Eco-environmental gradients: importance & short-cuts

Gradients are powerful tools in plant community ecology. However, several recent papers fail to quantify, even coarsely, the differences between sites. This is likely to become an increasingly critical issue in publishing future studies of plant-plant interactions.

I discuss gradients conceptually in this book chapter:
Lortie, C.J. 2010. Synthetic analysis of the stress-gradient hypothesis. Positive Plant Interactions and Community Dynamics: edited by F. Pugnaire, pages 125-147.
Please contact me if you would like a copy of this chapter as a pdf.

However, things are evolving rapidly. Here is an incomplete list of steps/solutions that I have seen in the last 6 months.
Solutions/steps to estimating abiotic gradients in nurse-plant studies
1. Select gradient points that are large enough to ensure a buffer between next location or at least do work in the middle so as to ensure independence. Basic community ecology for sure, but it seems to be a rule of thumb out there. Independence appears to be well executed in sampling.
2. Set up 3 Onset hobo prov2 loggers under shrub and 3 in open per gradient point to get a mean estimate of microsite differences. Very limited scale given how big each gradient location is in many systems, but at least it is something quantitative and environmental.
3. Use a surrogate measure such as shrub density, RDM, biomass estimates for each location/gradient point. Higher productivity = less stress. Assumption, but ok.
4. Use downscaled data of some sort. WorldClim, Modis for NDVI etc.. but need to have resolution that matches gradient.
5. Use a handheld soil moisture probe to get a load of estimates at each location/gradient point under shrubs and in open.  Do at peak biomass/flowering at least but ideally do early, mid, and end of season?  Do after all biomass cooked too so as to estimate potential for annuals and shrubs to mediate local climate effects.  Also, I have seen handheld lightmeter datasets etc. using same design. Lot’s of single point estimates distributed in space but not time.
Here is a nice sample paper on how to aggregate data from loggers for these studies:

The aggregation solution is hourly then do contrasts diurnally, pairwise for every hour between the microsites.

Here are the figshare links to a few conceptual illustrations I use in exploring the semantics of gradients.

Fig 4