We
aim to develop a tool that generates pseudo-random species-richness
matrices, incorporating several well-studied patterns of biodiversity.
A
common assumption in ecological models is that species are uniformly
and randomly distributed. Under such an assumption, few if any of the
patterns we observe in biodiversity data would be apparent.
For instance, if species were randomly distributed, a nestedness matrix
would likely have an artificially high level of disorder (figure
bottom-left). In reality, species distributions follow a more nested
structure, as can be seen by Illinois woodlot birds (Blake, 1991)
(figure top-left). The consequences of such invalid assumptions are
borne out in other statistics (right side; see other figure for
description) and greatly affect the outcome of ecological models.
Generating pseudo-data that mimics many of the patterns is challenging:
- Patterns are not independent
- Data on which patterns are parameterized are highly variable in quality, quantity, and biology.
- Mechanisms underlying patterns are poorly understood.
- Many patterns are scale-dependent
- Gaston K. In Pattern and Processes in Macro-Ecology. 1st Edition. Place: Blackwell Science Ltd; 2000
- Manel
S, Williams H, Ormerod S, Evaluating presence-absence models in
ecology: the need to account for prevalence. Journal of Applied Ecology
2001, 38: 921 - 931.
- ReVelle C, Williams J, Boland J:
Counterpart models in facility location science and reserve selection
science. Environmental Modeling and Assessment 2002, 7:71 - 80.
- Wright D, Reeves J: On the meaning and measurement of nestedness of species assemblages. Oecologia 1992, 92: 416 - 428
