You could use ordination or multivariate analysis which test ecological gradients. But I think you need to compile a list of questions you want to answer, then the analyses required will become clearer :-)

Hello Andrew,

Thanks for your response. I have been suspecting multivariate analyses will be employed. 

I would like to know:

What compositional and structural characteristics of vegetation promote the greatest species diversity and abundance of small mammals at the reserve?

The same question should be asked in respect to breeding birds present in the summer.

Lastly, I would like to know whether there is an interaction between small mammal and breeding bird species diversity and abundance.

Let me know if I need to clarify anything. Thank you so much!!.

Non-Metric Multidimensional Scaling could be a starting point. It is basically a plotting tool but it will help you get an idea of the structure of your data. By the same token, regression trees are also very useful: http://onlinelibrary.wiley.com/doi/10.1890/0012-9658(2002)083[1105:MRTANT]2.0.CO;2/abstract

I hope this helps

Exploration of your data is very important. The questions you have asked are multilayered. Just to approach one aspect of the analysis, consider your vegetation structure data. You will need to assess if the variables you collected are independent measures. You may have levels of autocorrelation that will bias the results. This can be achieved relatively easily through the use of linear and non-linear correlations. Then you have to ask whether the data described more than one vegetation community. Ordination or even clustering would provide some quick answers. More than one community, in some ways, simplifies the analysis as you compare against a community type rather than specific individual elements. 

My own experience is that the overall vegetation structure is more of a measure than any specific component within that structure. Comparisons of biotic assemblages with communities tends to provide more of an explanation in most cases.

You also have to take into account that avifauna and mammal fauna also work on different scales within the landscape. The comparison at the broader level may be more useful for a wide ranging bird species. Small mammals with restricted home ranges are best dealt at a smaller more detailed level in the landscape.

Now that you have formulated some questions, I would revisit them and try and break them down into more specific concepts to be investigated, i.e., do you consider individual species, trophic groups or niches etc.?; what is the role of seasonality?; is rarity an influencer in your data?. There is definitely not one test that will do all; you will need to build up the story using building blocks.

Hi Niko,

See: European Journal of Wildlife Research February 2012, Volume 58, Issue 1, pp 59–70 and Bird Study Volume 53, 2006 - Issue 2 Pages 112-125 for two examples of studies of relationships between species abundance and the composition and structure of vegetation.

Andrew :-)

As others said, multivariate analysis would be a good starting point. There are books about this theme (I recommend Numerical Ecology from Legendre & Legendre). If you use R, there are several tutorials on multivariate analysis for community data that may help.

Thank you all for the responses! I am still a bit hung up on how to organize my data to analyze them. I have large sets of data that do not necessarily correlate. What I am interested in doing specifically is to analyze the distributions of my data and determine whether there are any associations. For instance, I have many samples of vegetation variables from a plot that I have also collected small mammal abundance data in. The sample size for my small mammal data in each plot is 8 throughout the year of sampling. However, I have 48 samples for each vegetation variables in the same plot. I would like to see if there is some relationship between the distribution of abundance in small mammals and the distribution in any of various vegetation variables (e.g. litter height, basal cover of forbs, basal cover of grass). I do not know how to tabulate the data for analysis in software due to the differences in sample sizes. I also have environmental data that could serve as another variable to analyze. Is there some sort of statistical method that tests for associations between distributions of independent variable data with that of a dependent variable such as abundance?

You should first define your unit of analysis. If you are interested in animal abundance, it seems that the unit would be the plot. However, taking many samples from the same plot would typically lead to non-independent units or pseudoreplicates. You can average these values, or go through more complicated models such as generalized linear mixed models that allow using your nested design. When you have many samples from the same unit, there is no problem to tabulate data. These would like something like

plot  sp1 sp2 sp3     veg.sample    height

1        3      5      0             1                   2.5

1        3      5      0             2                   1.0

1        3      5       0            3                  1.5

2        1      2        1            1                   0.8

2         1     2       1           2                    1.7

2         1      2      1           3                    2.3

which is called "long format". This format is the admitted one for most statistical softwares.

Good luck!

 Facundo,

My apologies. I oversimplified my study site. I have four land units differing in management history, and thus differing in expected vegetative communities. Within each of the four units, I have two sampling plots where I have collected abundance data during each season. Each of the two plots have abundance data collected each season throughout the year, and are pooled together to create a sample size of 8 abundance points for each land unit. 

Vegetation variables such as litter depth and others were collected within each plot with quadrat sampling. Six quadrats were sampled in each plot, each season, totaling to 48 quadrat samples per year, per land unit.