Table 1 Seven models for predicting weighted network structure at new field sites in a novel environment
From: Predicting the effect of habitat modification on networks of interacting species
Model | Description | Data requirement | Application |
|---|---|---|---|
Null | Biologically plausible interactions at a field site occur with the same frequency | Presence or absence of an interaction at a field site in the novel environment | Reference predictions that assume recorded interaction counts are uninformative |
Aggregate counts | Recorded interaction frequencies are informative at all other field sites without additional data processing | Weighted interaction networks from multiple field sites not in the novel environment | Reference predictions that assume recorded interaction counts have intrinsic predictive value |
Random encounter | Interaction frequencies are proportional to the product of host and parasitoid species’ abundances | Relative species abundance in the novel environment | Reference predictions for a minimally complex mechanistic model |
Alternative preferences | Species-level processes and other ecological mechanisms do not change between different environments | Relative species abundance in the novel environment and existing network data to derive a preference matrix | Predicting between similar habitat types |
Correlated preferences | Altered resource selectivity by parasitoid species (consumers) based on habitat complexity | Relative species abundance, an existing preference matrix and a known general pattern for reordering entries according to the level of habitat complexity in the novel environment | Predicting between different habitat types |
Specified preferences | New parasitoid species (consumer) foraging strategies in the novel environment | Relative species abundance, an existing preference matrix and a subset of network data from the novel environment on the interactions involved in new foraging strategies | Predicting between different habitat types |
Complete characterisation | Species behaviour is so complex that all interaction preferences must be individually characterised in the novel environment | Relative species abundance and weighted interaction networks from multiple field sites in the novel environment | Reference predictions for a maximally complex mechanistic model |
