You can use your data to calculate a large number of diversity indices such as Shannon H' or Simpson's D. While these indices are intended for data at the species level, they are on occasion applied to higher taxonomic levels and even to chemical data in some cases. Keep in mind that if your data is at the family level you will want to be careful about comparing your results to studies that use data to the species or morpho-species level.

It sounds like your interested in correlating measures of alpha diversity, but you might also consider testing how beta diversity of vegetation influences beta diversity of arthropods. Definitions of beta diversity abound and can be controversial, but you could start out comparing the Bray-Curtis dissimilarity index of both groups. This would give you a sense of how heterogeneity of vegetation biomass and type influences heterogeneity in arthropods.

When we talk about biodiversity is not just a number per area. We talk about a biological diversity, it means there are interconnected aspects of species as individual and composite communities in the system. There are three levels of biodiversity either genetic, species, or ecosystem; or you could use alpha, beta, or gamma diversity as point of views depending on your objectives in measuring biodiversity of plants and insects (as also mentioned by Scott Ferrenberg).

Correlated to the condition of clear-cut: its not clear, it is newly clear-cut or several months/years after clear-cut? However, you could investigate for what plant species that occupy and dominance at that stage and what insects that exist following the plant species colonization (Pioneer species after clear-cut). So, species, density, richness and diversity of both plants and insects are valuable indices. Furthermore you could also investigate the plant-insect relationship (predation, mutualism) to examine the importance of existing pioneer species diversity at that stage. Do not forget to investigate abiotic or physical factors that support those species diversity such as woody debris and litter density and decomposition stages; soil bulk density, texture, nutrient, moisture, and temperature; relative humidity and air temperature.

By over viewing to all levels/biophysical condition you will get complete data to support your objective in measuring biodiversity of plants and insect diversity in clear-cut forests.

I hope this will helpful, and good luck for your project

I find it useful to differencitate between "Indicator" and "Index". An Indicator is a measure with a unit such as "number of species per 100 sqm" or "number of carabid beetles per ..." (your abundance). An Index is computed on the bases of several indicators such as the Shannon Index, which is calculated from species richness and abundance). From your question I get the feeling that you are looking for an "aggregated biodiversity index" reflecting "total biodiversity" of your plots. I'm not sure how helpful such an Index can be because those indices are often difficult to interpret. If for example the index increases, there may still be some species groups which are in decline, their decline is just over-compensated by the other (increasing) groups.But maybe it's the groups in decline which are the ones which matter?

I would rather work with a set of indicators and select them based on the hypothesis you are investigating, as suggested by Sulistiyowati.

I agree with colleagues' suggestions above. If you have two datasets: e.g. clear-cut forests and not-clear-cut ones, you can use ANOVA or non parametric equivalents to test differences in diversity indices (such as Shannon, or species richness, or weighted species richness if endemism or conservation status is considered). That would be also interesting in stepping into functional diversity. I mean, you could distinguish among different guilds of species (e.g. phytophagus, omnivorous, predators for invertebrates) and estimate their diversity. I understood that you would need simple diversity comparison indices, so in that case you calculate alpha and beta diversity. But if you would like to see the indicator value of one taxon for another, then this is a surrogate approach and there are different methods and techniques to test the surrogate value of one taxonomic group for another, including congruence of species richness patterns or the complementary approach. If this is the case, you can have a look at a paper of mine [ Kati et al. 2004 in Cons Biol). Good luck.

The diversity should decrease and opportunistic species should achieve dominance where habitats are affected by increased disturbance. Considering each species separately together with its ecological role is significantly more informative than collective measures such as richness or diversity. I agree with Felix Herzog that we can't rely upon the diversity indices only. I have some data, for comparing between protected and unprotected areas, indicate that some species of ants, tenebrionids and scarabaeids are more abundant and rich in unprotected area than protected. While, carabid species are more rich and abundant in protected areas. I found that it is important to point out the ecological role of each species/functional group separately in relation to the habitat/microhabitat/nich.

Biodiversity indices (Species Richness, Evenness, Shannon, Simpson etc) are universal in the sense that you can use them in any habitat and with any group of living beings (animal or plant species), once you can construct community matrices. If you have many plots on gradients (or just transects or scatterred randomly (haphazardly) over the study site, you can use multivariate analyses (PCA, DCA, cluster analyisis). If you have not only community matrices but also some environmental data for each sampling site or plot (sky openness, soil tempereture, moisture etc) CCA will be very useful.

I second Scott on trying beta-diversity approaches too (mul;tivariate methods include them).

I would like to add to this interesting discussion a word of caution about indices like Shannon and Simpson that combine species richness and evenness. That is to say, more even communities score higher in diversity for the same species richness. But in my experience, the richest communities are frequently incredibly uneven - a small number of common species and a very long tail of rare species. Taking that perspective, I've always seen these classic diversity indices as somewhat off the mark. I suggest staying pretty close to the primary data (species abundance by site) and not getting too lost in the thicket of available methods. The way to do this is to keep returning to your question and make sure the analyses you select are in service of your question.

