Quantifying the extent of environmental variability along the JFL gradient

We have soil data!  What a glorious day it is! Anytime you actualize a goal, it's a good feeling. I want to thank you - my supporters, who have made the acquisition of these data possible. It was not in the initial plan or budget to include soil data in this study, but when I got to Jianfegling, I quickly realized it would be necessity to our study question. More on that later. I also want to acknowledge Dr. Han Xu, from the Chinese Academy of Forestry, my main collaborator in China, and Ms. Xizhen Zhu, a technician at the Guangzhou Xinhua Agricultural Development limited company. Without their help, we would not have these data.

We shipped over a hundred pounds of soil to Ms. Zhu for soil analyses (see image below). She worked very diligently and efficiently to analyze 17 soil variables for each of the 300 samples. The variables measured were: soil texture (percent small [<0.002 mm diam.] medium [ 0.005 - 0.02 mm diam. ] , and large [0.05 - 2.0 mm] soil grain size); soil pH; organic matter content; total Nitrogen (N); total Phosphorus (P); alkali-hydrolyzable N; available P; available Potassium (K); exchangeable Sodium (Na); Calcium (Ca), and Magnesium (Mg); total exchangeable bases (TEB), and base saturation (BS). These data help characterize the nature of the environmental gradient that we sampled.

 

7 large bags containing 300 samples were sent to the lab for analysis. Thank you for your support!

Let me take a moment to refresh your memory of our research objective. We are interested in understanding how plants live and function across an environmental gradient: a fundamental question in plant ecology. To do this we measured root and leaf functional traits - morphological metrics of the physiological behavior - for 300 saplings along a 6.6 km gradient, spanning two peaks, in subtropical wet forest in Jianfengling (JFL), Hainan Island. Approximately the first half of the gradient is in secondary forest of about 60-80 years of age that has been selectively logged, and the second half of the gradient is untouched, 'primary' forest (the fifth flag on the below figures shows the boundary between the two areas; right after the descent of the first mountain). These forest type differences are clear, but we still needed the soil data - to quantitatively understand the environmental heterogeneity encompassed along the gradient.

 

Satellite view of the JFL transect

 

Topographic profile of the JFL transect.

This last week I revived the soil data matrix from Ms. Zhu. We can boil down the 17 measured soil variables into one graphic using multivariate statistics. A Principle Components Analysis can reduce many possibly correlated variables into a few orthogonal (statistically-independent) dimensions of variation. A PCA of the soil data matrix showed that the first two axes account for over one-third of the total variability in the soil conditions alongthe transect. Axis 1 illustrates differences in texture and maturation of the soil matrix attributed to forest age, and is positively-related to soil pH (r = 0.78) soil base saturation (r = 0.79) and large grain size (r = 0.66) and negatively related to soil organic matter (r = -0.62). Axis 2 represents a the range of soil fertility along the gradient, and is positively-related to total N content (r = 0.68), alkali-hydrolyzable N (r = 0.66), total P (r = 0.65), and a finer soil texture (large grain size r = -0.29). You can see that there is some separation between sites with respect to forest type (especially toward the extremes of the axes), but also that there is a large amount of overlap between them.

 

Principle Components Analysis of 17 soils variables from 300 samples taken at the base of sample individuals along the JFL gradient.

These results are consistent with what we might expect for soils along such a gradient. Generally, soil fertility decreases, the amount of organic matter increases, and soil texture becomes finer with forest succession. As the forest matures, the plants get larger and competition for soil nutrients increases. Accordingly, total root biomass increases, as does above-ground biomass. But as competition increases, the relative investment in fine root production should increase, unless the plants utilize other pathways of nutrient foraging (mycorrhizae, etc.). It will be interesting to see what we find as we continue to analyze the functional trait data in light of these findings. We are also doing leaf and root nutrient analyses, so much more to come from this project.

Again, thanks for your support and interest. All the very best, Aaron