David Makowski

Lack of evidence for a decrease in synthetic pesticide use on the main arable crops in France

The frequent, widespread use of pesticides in agriculture adversely affects biodiversity, human health, and water quality. In 2008, the French government adopted an environmental policy plan, Ecophyto 2018, to halve pesticide use within 10years. Trends in synthetic pesticide sales and use in France were described, through three different indicators: the number of unit doses (NUD), the quantity of active ingredient (QAI), and the treatment frequency index (TFI). Changes in pesticide use on seven of the principal arable crops in France since the implementation of this policy plan were analyzed, together with the impact of changes in pesticide use on water quality. No evidence was found for a decrease in pesticide sales at national level between 2008 and 2013. In terms of the TFI values for individual crops, the only decrease in pesticide use observed since 2001 was for soft wheat. This decrease was very slight, and pesticide use did not decline more rapidly after 2006 than before. Changes in pesticide use differed between French regions and crops. Water pollution did not decrease during the period studied. Possible explanations for the lack of effectiveness of the French environmental plan are considered in the context of European legislation.

Causes and implications of the unforeseen 2016 extreme yield loss in the breadbasket of France

In 2016, France, one of the leading wheat-producing and wheat-exporting regions in the world suffered its most extreme yield loss in over half a century. Yet, yield forecasting systems failed to anticipate this event. We show that this unprecedented event is a new type of compound extreme with a conjunction of abnormally warm temperatures in late autumn and abnormally wet conditions in the following spring. A binomial logistic regression accounting for fall and spring conditions is able to capture key yield loss events since 1959. Based on climate projections, we show that the conditions that led to the 2016 wheat yield loss are projected to become more frequent in the future. The increased likelihood of such compound extreme events poses a challenge: farming systems and yield forecasting systems, which often support them, must adapt.In France, the 2016 winter wheat harvest was at its lowest since over 50 years. Here, Ben-Ari et al. show the role of seasonal temperature and precipitation extremes in this loss, and accounting for both of these variables explains large, historical yield loss events.

Decomposing global crop yield variability

Recent food crises have highlighted the need to better understand the between-year variability of agricultural production. Although increasing future production seems necessary, the globalization of commodity markets suggests that the food system would also benefit from enhanced supplies stability through a reduction in the year-to-year variability. Here, we develop an analytical expression decomposing global crop yield interannual variability into three informative components that quantify how evenly are croplands distributed in the world, the proportion of cultivated areas allocated to regions of above or below average variability and the covariation between yields in distinct world regions. This decomposition is used to identify drivers of interannual yield variations for four major crops (i.e., maize, rice, soybean and wheat) over the period 1961–2012. We show that maize production is fairly spread but marked by one prominent region with high levels of crop yield interannual variability (which encompasses the North American corn belt in the USA, and Canada). In contrast, global rice yields have a small variability because, although spatially concentrated, much of the production is located in regions of below-average variability (i.e., South, Eastern and South Eastern Asia). Because of these contrasted land use allocations, an even cultivated land distribution across regions would reduce global maize yield variance, but increase the variance of global yield rice. Intermediate results are obtained for soybean and wheat for which croplands are mainly located in regions with close-to-average variability. At the scale of large world regions, we find that covariances of regional yields have a negligible contribution to global yield variance. The proposed decomposition could be applied at any spatial and time scales, including the yearly time step. By addressing global crop production stability (or lack thereof) our results contribute to the understanding of a key aspect of global food availability.

Sensitivity analysis of the STICS-MACRO model to identify cropping practices reducing pesticides losses

STICS-MACRO is a process-based model simulating the fate of pesticides in the soil-plant system as a function of agricultural practices and pedoclimatic conditions. The objective of this work was to evaluate the influence of crop management practices on water and pesticide flows in contrasted environmental conditions. We used the Morris screening sensitivity analysis method to identify the most influential cropping practices. Crop residues management and tillage practices were shown to have strong effects on water percolation and pesticide leaching. In particular, the amount of organic residues added to soil was found to be the most influential input. The presence of a mulch could increase soil water content so water percolation and pesticide leaching. Conventional tillage was also found to decrease pesticide leaching, compared to no-till, which is consistent with many field observations. The effects of the soil, crop and climate conditions tested in this work were less important than those of cropping practices. STICS-MACRO allows an ex ante evaluation of cropping systems and agricultural practices, and of the related pesticides environmental impacts.