I also have to advise caution when using taxa above the species/morphospecies level ["I have data on plant and arthropod family richness, abundance, and biomass"]. The function of higher taxa is to organize information about biodiversity, but there are many equally good possible systems for this purpose. The one we use today is in part a product of human history, as well as evolutionary history. You can discover a species, or new data about a species; you can't really discover a genus or a family, you can only establish a new genus or family in order to classify a peculiar species.

Personally, I favor methods that extrapolate estimated species richness based on proportions of rare species (see Colwell's EstimateS program and documentation for a start if you're not familiar with this).

Are you really investigating beetle abundance against plant diversity, or is it diversity vs diversity?

Don't be put off by high diversity values in disturbed places. This is indeed often the case. Keep in mind that biodiversity is really two things: the number of species in a place, and the rate at which the community changes over space and time. So it may be that species in a rich (and disturbed) habitat are very widespread, perhaps because they do well in the ample disturbed habitat available. By the same token, an old Oregonian rainforest may have relatively few carabid species, but where else are they found?

1) number of plant species X square meter , usually random sampling plants in a 1meter x1meters square sampling plot ( collecting and identifcation must be done for erbaceos species too using dicotomic keys) than to relate plants and carabid beetles2) beetles collecting and identification (to genus level ) in the same square sapling plot . Finally just 3)do the ratio between the number of plants and the numeber of beeetles ( obviously collected in the same square plot) This can be a syntetic indicator. For plats and beeltes y can also use the jacard matrix to highlight and caculate differences in biodiversity ( jaccard matrix is largely used in plants ecology. Bye

Dear Collin

We proposed a new indice based on Bayesian statistics (SSP indice), in my mind much more efficient than all the others.

Luque C, Gers C, Lauga J, Mariano NA, Wink M and Legal L (2007) Analysis of forestry impacts and Biodiversity in two Pyrenean Forests through a Comparison of moth communities (Lepidoptera, Heterocera). Insect Science 14(4): 323-338

Check my RG profile you can download it.

But if you wish to compare diversity between Two or more sites...it is finally a beta diversity story. Then check the recent approach of some Portuguese colleagues:

Carvalho, J.C.; Cardoso, P.; Gomes, P. Determining the relative roles of species replacement and species richness differences in generating beta-diversity patterns. Glob. Ecol. Biogeogr. 2011, doi:10.1111/j.1466-8238.2011.00694.x.

We published an application of their theoretical calculation (which is the following story of our former paper of 2007). If you want to use it check well our Material and methods as we had to include few changes which were necessary in the script in "R".

Luque C, Legal L, Winterton P, Mariano N.A, Gers C (2011) Illustration of the Structure of Arthropod Assemblages (Collembola and Lepidoptera) in Different Forest Types: An Example in the French Pyrenees. Diversity 3(4):693-711.

This is also available to download here.

Hoping it will be helpful.

Cheers from Luc

Collin,

It might be worth having a look at the Renyi diversity profile as well. You can use the BiodiversityR (or vegan) package in R to calculate it. For more detailed information, please have a look at the link: http://licenciamento.ibama.gov.br/Cursos/Curso%20Junho%202013/Introducao_a_Estimativas_de_Biodiversidade_5_a_7_junho_2013/Livros/Bibliografia%20espec%C3%ADfica%20para%20R/Livro%20-%20Tree%20diversity%20analysis%20-%20BiodiversityR/Chapter%205-analysis%20of%20diversity.pdf

Totally agree with Gabor we used also Renyi diversity profile in our paper of 2007.

My suggestion is to avoid a two-step dimensionality reduction process based on diversity indices and analyze your data once using a multivariate regression process like DISTLM in PRIMER with the PERMANOVA add-in; PERMANOVA is produced by Marti Anderson, and is available as stand alone freeware I believe, direct from https://www.stat.auckland.ac.nz/~mja/

Why not also look at other facets of diversity such as phylogenetic and taxonomic diversity and distinctness or functional diversity? All are easily computed using freewares under R.

Regards

AF

Collin,

you are collecting a very broad array of opinions in this thread... with a little more focuse this discussion could be more useful. Would you tell us your specific goal?

Colin,

I suggest you also look at using some species estimation techniques to account for rarity profiles in your datasets, see Robert Colwell's excellent EstimateS freeware and the paper by Robert and Nick Gotelli , see: http://viceroy.eeb.uconn.edu/EstimateS/

However, I also echo Zaal's request for some mote detail, good thread!

regards

For a question like this, I think interaction diversity is the most relevant measure, if you are able to measure it. Species richness, evenness, relative abundance, or any other diversity measures are not as tightly linked to ecosystem function, stability, or other ecosystem parameters as interaction diversity. Here is a link to a paper on the topic:

http://wolfweb.unr.edu/~ldyer/dyeretal2010c.pdf

Also, I agree with many of the comments about rarefaction and caveats about using diversity indices, but I think Lou Jost has some of the best approaches to understanding and quantifying diversity metrics. Here is a link to one of his original papers on this:

http://wolfweb.unr.edu/~ldyer/classes/394/jost06.pdf