogeny of figs with their mutualistic pollinating wasps was significantly congruent and involved more cospeciation, while that of figs and their parasitic galling wasps did not show congruence and involved more host switching. Codiversification of pollinating and their competitor galling wasps involved more cospeciation and loss and showed significant phylogenetic congruence. The parasitoid wasps exhibited significant phylogenetic congruence with their host wasps and with their indirect host figs. Codiversification of parasitoid wasps with their hosts involved more cospeciation and more host switching. The results indicated that beneficial interactions between individuals promote coalitions, and negative interactions promote dissociation. Mutualism tends to facilitate cospeciation, competition promotes interaction loss, and antagonism encourages association switching. However, galling wasp species likely experience highly similar resource with competitor pollinating wasps, they were involved more cospeciation and showed high phylogenetic congruence. If hosts combat diverse antagonistic species, individual interactions may have little influence on speciation but cause more host switching. Moreover, this host switching may depend on the strength of the competition within parasites and limited resource. Parasitoids may have strong dependence on their host wasps because of the high congruence and cospeciation between the phylogenies.
Secondly, we collected 28 species of pollinating wasps, 45 species of galling wasps, and 84 species of parasitoid wasps associated with 13 monoecious and 13 dioecious fig species in Xishuangbanna, Yunnan. Using 16S rRNA and MLST genotypes of Wolbachia, we surveyed Wolbachia infection frequency and inferred the phylogenies. We then compared the phylogenies between fig wasps and Wolbachia as well as figs and Wolbachia to assess phylogenetic congruence and the relative frequencies of cospeciation, association duplication, switching and loss. Fig sexual system can significantly influence fig wasp Wolbachia infection frequency. Specifically, wasps associated with dioecious figs presented higher Wolbachia infection frequency than wasps with monoecious figs. Fig sexual system was slightly consistent with Wolbachia strains, and dioecious figs showed high phylogenetic congruence with Wolbachia. We suspect that Wolbachia in dioecious figs may under high positive selection which might help pollinating wasps distinguish male trees from costly female trees. The high Wolbachia infection frequency might be a factor for promoting fig wasp speciation in dioecious figs. Different types of fig wasp show distinct Wolbachia infection frequency. Phytophagous wasps including pollinating and galling wasps present higher Wolbachia infection frequency than carnivorous parasitoid wasps. Wolbachia may also under high positive selection in phytophagous wasps and can help them defend parasitoid wasps. Moreover, neither the close ecological association within specific host fig nor food chain of host and parasitoid wasps has similar Wolbachia strain. It indicated that Wolbachia might not transmit through close contact or food chain. The highly congruent phylogenies between wasps and Wolbachia indicated that Wolbachia mainly transmitted through vertical transmission and involved long term codiversification with their wasp hosts in local community in Xishuangbanna. Parasitoid wasps presented the lowest Wolbachia infection frequency and involved more cospeciation and less host switching. It may because parasitoid have resistance of Wolbachia and have strong selection on Wolbachia.
Thirdly, we collected 10 fig pollinating wasp species, 3 parasitic galling wasp species and 1 parasitoid wasp species associated with 10 common fig species near Panama Canal, Republic of Panama. We sequenced endosymbiotic bacteria 16S rRNA cultured from the wasp ovaries and conducted metabarcoding of several wasp ovaries through Illumina sequencing. We finally obtain the community structure and phylogeny of these endosymbiotic bacteria. Some prominent taxon, Enterobacteriaceae and Pantoea, were identified in several pollinating wasp samples, while Wolbachia was more common in parasitic wasp samples. Pollinating wasps and parasitic wasps showed significant difference in endosymbiotic bacteria community structure while that of active and passive pollinating wasps was not significantly different. We suspect Enterobacteriaceae and Pantoea may play a role in the mutualism between figs and their pollinator hosts. For parasitic wasps, Wolbachia may help them escape from fig sanction.
This study shed light on the influence of different interspecific interactions on codiversification in fig microcosm, while also revealing its dependence on specific characteristics of those interactions.