Water stress in Montana cropping systems: effects of cultivar, management, and environment on crop production in dryland systems

Abstract

Crop productivity--defined as yield, protein, and economic returns--hinge on crop water use. Crop water use is a function of genetic, environment, and management factors. This thesis addresses how these factors interact with crop water use and productivity in Montana. In chapter 2, a two-year (2014-2015) study compared winter wheat yield and protein following fallow and three intensive sequences on deep and shallow soils. Water extraction was measured on deep soils, and kriged soil depth estimates served as a surrogate for stored soil moisture on shallow soils. On deep soils, yields ranged from 72-84% of fallow-wheat from 20.5 mm less water extracted below 45 cm, while protein was ~0.63% greater in intensified sequences. On shallow soils, sequence did not affect yield or protein. Yields increased with soil depth while protein decreased in 2014, but no trends were observed in 2015 due to 47 mm greater precipitation from joint to heading. Intensive sequences diminish wheat productivity on deep soils, whereas soil depth and precipitation timing control productivity on shallow soils. In chapter 3, state-wide cultivar testing, soils, and climate data was used to quantify four general drought patterns in winter wheat and five in pea. Cultivar had little impact on yield compared to drought pattern with winter wheat yields ranging from 4421 kg ha -1 to 2539 kg ha -1 and pea yields ranging from 2877 kg ha -1 to 975 kg ha -1. Yields negatively correlated with drought intensity at heading in wheat (r 2=-0.79) and flowering in pea (r 2=-0.76). Quantifying drought patterns provides a physical interpretation to improve management and breeding efforts. In chapter 4, yield-evapotranspiration (ET) functions were derived for spring wheat, pea, and chickpea from a three-year (2002-2004) seeding date trial. Yield-ET functions were coupled with ten-year (2005-2015) climate records to predict yields at four staggered seeding dates. Yield predictions were converted to marginal revenues based on high, medium, and low markets and fixed production costs. Across seeding dates and markets, simulated returns were highest for chickpea (~601 $ ha -1) followed by wheat (372 $ ha -1) and pea (202 $ ha -1). This indicates chickpea should be seeded before wheat and pea.