L'Unité Expérimentale de Gotheron développe des programmes d'expérimentation-recherche sur les systèmes de production durable en  ARBORICULTURE  fruitière (abricotiers, pêchers, pommiers, poiriers). © MAITRE Christophe

Water and agriculture

Drought forecasting and measurement 

The key issue for agriculture is to anticipate the consequences of drought on agricultural yields. It is a complex issue – but one with major economic stakes – for which research employs a range of complementary methods.

The years 1976, 1997, 2003, 2005, 2011 were historical periods of drought in France. Droughts cannot be predicted several months ahead of time. However, based on long-term trends, an average temperature increase of 2°C to 4°C can be expected between now and the end of the century, with decreased rainfall during spring and summer, which will be more marked to the west of the country (source: Météo France and IPCC). Agricultural practices must take into account the risk of drought, as pointed out in the 2006 Collective Scientific Expert report Drought and Agriculture*. There are several methods used to quantify the severity of a drought and to plan for its consequences on crop yields.

*http://www.inra.fr/l_institut/expertise/expertises_realisees/secheresse_et_agriculture_rapport_d_expertise

 

Monitoring Crop Water Balance

From an agricultural standpoint, the severity of a drought is defined by the deficit in groundwater levels. One way to determine the deficit is looking at the available water (AW) in the uptake zone compared to what that zone (Z) can hold. A plant can optimally cover its water needs if the reservoir is at least half full (Z>AW/2). Below this point, the plant will close off its stomata and evapotranspiration (ET) will not be optimal.

The water cycle at the plant level. © INRA
The water cycle at the plant level © INRA

The quantity of water available to the plant at any given moment (Z) equals AW + (rainwater + irrigation – runoff and infiltration) - ET

AW:available water in the uptake zone which depends on the ground level and root depth of a given crop.

ET: Evapotranspiration, or the sum of soil evaporation and water consumption by the plant, which releases almost all the water it absorbs into the atmosphere through pores on its leaves (i.e. stomata).

 

The influence of climate on yield: agro-climatic monitoring

The hydric state of the soil (Z/AW) is one of twelve parameters of INRA’s agro-climatic monitoring tool (VAC), developed following the heat wave of 2003 to analyse the influence climate has on crop yields. This tool calculates production characteristics (yield, quality, cycle duration, water stress indices, nitrogen stress, etc.) according to meteorological data (minimum and maximum temperatures, rainfall, sunshine, humidity, wind) recorded at INRA’s ten experimental sites which cover France’s diverse climate. It relies on a plant development model (STICS) and is applicable to seven crop types (including wheat, rapeseed, maize and sunflower). It evaluates a simplified theoretical system for which only the climate is variable. Two components which are arbitrarily fixed and common to all the sites: an average soil and single, but representative, technical crop management sequence. Real yield cannot be calculated, but rather a climatic yield which only takes into consideration climatic influences.

 

Yield forecasting: the MARS system

At the European level, another complementary tool* gives an idea of real yields, but on a larger geographical scale. This tool, called MARS (Monitoring Agricultural ResourceS), calculates the meteorological variables (cumulated temperatures, rainfall, etc.) over areas of 25x25km. It provides month by month estimations of future yields of various crops (winter wheat, barley, maize, potatoes, etc.) by using crop management sequences identified for each zone and soil cartography. For example, it was forecast that 2011 yields for common wheat in France would be 11.5% below 2010 levels.

This result was corroborated by the French Ministry of Agriculture’s department of statistics and forecasting (10% decrease).

* Developed by the European Commission’s Joint Research Centre based in Ispra, Italy