A global yield dataset for major lignocellulosic bioenergy crops based on field measurements

Reliable data on biomass produced by lignocellulosic bioenergy crops are essential to identify sustainable bioenergy sources. Field studies have been performed for decades on bioenergy crops, but only a small proportion of the available data is used to explore future land use scenarios including bioenergy crops. A global dataset of biomass production for key lignocellulosic bioenergy crops is thus needed to disentangle the factors impacting biomass production in different regions. Such dataset will be also useful to develop and assess bioenergy crop modelling in integrated assessment socio-economic models and global vegetation models. Here, we compiled and described a global biomass yield dataset based on field measurements. We extracted 5,088 entries of data from 257 published studies for five main lingocellulosic bioenergy crops: eucalypt, Miscanthus, poplar, switchgrass, and willow. Data are from 355 geographic sites in 31 countries around the world. We also documented the species, plantation practices, climate conditions, soil property, and managements. Our dataset can be used to identify productive bioenergy species over a large range of environments.

Preceding cultivation of grain legumes increases cereal yields under low nitrogen input conditions

The European Union (EU) has advised to increase the production of grain legumes, both to reduce EU dependency on soybean imports from the Americas and to reduce pollution from intensive cereal production. Several studies have indicated that preceding grain legume had a positive effect on the yields of subsequent cereals; this argument being often used to promote cultivation of grain legumes. However, no quantitative synthesis of the data has been performed on a global scale to estimate the relative increases in cereal yields arisen from cultivation ofxa0preceding grain legumes. Here, we performed a meta-analysis of 1181 yields of cereals cultivated in 15 countries. The results show that the yields of cereals cultivated after grain legumes were, on the average, +xa029% higher than the yields of cereals cultivated after cereals. Our findings also show that the positive effect of grain legumes decreased with the nitrogen (N) fertilization applied to subsequent cereals, then became negligible when the mean N fertilization exceeded 150xa0kgxa0Nxa0ha−1. This threshold is often exceeded in European conventional cereal systems.

Lower average yields but similar yield variability in organic versus conventional horticulture. A meta-analysis

Organic agriculture prohibits the use of almost all synthetic inputs and it is expected to have lower impacts on natural resources than conventional agriculture. However, previous meta-analyses have shown that yields in organic systems are in average 8 to 25% lower compared with conventional systems. Here, we focus on horticulture (fruits and vegetables) and we refine our knowledge by characterising the distributions of organic and conventional yields both in terms of average yield loss and in terms of variability across experiments and across years. We built a new dataset including 636 ratios of organic versus conventional yields covering 37 horticultural species and 17 countries and estimated (i) mean yield ratios, (ii) yield ratio probability distribution across experiments and (iii) interannual yield variances in organic and conventional systems. Our results show that yields in organic horticulture are indeed on average 10 to 32% lower than those in conventional horticulture but they exhibit large variation across experiments. An analysis of yield ratio probability distribution shows that yield loss in organic horticulture has about 10% chances to exceed 50% compared to conventional systems. The analysis gives also around 20% chances to get higher yields in organic horticulture compared to conventional systems. None of the tested covariates (e.g. crop type, climate zone) was able to explain a significant part of the yield ratio variability. We find no evidence of a larger interannual variability (i.e. lower yield stability) in organic versus conventional horticulture. Longer-term trials could nonetheless help substantiate this result. Our results support also the needs to conduct new experiments in countries from the Southern Hemisphere and to collect standard data on crop management and environmental characteristics.

