For the most part research into coevolution in ecological networks has studied the network in isolation. However, in nature the tangled bank analogy rules and communities involve mutualistic networks that are intricately interwoven into trophic interactions and other antagonistic networks as well. Thus an important goal for the network-based approach to study coevolution should be to investigate the interactions between the forces shaping each of these networks. The persistence of mutualistic interactions, for example, may be influenced by trophic complexity (Hoeksema and Bruna 2000). Research on integration of multiple networks could be facilitated through the relatively recent interest in community genetics. Specifically, studies that examine the feedback between ecological and evolutionary dynamics could demonstrate how organisms affect their environment (the community) and are in turn affected by the environment (Post and Palkovacs 2009).
Additionally, a better understanding of the influence of phylogeny on the development of ecological networks is vital. A phylogenetic perspective helps to clarify the “forbidden links” within a system and thus is essential to determine how the structure of the network grows and changes. Phylogenetic community assembly has shown ecologists the importance of species relationships in the development and structure of communities over time (Emerson and Gillespie 2008). Phylogenetic analyses could also inform ecologists about the likelihood of trait convergence on the core of generalist species. If species were assimilating into the network through convergence, then we might expect the trait to be phylogenetically overdispersed rather than clustered (Emerson and Gillespie 2008).
More importantly, empirical research into coevolutionary mechanisms within ecological network is severely deficient. The majority of work currently being done to understand the patterns of networks and the processes shaping them is done through simulation and examination of well-described webs. For example, Bascompte and colleagues (2003) used 27 plant-frugivore and 25 plant-pollinator networks in their analysis of nestedness in mutualistic networks. Thus, they were only able to describe the pattern of the various networks, and provide little support as to the validity of complementarity and convergence as a mechanism. In part this lack of empirical work on the subject can be attributed to the sheer complexity of the majority of ecological networks out there. Furthermore, there is some disparity among those individuals who are doing empirical work on the subject as to whether or not coevolution is truly diffuse (Hougen-Eitzman and Rausher 1994, Rausher 1996).
In order to clarify the mechanisms underlying the development of structure, as well as the implications of structure on coevolution more empirical work must be done on both mutualistic and antagonistic webs. To this extent it is important to utilize and understand the implications of motifs within networks (Bascompte 2009). Motifs, being small groupings of species interactions that repeat throughout webs, should provide excellent study systems. Through a full understanding of how motifs can be generalized (if they can be) could be instrumental to the development of a stronger body of coevolutionary theory. Currently those studies that do empirically examine coevolution often do utilize motifs. For example, Lapchin and Guillemaud (2005) investigated asymmetric diffuse coevolution in a host parasitoid system. Their study used a commonly seen motif in trophic networks, the tritrophic food chain of plant-aphid-aphid parasitoid. Studies such as theirs in combination with a better understanding of how these motifs can be applied to the network as a whole and knowledge of the structure of the network could be invaluable to the development of coevolutionary theory.