Theoretical and experimental studies of community organization have revealed some intriguing phenomena with implications for the development of stable perennial grain polycultures. Such studies have shown that complexity is often the consequence of specific events that occur during community assembly — events that are generally undetectable from an examination of extant community patterns alone (Post and Pimm, 1983; Gilpin, 1987; Robinson and Dickerson, 1987; Drake, 1991; Drake et al., 1993). Complex stable communities in nature are not created as is; they assemble by the addition of species through invasion and by the loss of species through local extinction. Species that form important intermediate states must be included within the appropriate time frame. The historical sequence in which species are added to and removed from a community determines its final structure.
Community stability changes as species are progressively added to or subtracted from model food webs (Tregonning and Roberts, 1979; Roberts and Tregonning, 1981). A principal result of this work is that most "new" complex systems are unstable; stability comes with age. Drake (1990a) found that with unlimited invasion opportunities communities eventually reach persistent species compositions where none of the species "outside" the community can enter them. In addition, there may be so-called Humpty-Dumpty effects (Drake, 1990b; Luh and Pimm, 1993), meaning it may be impossible to reassemble a community solely from the species present in the final, persistent state. This is because species present early in assembly that were important in determining the final structure may not be present in the final composition.
Such results suggest that mimicking the history of a natural community is a powerful tool for constructing stable perennial polycultures. Persistent, diverse grain fields may be created more efficiently by starting with many more than the desired number of species, and allowing the community to "shake down" or "collapse," than by experimenting with many mixes of the desired numbers of species. If this prediction is correct, it will confirm an important role for history in constructing stable, complex communities.
In 1994, the Land Institute began a study to explore whether plant biodiversity and the sequence of species introductions can contribute to the assembly of perennial grain polycultures. Treatments consist of four incrementally diverse mixtures of herbaceous perennial species that represent the C4 grasses, C3 grasses, nitrogen-fixing legumes, and composites. Several of the species are potential perennial grains. The initial seed mixtures comprise 4, 8, 12, and 16 grassland species (Table 2). None of the species was present at the site prior to the experiment. Each lower diversity treatment was nested within its higher diversity counterpart (see Naeem et al., 1994). The treatments vary the size of the species pool, while keeping guild representation constant.
Was this article helpful?