A Bayesian approach to model dispersal for decision support

In agricultural and environmental sciences dispersal models are often used for risk assessment to predict the risk associated with a given configuration and also to test scenarios that are likely to minimise those risks. Like any biological process, dispersal is subject to biological, climatic and environmental variability and its prediction relies on models and parameter values which can only approximate the real processes. In this paper, we present a Bayesian method to model dispersal using spatial configuration and climatic data (distances between emitters and receptors; main wind direction) while accounting for uncertainty, with an application to the prediction of adventitious presence rate of genetically modified maize (GM) in a non-GM field. This method includes the design of candidate models, their calibration, selection and evaluation on an independent dataset. A group of models was identified that is sufficiently robust to be used for prediction purpose. The group of models allows to include local information and it reflects reliably enough the observed variability in the data so that probabilistic model predictions can be performed and used to quantify risk under different scenarios or derive optimal sampling schemes. A Bayesian approach is proposed to model dispersal and to make probabilistic predictions which account for uncertainty.16 statistical gene flow models were designed, calibrated and compared within the Bayesian framework.Models with Zero-inflated Poisson distribution and with exponential decay turn out to provide the most reliable predictions.The proposed approach allows to set up context-specific isolation distances by providing accurate probabilistic predictions.Thanks to precise predictions of intra-field variability, our models allow to design optimal stratified sampling schemes.

A new framework to estimate spatio-temporal ammonia emissions due to nitrogen fertilization in France

In France, agriculture is responsible for 98% of ammonia (NH3) emissions with over 50% caused by nitrogen (N) fertilization. The current French national inventory is based on default emission factors (EF) and does not account for the main variables influencing NH3 emissions. To model the spatio-temporal variability of NH3 emissions due to mineral and organic N fertilization, we implemented a new method named CADASTRE_NH3. The novelty lies in the combined use of two types of resources: the process-based VoltAir model and geo-referenced and temporally explicit databases for soil properties, meteorological conditions and N fertilization. Simulation units are the Small Agricultural Regions. Several sources of information were combined to obtain N fertilization management: census and surveys of the French Ministry of Agriculture, statistics on commercial fertilizer deliveries, and French expertise on physicochemical properties of organic manure. The practical interest of this new framework was illustrated for France during the crop year 2005/06. Aggregation at crop year level showed a reasonable agreement between estimated values derived from CADASTRE_NH3 and those from the French inventory method, for N and ammoniacal-N (TAN) application rates, total NH3 emissions and NH3 EF. Discrepancies were large for organic manure only; national TAN application rates and NH3 emissions were 62-63% lower with CADASTRE_NH3. This was due to divergences in the representation of cattle farm yard manure and in the TAN:N ratio of solid manure. Annual emissions for fertilization in France were estimated to be 270u202fGg NH3, 29% lower than the French national inventory estimate. At the regional level, organic manure contributed to 73% of field NH3 emissions in intensive livestock husbandry areas and to 41% in the other areas. The CADASTRE_NH3 framework can be seen as a Tier 3 approach able to estimate specific regional EF for different mineral fertilizers and organic manure.

Ranking crop species using mixed treatment comparisons

The mixed treatment comparison (MTC) method has been proposed to combine results across trials comparing several treatments. MTC allows coherent judgments on which of the treatments is the most effective. It produces estimates of the relative effects of each treatment compared with every other treatment by pooling direct and indirect evidence. In this article, we explore how this methodological framework can be used to rank a large number of agricultural crop species from yield data collected in field experiments. Our approach is illustrated in a meta-analysis of yield data obtained in 67 field studies for 36 different bioenergy crop species. The considered dataset defines a network of comparisons of crop species. We introduce several Bayesian MTC models based on baseline treatment contrasts and evaluate the practical advantages of these models to produce yield ratio estimates. We explore the consistency of some estimates by node-splitting and compare our results to those obtained with a classical two-way linear mixed model. Results reveal that the model showing the lowest deviance information criterion (DIC) includes both study random effects and study-specific residual variances. But all the tested models including study random effects lead to similar yield ratio estimates. The proposed Bayesian framework allows an in-depth analysis of the uncertainty in the species ranking.