A

Figure 7.4 Three types of model for complex networks. (a) A scale-free model, characterized by a few, highly connected nodes; (b) a modular network, consisting of four highly interlinked units, connected to each other by a few links; and (c) an hierarchical network, which combines a scale-free topology with hierarchical clustering. Reprinted with permission from Ravasz et al. (2002). Copyright 2002 AAAS.

Figure 7.4 Three types of model for complex networks. (a) A scale-free model, characterized by a few, highly connected nodes; (b) a modular network, consisting of four highly interlinked units, connected to each other by a few links; and (c) an hierarchical network, which combines a scale-free topology with hierarchical clustering. Reprinted with permission from Ravasz et al. (2002). Copyright 2002 AAAS.

Network analysis is not uncommon in community ecology. Food webs can be considered as networks and ecologists are interested in what properties of food webs contribute to their stability. One such property, discussed by Neutel et al. (2002), is the pattern of interaction strengths across the web. Interaction strength of a trophic couple is a measure of the influence of one species on the population increase of another species when both are near equilibrium. The negative effect of a predator on its prey is usually larger than the positive effect of a prey on its predator; therefore, interactions strengths are greater for top-down relationships than for bottom-up relationships. Interaction strengths are particularly important when there are loops in the web; that is, paths that return on the same species without visiting other species more than once. The mean interaction strength of all the pairs in a loop is called the loop weight. Examination of 104 published food-webs revealed that the longer the loop, the lower its loop weight. Mathematical analysis showed that this property contributes greatly to the stability of the network. So, loop weight in a food web can be considered a network motif of a significance comparable to the ones identified by Zhang et al. (2005) for metabolic networks in the cell.

We believe that approaches similar in spirit to systems biology should ultimately be adopted to enable genomic answers to ecological questions. That is, a systems approach is needed to link genomics to ecosystem function, to life-history pattern, and to the ecological niche. Chapters 4, 5, and 6 of this book have shown that such links are currently far from complete. When ecological processes are governed by a limited number of signal transduction pathways, as in some of the stress responses discussed in Chapter 6, a network analysis of interactions seems to be a suitable option (see the figures in Chapter 6 showing induction of gene expression by stress). Also, in the cases of nutrient cycles catalysed by well-characterized genetic complexes in microorganisms, a systems approach seems to be feasible. However, in general it is difficult to forecast which type of systems biology is required as an integrative framework for ecological genomics. In addition, we see three main issues that lack in the present systems-biology approaches but which are nevertheless crucial for ecology.

Spatial considerations. Space is a very important aspect of ecological analysis. Many processes in communities require some degree of proximity between different organisms (e.g. in the case of syntrophy). Reproduction in animals usually requires physical contact between the sexes; in flowering plants pollen has to travel from anther to pistil and the distance between these organs matters. Stratification of the environment, or other types of heterogeneity, are often crucial for species to coexist with each other. Colonization events can alter the functioning of ecosystems dramatically if the colonizer is an invader and outcompetes local species. All these issues, so obvious for an ecologist, are completely absent from present-day systems biology. It is unclear how genomics and systems biology can contribute to spatial ecology.

Temporal considerations. Like space, time is another important dimension of ecological analysis. Any ecosystem bears all kinds of traces from its prior development; the way in which an ecosystem functions is determined partly by what went before. Historical issues are, for example, reflected in the build-up of the soil profile and the presence of peat accumulated from previous plant growth. If birds for some reason in the past decided to use one piece of land rather than another, this situation may continue for a long time because habitat use is often culturally transmitted or imprinted in the offspring. So, ecologists are accustomed to the fact that some phenomena in nature can only be understood by referring to past events, and that there is a more or less predictable succession of events during the development of an ecosystem. It is unclear how such temporal phenomena can be reconciled with a systems-biology perspective.

Bi-directional interaction with the environment. In the systems-biology treatment of the living cell, the environment of the cell is considered as given; expressed in ecological terms, it is a set of external conditions that are not altered by activities of the cell. In real ecological systems, the environment is altered significantly by organisms, for example by gas exchange, by consuming resources, or by altering physical structure (e.g. burrowing by earthworms). Including such interactions in a systems-biology model would imply that resources and nutrients are considered dynamic variables that can be depleted or otherwise altered by the organism. Such bi-directional interactions are not yet part of systems-biology analyses.

0 0

Post a comment