Organisms that inhabit streams and rivers experience a high degree of spatial and temporal environmental variability overlaying a dendritic habitat geometry.  I have previously explored methods for predicting the responses of populations to spatial and temporal variability at a variety of scales using methods that characterize how the input of environmental variation results in the output of population variation.  These methods are derived using functional relationships between environmental variables (e.g. flow, productivity, recruitment sites) and demographic and dispersal rates that can be quantified from small-scale experiments.  My previous efforts have focused on dynamics in simple one dimensional stream geometries, yet the dendritic geometry typifying most streams and rivers could potentially complicate the application of simple modeling results to real systems.   

Results/Conclusions

Here, I explore the ability of previously described methods for modeling relationships between environmental variables and population dynamics to predict such relationships derived using empirical methods in dendritic networks.  I show through simulations of spatially explicit population dynamics in hypothetical dendritic riverscapes how connectivity among branches can at times obscure relationships estimated using standard analyses as such spatial cross-correlation or variogram calculations, especially when organisms disperse frequently.  However, I show that analyses of simple population models can provide predicted spatial cross-correlations and the degree that connectivity and dispersal influence these predictions.  I conclude by examining data requirements for estimating the effects of spatial and temporal variability on population dynamics in dendritic networks which, while potentially extensive, are not unreasonable given the scope of sampling typically employed by biologists and river managers.