Networks which are aggregated based on trophic similarity and taxonomic groups will be affected by aggregation and will not give an accurate description of the networks. We have to be careful to use aggregated networks when the focus of a study is to analyse differences between species or when comparing networks with different levels of resolution.

 

Comparisons between networks with different levels of aggregation can be a problem and give inaccurate conclusions. If two networks with different level of resolution are used in a study, the recommendation is to aggregate the better-resolved network to match the poorly-resolved web. This would give two networks with the same level of resolution. Comparisons between these network would be more accurate and differences between them can be detected. The networks would be different from the original network, however, it would still be possible to determine if the networks are structurally similar and if one of them have more loops than the other.

 

The effects of the aggregation are not as severe when the aggregation is based on spatial scale. There is less need for caution when comparing networks covering different spatial scales. This can, however, differ for networks with more variation over spatial scales.