Date of Award
Master of Science
Soil Science; Environmental Science
Michael J. Castellano
Bradley A. Miller
Understanding the interactions of land-use management and landscape position on soil health is important in assessing how to measure soil health effectively. My goal was to optimize soil health assessments by evaluating: (i) which soil health metrics were most responsive to management, (ii) where differences were most easily observed on a landscape, and (iii) how many soil samples were needed to statistically detect the effects of management. Soil from two long-term research experiments in central and northeast Iowa were sampled from 0-15 centimeters, using a modified stratified sampling design based on management practice and landscape position.
My results indicated that soil respiration was a more responsive biological soil health metric than dissolved organic carbon, both in terms of differences between management practices and number of samples needed to detect these differences. Additionally, I found that landscape position and management effects on soil health were interdependent. Based on results presented herein, I recommend soil sampling at the relatively flat, stable hillslope positions, i.e., the summit, footslope, or toeslope. In Iowa, differences in soil health were likely to be more strongly observed on east facing slopes. In all cases, comparison of soil health metrics should be made on matching hillslope positions and aspects. Lastly, results of the power analysis suggested there was insufficient statistical power to detect significant differences in soil health without ample soil samples. I recommend measuring at least 100 soil samples per dataset, 25 per management practice, and 3 separate samples per plot.
Vizka, Elaine, "Detecting changes in soil health due to management in the context of landscape position" (2018). Graduate Theses and Dissertations. 16686.