Aggregation of incomplete food web data may help to suggest sampling strategies

Aggregation of data and incomplete sampling are two notoriuos problems of food web research. We suggest to look at them in parallel since their effects are interdependent. Different aggregation methods are not equally sensitive to missing data and they lead to different biases in describing food web structure. In this paper, we construct a low-quality food web of Lake Balaton (based only on high-quality literature), aggregate it in several ways, compare the different versions of the food web by network analysis and discuss how the results can help future sampling, field work and data management. We identify groups where resolution or aggregation should be increased.