Joseph Amsili, Harold van Es, and Bob Schindelbeck
Soil and Crop Section, School of Integrative Plant Science, Cornell University
- Soil health metrics were sensitive to long-term tillage practices.
- Biological indicators related to labile carbon and nitrogen were best correlated with crop yields.
Soil health has been defined as the “the capacity of the soil to function as a vital living ecosystem that sustains plants, animals, and humans” (NRCS). The soil health concept recognizes that biological, physical, and chemical constraints must all be addressed for the soil to reach its full potential. Soil health assessment has become a powerful tool to diagnose biological and physical constraints in addition to the more traditionally measured chemical limitations (nutrients and pH).
In 2017, Roper et al. published an article using Cornell University’s Comprehensive Analysis of Soil Health (CASH) approach to link soil health indicators to different agronomic management systems in North Carolina. The results from this experiment allowed us to explore the linkages between soil health and yield, which has remained an important gap in the soil health literature.
Soil samples were collected from three long-term experiments (20+ years) in the coastal plain, piedmont, and mountain physiographic regions of North Carolina. Samples were sent to Cornell University to be analyzed for the standard CASH package, which includes two physical indicators – wet aggregate stability (AgStab), and available water capacity (AWC); four biological indicators – organic matter (OM), active carbon (ActC), autoclavable citrate extractable protein (Protein), and respiration (Resp); and seven chemical measurements (pH, and extractable P, K, Mg, Fe, Mn, and Zn). The results were made available in a supplemental table in the Roper et al. paper and re-analyzed by us.
Results and Discussion
Sensitivity of Soil Health Measurements
The coastal plain and mountain region experiments involved different tillage practices combined with organic vs. conventional management. When the CASH values were analyzed, we found significant differences in AgStab, ActC, P, and Zn at each site (Table 1), while the mountain site also showed differences in Protein and several other chemical indicators. The effects on chemical indicators were primarily related to the organic vs. conventional treatments as compost applications tended to increase nutrient levels. But the physical (AgStab) and biological indicators (ActC and Protein) were mostly influenced by tillage practices.
The piedmont experiment involved different levels of tillage intensity including moldboard plowing, chisel-till and no-till, and almost all CASH indicators were affected by the treatments (Table 1). Over all three sites, no-till treatments had higher soil physical and biological indicators than conventional tillage (data not shown). These findings are in agreement with many other studies that have demonstrated that CASH indicators are sensitive to management.
Soil Health and Yield
Establishing positive relationships between soil health and crop yield is a difficult task because soil health effects are often masked by non-soil factors such as pest and weed pressures, erratic weather, and management effects. However, in the piedmont study, several individual CASH indicator values were positively correlated with mean corn and soybean yield (annual yield variability was not considered here). Specifically, eight CASH indicators (in order of correlation strength: Protein, ActC, Mn, Resp, AgStab, P, Mg, and OM) displayed significant linear relationships to mean corn yields (Figure 1). Each indicator tended to be highest under minimum-till, intermediate under chisel-till, and lowest under moldboard plow.
This implies that reducing tillage positively impacted these SH indicators, which in turn benefitted crop yield. Most notable is the fact that the indicators related to labile carbon and nitrogen (Protein, ActC, Resp, and AgStab) showed the strongest correlations with mean corn yield, while total organic matter content showed weaker relations (Figure 1). This suggests that organic matter quality may be more important for improving corn yield than total organic matter quantity. When we analyzed the relationship between soil health indicators and mean soybean yield at this site, we found that Protein was a much weaker predictor of soybean yield than it was for corn yield (not shown). This makes sense because a legume crop would be less influenced by nitrogen supply from soil organic nitrogen sources (Protein reflects the largest such pool) than a non-legume crop. Additionally, we found that Mn showed a strong positive relationship to both corn and soybean yield, which is explained by Mn’s important role in organic matter decomposition.
Figure 1. Linear regression of mean corn yield (bu/acre) as a function of different soil health indicators and tillage practices at the Piedmont site. Plots are organized from highest to lowest R2adj value. Error bars represent standard deviations associated with annual yield variability.
Soil health data from three long-term field experiments in North Carolina showed that the CASH indicators were sensitive to agricultural management at each location, which is in line with previous results from our work in New York. Furthermore, soil health indicators were positively correlated with long-term average cash crop yield at the piedmont site, which illustrates the benefit of labile organic matter for crop growth. These results help to demonstrate the linkage between tillage, organic additions, soil health, and crop yield, which had remained an elusive goal.
We acknowledge the work and funding of the original paper by Roper et al., 2017 (Soil Science Society of America Journal 81: 828-843).
This article was based on Harold van Es and Douglas Karlen’s article titled, “Reanalysis validates soil health indicator sensitivity and correlation with long-term crop yields” (van Es and Karlen, 2019; accepted in Soil Science Society of America Journal).