This paper is the first publication of hopefully a series of studies regarding abiotic stress tolerance dynamics in (woody) plants. I began with a simple study that needed no fieldwork, just data from the internet. Ülo had previously published, together with Fernando Valladares, a great paper regarding Northern hemisphere woody species drought, shade and waterlogging tolerances (Niinemets & Valladares 2006 Ecol Monogr). They had cross-calibrated all the species- and stress-specific tolerance data for 806 woody species (about 40 % of the total species pool of Northern hemisphere woody plants). And found out that all the different stresses have significant negative relationship between them (Fig 1).
Fig 1. Figure 3 from Niinemets and Valladares 2006 Ecol Monogr paper.
These results were in accordance with the dominating model of Smith and Huston 1989/1990 (Fig 2), which draws a really solid trade-off line between shade and drought tolerance. This model is highly cited (currently holds 459 citations according to Google Scholar) and mostly cited as the theoretical basis of a case-study. But the actual results from actual species showed quite a lot of variability in such trade-offs – more or less half of the dots in Niinemets and Valladares 2006 results were on the wrong side of the line.
Fig 2. Panel B from Smith and Huston 1989/1990. Legend: “(b) Division of the continuum of woody plant strategies into the 15 discrete functional types used in the computer simulations. Each functional type is defined by maximum growth rate, shade tolerance, and tolerance to low moisture levels, as defined by the parameters in Table 1. The same labeling for these 15 functional types is used throughout the paper.” [N.B.Note that the axis are the other way around than in previous figure – here the tolerances are high near the zero.]
Valladares and Niinemets had proposed in a subsequent review paper published in 2008 (in Annual Rev in Ecol Evol and Syst) that the trade-off between abiotic stress factors, especially between shade and drought tolerance, depended on climatic gradient. They assumed that if the vegetation period is longer, it is more probable that different types of stresses do not occur at the same time and also that the plant has more time to deal with different kind of stresses. Let´s say there are floodings in every spring, so it would make sense for a tree to be still dormant during that period. And if the vegetation period is long enough, the plant can afford such strategy, because there is plenty of temporal resources available. It´s still possible for the tree to yield ripe seeds and so on.
But this had not been tested before. So by building on top of the Niinemets and Valladares 2006 paper, we decided to see if the shade-drought trade-off was affected by vegetation period length and whether woody species that usually have dormancy (angiosperms) differ in any way from woody species that are usually evergreen (gymnosperms).
As the stress tolerance data was species-, but not location-specific, we could not carry out spatially explicit analysis. We decided that the best species-specific indicator of vegetation period length would be species cold tolerance. Luckily United States Department of Agriculture (USDA) has been long working on Plant Hardiness Index and while species-specific data for cold hardiness is not directly available from USDA page (it used to be about 10 years ago), many gardening sites have cold tolerance information available for lots of species. I was able to extract the cold tolerance data for about 90% of 806 species from different gardening sites. For the rest of species I just compared USDA Plant hardiness zone map with species distribution map and concluded its cold tolerance, or extracted data from published literature. I flipped the hardiness scale and turned it into 5-point scale in order it to fit with our tolerance data (all our stress factors have a range from 1-low tolerance, to 5-high tolerance) and started testing.
And our results indeed show that both cold tolerance (indicator of vegetation period length) and growth form (indicator of dormancy mechanism) significantly affected plant abiotic stress tolerance. But all this is already better explained in the paper itself.
Although our results still contain quite a lot of noise (which we assume could be largely intraspecific variability), we still proved quite clearly that the strict tolerance trade-off model by Smith and Huston (see Fig 2 above) does not really hold. Another textbook knowledge rebutted. Well, it´s all in a day´s work for a macroecologist…
Finally, some rejecton stats. We had almost no trouble with publishing this manuscript. Of course, at first Ecology Letters rejected it without review. But GEB was our second choice journal and we got a major revision right away, with some really constructive comments from the reviewers. It all went rather smoothly and quickly from there on.
By Lauri Laanisto