Automated experimentation in ecological networks

Abstract

Background: In ecological networks, natural communities are studied from a complex systems perspective by representing interactions among species within them in the form of a graph, which is in turn analysed using mathematical tools. Topological features encountered in complex networks have been proved to provide the systems they represent with interesting attributes such as robustness and stability, which in ecological systems translates into the ability of communities to resist perturbations of different kinds. A focus of research in community ecology is on understanding the mechanisms by which these complex networks of interactions among species in a community arise. We employ an agent-based approach to model ecological processes operating at the species' interaction level for the study of the emergence of organisation in ecological networks.

Results: We have designed protocols of interaction among agents in a multi-agent system based on ecological processes occurring at the interaction level between species in plant-animal mutualistic communities. Interaction models for agents coordination thus engineered facilitate the emergence of network features such as those found in ecological networks of interacting species, in our artificial societies of agents.

Conclusions: Agent based models developed in this way facilitate the automation of the design an execution of simulation experiments that allow for the exploration of diverse behavioural mechanisms believed to be responsible for community organisation in ecological communities. This automated way of conducting experiments empowers the study of ecological networks by exploiting the expressive power of interaction models specification in agent systems.

Figure 1

Examples of mutualistic relationships in nature. From left to right: plant-pollinator interaction between a bee and a flower, plant-frugivore interaction between a bird and a fleshy fruit plant, and mutualistic association between fungi and algae (i.e. a lichen).

Figure 2

Example of a complex food web in a real community (grassland in the United Kingdom). Nodes and edges (in the graph) represent species and trophic interactions among them respectively. (Image produced with FoodWeb3D, written by R.J. Williams and provided by the Pacific Ecoinformatics and Computational Ecology Lab http://www.foodwebs.org).

Figure 3

Example of a mutualistic network of a real community. Nodes and edges represent species and their mutualistic relationships respectively. (Figure provided by P. Jordano. Reproduced with permission.)

Figure 4

Syntax of LCC dialogue framework.

Figure 5

Distributions of frequencies of nodes degrees in simulated networks. Frequency distribution of the mean (with its standard deviation) of the number of interactions per node in fifteen of our simulated networks.

Figure 6

Example of the output of a simulation run performed using the model described. A network of interactions between artificial species in a simulated ecosystem. Host and visitor agents are represented by green and red nodes respectively. The thickness of the arcs represent the relative strength of that interaction relative to other coming from the same host species, while numbers on them indicate the number of times that particular interaction has been observed.

Figure 7

Plots corresponding to the network displayed in figure 6. The plots show the frequency distribution of the number of interactions and the distribution of degrees among the nodes on the network.

Figure 8

Connectance versus NODF nestedness values in natural and simulated communities. The values plotted are the connectance vs the NODF nestedness indexes obtained from the natural (blue dots) and simulated (yellow dots) communities presented in table 1 (see text).

Figure 9

Network representation of community number 3 in our simulations as presented in table 1. Host and visitor agents are represented by green and red nodes respectively. The thickness of the arcs represent the relative strength of that interaction relative to other coming from the same host species, while numbers on them indicate the number of times that particular interaction has been observed.

Figure 10

Network representation of the OFLO natural community as introduced in table 1. Host and visitor agents are represented by green and red nodes respectively. The thickness of the arcs represent the relative strength of that interaction relative to other coming from the same host species, while numbers on them indicate the number of times that particular interaction has been observed.

Figure 11

Properties derived from the network of interactions in community number 3 in our simulations. The community network is displayed in figure 9.

Figure 12

Properties derived from the network of interactions in the OFLO natural community. The community network is displayed in figure 10.

Figure 13

Properties of the communities number 3 (Figure 11) and OFLO (Figure 12) overlaid. In this plot we can see Figures 11 and 12 overlaid on each other, which facilitates the comparison between the properties of the community number 3 and the OFLO community. The datasets presented in grey in each of the plots correspond to the OFLO natural community, while the plots shown in their original colour correspond to community number 3.

Figure 14

Ecologically inspired interaction model, written in LCC, for agents coordination in an artificial ecosystem. The "visitor" role is specified in this section of the protocol.

Figure 15

Ecologically inspired interaction model, written in LCC, for agents coordination in an artificial ecosystem. The "host" role is specified in this section of the protocol.