Assessing the adaptive capacity of durum wheat cultivars to future climate

The perspective of climate change requires an analysis of the adaptation possibilities ofspecies currently cultivated. A powerful tool for adaptation is the relevant intra-specificbiodiversity of crops. The knowledge, for different crop cultivars, of the responses todifferent environmental conditions (e.g. yield response functions to water regime) canbe a tool to identify adaptation options to future climate. Moreover climate scenarioneeds to be downscaled to the spatial scale relevant to crop and farm management.Distributed models of crop response to environmental forcing might be used for thispurpose, but severely constrained by the very scarce knowledge on variety-specificvalues of model parameters, thus limiting the potential exploitation of intra-specificbiodiversity towards adaptation.We have developed an approach towards this objective that relies on twocomplementary elements: A) database on climatic requirements of durum wheatvarieties: the yield response functions to water availability were determined fromscientific literature. These functions were applied to describe the behaviour of thecultivars with respect to the soil water availability; B) the simulation performed by theagro-hydrological model SWAP (soil-water-plant and atmosphere), to determine thefuture soil water regime at landscape scale.The case-study presented here shows how the yield response of durum wheat cultivarsto soil water availability can be defined by means of variety-specific threshold values ofsoil water (or evapotranspiration) deficit. The soil water regime calculated by thedistributed model is compared with the threshold values to identify varieties compatiblewith expected climate. The operation is repeated for a set of realizations of eachclimate scenario. This analysis is performed in a distributed manner, i.e. using the timeseries for each model grid to assess possible variations in the extent and spatialdistribution of cultivated area of durum wheat cultivars.The selected study area is a hilly region of about 40,000 ha in Southern Italy (FortoreBeneventano, Campania Region), characterized by a complex geomorphologyincluding clayey and marl flysch hills and highlands, narrow alluvial planes, and smallsandstone relieves. Future climate scenarios in the area were generated within theItalian National Project AGROSCENARI. Climate scenarios at low spatial resolutiongenerated with general circulation models (AOGCM) were down-scaled by means of astatistical model (Tomozeiu et al., 2007). The downscaled climate scenario includes 50realizations of daily minimum, maximum temperature and precipitation data, on aregular grid with a spatial resolution of 35 km, for the 2021-2050 period. The downscaledclimate scenario was further refined by using the distributed model whichdescribes the soil water regime in four soil systems units. Spatial pattern of soil waterand evapotranspiration deficit was determined for the 50 realizations of the daily timeseries, taking into account the four soil systems, and was compared with thresholdvalues to evaluate cultivars' adaptation options to the foreseen future climate. The casestudy shows how, in the future climate scenario, the intra-specific variability will allowto maintain current crop production system.