How to correctly assess the impact of invasive species on a community using the diversity index?

Question

How can the impact of invasive species on a community be correctly assessed using the diversity index if the abundance of the target species, i.e. the invasive species, is not used to calculate the index?

We investigated the impact of two invasive alien plant species (goldenrod S. canadensis and S. gigantea) on plant diversity in semi-natural grasslands. Plant diversity is expressed in terms of Hill diversities (Hill species richness, Hill-Shannon diversity and Hill-Simpson diversity). We wanted to investigate whether the invasion of the perennial goldenrod species affects the species diversity of three plant guilds - herbs, graminoids and bryophytes. Therefore, we calculated Hill diversity indices separately for each guild. We excluded Solidago species from the species richness estimates for forbs (Solidago is a forb), as we wanted to assess the impact of invaders on the number of remaining species. However, for the Hill-Shannon and Hill-Simpson indices of herb diversity, Solidago was included because it would be meaningless to calculate these two indices without considering the abundance of Solidago in the invaded plots. The Hill-Shannon and Hill-Simpson diversity indices of graminoids and bryophytes do not "take into account" the abundance of Solidago, but an effect of Solidago invasion on the diversity of graminoids and bryophytes can be expected.

We found that invasion by Solidago spp. had the most pronounced negative effect on forb species richness and diversity. On the other hand, graminoid species richness and diversity were not affected by Solidago invasion; a similar pattern was observed for bryophytes. While the Hill-Shannon and Hill-Simpson diversity indices of graminoids and bryophytes are mathematically sound, I think they are not (bio/eco)logically sound because they do not "take into account" the abundance of Solidago. (Or am I wrong and this approach is OK?)

What is your opinion? How can the effect of Solidago on the Hill-Shannon and Hill-Simpson diversity indices of graminoids and bryophytes be assessed in a logical way? Or is it just a matter of interpretation?

I was thinking of something like taking Solidago into account when calculating diversity indices in the same way as covariates are used in a statistical model. But I don't know how to implement this...

Answer 1

My best recommendation would be to consider what each measure is actually telling you, and only using that measure for an appropriate question. Abundance and diversity are different things. As you recognize, a diversity index can tell you, for example, how the mix of grasses changes, but not whether grasses overall increase or decrease. If the question of grass abundance is important, then just measure that directly.

When faced with analytical decisions, I would go back and consider the core question of your research and what is motivating your analytical decisions. Why not calculate overall species diversity? What would it mean to include or exclude Solidago from that calculation? Does it make sense to include the invasive as part of diversity? What about the other non-native species in your plots?

You will likely need multiple observations to describe the important aspects of your results. You might describe the outcome by saying, "Invasion by Solidago had little effect on the mixture of grasses present, but reduced overall grass abundance by X%." You would then use the index statistic and the abundance statistic to support that claim.