Decoupling of impact factors reveals the response of German winter wheat yields to climatic changes

Abstract

Yield development of agricultural crops over time is not merely the result of genetic and agronomic factors, but also the outcome of a complex interaction between climatic and site-specific soil conditions. However, the influence of past climatic changes on yield trends remains unclear, particularly under consideration of different soil conditions. In this study, we determine the effects of single agrometeorological factors on the evolution of German winter wheat yields between 1958 and 2015 from 298 published nitrogen (N)-fertilization experiments. For this purpose, we separate climatic from genetic and agronomic yield effects using linear mixed effect models and estimate the climatic influence based on a coefficient of determination for these models. We found earlier occurrence of wheat growth stages, and shortened development phases except for the phase of stem elongation. Agrometeorological factors are defined as climate covariates related to the growth of winter wheat. Our results indicate a general and strong effect of agroclimatic changes on yield development, in particular due to increasing mean temperatures and heat stress events during the grain-filling period. Except for heat stress days with more than 31°C, yields at sites with higher yield potential were less prone to adverse weather effects than at sites with lower yield potential. Our data furthermore reveal that a potential yield levelling, as found for many West-European countries, predominantly occurred at sites with relatively low yield potential and about one decade earlier (mid-1980s) compared to averaged yield data for the whole of Germany. Interestingly, effects related to high precipitation events were less relevant than temperature-related effects and became relevant particularly during the vegetative growth phase. Overall, this study emphasizes the sensitivity of yield productivity to past climatic conditions, under consideration of regional differences, and underlines the necessity of finding adaptation strategies for food production under ongoing and expected climate change. K E Y W O R D Sclimate change impact, climate trend, long-term yield development, phenology trend, R 2 for mixed effect models, soil yield potential, weather extremes, Winter wheat

Crop adaptation to climate change as a consequence of long-term breeding

“…In the face of climatic changes, it is expected that total annual precipitation levels might not necessarily change significantly; however, the frequency and distribution of rainfall is expected to change. In case of phenologically adapted crops, this can result in a situation where the timing of water supply can be unfavourably related to the needs of the plant, for example by a shift towards off-season rain events (Bönecke et al 2020 ). Furthermore, even a small increase in spring temperatures may lead to earlier and faster growth of crops, causing them to consume more water earlier in earlier developmental stages and potentially run out of water in early summer (Lian et al 2020 ).…”

Methods of yield stability analysis in long-term field experiments. A review

“…Classical detrending techniques that have been applied to crop yield data can be grouped into two main categories: (1) approaches that use time and (2) approaches that use time and one or more additional variables to separate trends from variability in the time series. An example for the latter approach is the study by Bönecke et al (2020) who separated climatic from genetic and agronomic yield effects using linear mixed effect models and estimated the climatic influence based on a coefficient of determination. Mathematically related to both approaches, though often a rather technical consideration, is the transformation of the data (e.g.…”

“…Corresponding statistical approaches, such as mixed effect models, allow for explicitly assessing the contribution of factors, such as crop genotype, fertilizer amount and type, and site characteristics (environment) to the overall variability (see, e.g. Bönecke et al 2020).…”

“…Over the past decades, many different regression-and variance-based measures of yield stability have been proposed (Becker 1981;Becker and Léon 1988;Dehghani et al 2008;Eberhart and Russell 1966;Eghball and Power 1995;Huehn 1990;Hussein et al 2000;Lin et al 1986;Piepho 1998;Francis and Kannenberg 1978;Lin and Binns 1988;Plaisted and Peterson 1959;Sneller et al 1997;Tai 1971;Plaisted 1960;Jensen 1976;Wanjari et al 2004;Kang 1988; Rothamsted Research in the UK to compare various combinations of inorganic fertilizers and organic manures on the yield of winter wheat (Picture: Janna Macholdt/JLU Gießen) Kataoka 1963;Eskridge 1990;Fox and Rosielle 1982;Abou-El-Fittouh et al 1969;Cotes et al 2006;Perkins and Jinks 1968;Freeman and Perkins 1971;Hernandez et al 1993;Sadras and Bongiovanni 2004;Heath 2006;Fernández-Martínez et al 2018;Bacsi and Hollósy 2019). There is also the GGE (genotype plus genotype-by-environment) biplot analysis available, e.g.…”

Machine Learning With GA Optimization to Model the Agricultural Soil-Landscape of Germany: An Approach Involving Soil Functional Types With Their Multivariate Parameter Distributions Along the Depth Profile

“…More than 25% of the respective measurement sites report average values above the threshold of 50 mg L −1 (Jakobs et al, 2020). Overall, economic and environmental risk is site-specific and depends on soil characteristics (Bönecke et al, 2020;Webber et al, 2020). Hence, the estimation of environmental impact and vulnerability of the farmer's income as well as the development of adequate agricultural management strategies, policies, and farmers' subsidies require site-specific soil information at national scale.…”