Sequentially assembled food webs and extremum principles in ecosystem ecology

Abstract

Summary1. Successional changes during sequential assembly of food webs were examined. This was carried out by numerical methods, drawing one species at a time from a species pool and obtaining the permanent (persistent) community emerging at each step. Interactions among species were based on some simple rules about body sizes of consumers and their prey, and community dynamics were described in terms of flows of biomass density. 2. Sequential assembly acted as a sieve on the communities, assembled communities having many properties different on average from those of feasible, stable communities taken at random from the species pools. 3. Time-series of community development were consistent with certain functions thought to go to an extremum (maximum or minimum) in ecosystem ecology, including a rapid early increase in net primary productivity and ascendency, although a clear trend in total biomass density was not evident and resilience decreased rather than increased. 4. In addition, more gradual changes in food web structure took place during succession to which the ecosystem goal functions were relatively insensitive. These changes included gradual increases in the number of species, invasion resistance, number of loops of length > 2 and number of prey species per consumer species. 5. We therefore argue that ecosystem and community dynamics can offer complementary insights into the process of ecological succession. The extremum principles of ecosystem ecology highlight some of the major properties of succession, whereas the community ecology sheds light on some more subtle changes taking place within the networks.

Increasing community size and connectance can increase stability in competitive communities

“…Various features of randomly drawn communities have been shown to differ to communities drawn using sequential assembly rules (Virgo et al, 2006), while Haydon (2000) has demonstrated the conditions required to produce maximally stable communities. However, 5 randomly assembled communities have proven a popular starting point when asking questions of community stability (Chen and Cohen, 2001;Cohen and Newman, 1985;Ives and Hughes, 2002;Ives and Carpenter, 2007;Ives et al, 1999;Ives et al, 2000;Jansen and Kokkoris, 2003;Rozdilsky and Stone, 2001) and can serve as a useful null hypothesis for comparison with communities assembled under different ecological and evolutionary rules.…”

Permanence Does Not Predict the Commonly Measured Food Web Structural Attributes

“…This work also complements previous permanent-assembly algorithms by increasing the set of food web attributes for which the effect of permanence is known. Virgo et al (2006) measured the value of thermodynamic goal functions arising from permanence, whereas in this work, the structural attributes that are commonly used to verify the ordered-niche models are investigated instead. …”

“…DOI: 10.1086/524953 step, according to some prespecified rules. There is already a large body of food web assembly literature (Tregonning and Roberts 1979;Post and Pimm 1983;Mithen and Lawton 1986;Taylor 1988;Drake 1990;Law and Blackford 1992;Luh and Pimm 1993;Law and Morton 1996;Lockwood et al 1997;Drossel et al 2001;Bastolla et al 2005;Virgo et al 2006) and a complimentary literature testing the predictions of the algorithms in experimental systems such as microcosms (Dickerson and Robinson 1986;Robinson and Dickerson 1987;Drake 1991;Wardle et al 1995;Moyle and Light 1996;Weatherby et al 1998;Chase 2003;Fukami 2004). Because assembly algorithms simulate invasion and extinction in a whole food web, they hold great promise for answering questions about succession, rehabilitation, and invasibility.…”

“…To isolate the effect of permanence from the more minimal assumptions that underlie it, three food webbuilding algorithms are created: one with feasibility and permanence as constraints, one with only feasibility as a constraint, and one with no dynamical constraint at all. The predictions of these algorithms are then compared to real empirical webs, and their predictive success is contrasted with that of the ordered-niche models.To date, all published food web-building algorithms using permanence as a constraint have also included physiological considerations (Law and Morton 1996;Virgo et al 2006), as one would expect when the objective is to make the model as realistic as possible. However, given that the objective in this article is to isolate the predictive value of permanence, no physiological considerations are included.…”

“…To date, all published food web-building algorithms using permanence as a constraint have also included physiological considerations (Law and Morton 1996;Virgo et al 2006), as one would expect when the objective is to make the model as realistic as possible. However, given that the objective in this article is to isolate the predictive value of permanence, no physiological considerations are included.…”