Assessment of multiple- and single-factor stress impacts on corn

Abstract

Impacts of abiotic stress on growth and development of corn ( Zea mays L.) plants have typically been investigated in isolation from influences of other stress factors. This approach usually results in both lost information and overly simplistic conclusions concerning the effects of single stress factors on the physiological response of corn. In reality, the impact of an individual stress on corn growth depends on the intensity and possible interactions among other stress factors. A superior approach for investigating the effect of a single stress would consider the effect of other stress factors using multivariable modeling. We used a mixed-model framework to estimate the relative impacts of soil physical stresses for a multiple-location-and-year experiment involving the response of corn to a range of tillage systems from no-till to conventional fall tillage. Covariates were treated as fixed effects in a multivariable mixed model; effects due to location, year, replication, and their interactions were treated as random effects. Relationships among soil physical factors that limited corn root and shoot growth in the field were identified. We present the underlying analysis in detail, along with several examples where interactions were identified among soil physical stress factors and early corn growth in specific tillage systems. In an example used to illustrate the model, high soil water contents (>0.30 m 3 m −3) during the first 4 weeks after seeding were more detrimental to no-till corn root and shoot development at low (19 °C) mean soil temperatures. In contrast, high soil water contents benefited no-till corn root and shoot growth at warm (23 °C) soil temperatures. Therefore, when corn seedlings encountered multiple stresses of low soil temperatures and high water contents, reductions in corn root and shoot growth were more likely to occur in no-till relative to conventional tillage. Future assessment of multiple stress impacts on corn development should employ multivariable statistical models more routinely to improve our understanding of interactions among the various stresses and also to help guide efforts in stress amelioration.