Adaptability of global olive cultivars to water availability under future Mediterranean climate

Abstract

Adaptation to climate change is a major challenge facing the agricultural sector worldwide. Olive (Olea europaea L.) is a global, high value crop currently cultivated in 28 countries worldwide. Global data to assess the vulnerability of the crop to climate variability are scarce, and in some notable cases, such the United Nations Food and Agricutlure Organization database (FAO, 2006), qualitative assessments rather than quantitative indicators are provided. The aim of this study is to demonstrate a new approach to help overcome these constraints toward a globally applicable method to assess the adaptability of olive cultivars. The adaptability of 11 cultivars, widely used in 11 countries worldwide, was studied using a new generic approach based on the evaluation of soil hydrological regime against cultivar-specific hydrological requirements. The approach requires local data, notably on soil hydrological properties, but it is easily transferable to other countries and regions. We applied an agrohydrological model in 60 soil units to determine hydrological indicators both in a reference (1961–1990) and a future (2021–2050) climate case. We compared indicators with cultivar-specific requirements to achieve the target yield; requirements were established using experimental yield response curves. We estimated the probability of adaptation, i.e., the probability that a given cultivar attains the target yield, and we used it to evaluate the cultivar potential distribution in the study area. At the locations where soil hydrological conditions were favorable, the probabilities of adaptation of the cultivars were high in both climate cases. The results show that the area with suitable conditions for the target yield (area of adaptability) decreased under future climate for all the cultivars, with higher reduction for Frantoio and Maiatica and smaller reduction for Itrana, Nocellara, Ascolana, and Kalamata. These cultivars are currently grown in Argentina, United States (US), Australia, France, Greece, and Italy. Our results indicate also that these cultivars require higher available soil water to attain the target yield, i.e., we may expect similar vulnerability in other parts of the world. Based on these findings, we provide some specific recommendations for enrichment of global databases and for further developments of our approach, to increase its potential for global application.

Appendices

Appendices

1.1 Appendix 1: Description of the procedure for producing the climate datasets

1.1.1 Reference climate case

The reference climate dataset, produced by the Research Unit for Climatology and Meteorology applied to Agriculture (CREA-CMA), was derived by applying the kriging with external drift method (Hengl et al. 2003; Wackernagel 2003) to the meteorological data of the national agrometeorological database (Ministero dell’Agricoltura e delle Foreste 1990). Three auxiliary variables were applied: distance, exposure, and difference of altitude between points. Daily meteorological data were gridded at 35 × 35-km resolution for the period 1950–onwards (Esposito 2010).

1.1.2 Future climate case

Daily values of maximum and minimum temperatures as well as precipitation in the future climate case were produced by the Agenzia Regionale Prevenzione e Ambiente (Arpae – Emilia Romagna) in two phases. Firstly, seasonal mean and standard deviation of the meteorological variables were calculated by a statistical downscaling model (Tomozeiu et al. 2007) using predictions by means of atmosphere-ocean-coupled general circulation models (AOGCM) under emission scenario A1B¹ (ENSEMBLES, Van der Linden and Mitchell 2009). The results were then applied to a weather generator to produce 50 realizations of the daily values of the variables for a representative year of the period between 2021 and 2050. Further details about the procedure were provided by Villani et al. (2011) and Tomozeiu et al. (2013).

1.2 Appendix 2: Description of the soil-water-atmosphere-plant (SWAP) model

The water balance analysis was performed using the SWAP model (Kroes et al. 2009). Assuming one-dimensional vertical flow processes, it calculates the soil water flow through the Richards’ equation that can be written as:

$$ C(h)\cdot \frac{\partial h}{\partial t}=\partial \left[k(h)\cdot \left(\frac{\partial h}{\partial z}+1\right)\right]/\partial z-S(h) $$

(9)

where C(h) = ∂θ/∂h is the soil water capacity, θ (cm³ cm⁻³) is the volumetric soil water content, h (cm) is the soil water pressure head, t (days) is the time, z (cm) is the vertical coordinate taken positively upward, k (cm day⁻¹) is the hydraulic conductivity, and S (cm³ cm⁻³ day⁻¹) is the water extraction rate by plant roots. The Richards’ equation is solved for the unsaturated-saturated zone using an implicit finite difference scheme for numerical integration. Soil water retention is described by the unimodal θ(h) relationship proposed by Van Genuchten (1980) and expressed here in terms of the effective saturation, Se, as follows:

$$ {\displaystyle \begin{array}{l}\mathrm{Se}={\left[\frac{1}{1+{\left(\alpha \left|h\right|\right)}^n}\right]}^m\\ {}\end{array}} $$

(10)

where Se = (θ − θ_r)/(θ₀ − θ_r), θ_r and θ₀ are the residual water content and the water content at h = 0, respectively, and α (cm⁻¹), n (–), and m (–) are curve-fitting parameters. Mualem’s expression (Mualem 1976) was applied to calculate relative hydraulic conductivity (k_r). Assuming m = 1 − 1/n, Van Genuchten (1980) obtained a closed-form analytical solution to predict k_r at specified volumetric water content:

$$ {k}_{\mathrm{r}}\left(\mathrm{Se}\right)=\frac{k\left(\mathrm{Se}\right)}{k_0}={\mathrm{Se}}^{\tau }{\left[1-{\left(1-{\mathrm{Se}}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$m$}\right.}\right)}^m\right]}^2 $$

(11)

where k₀ is the hydraulic conductivity at θ₀ and τ is a parameter which accounts for the dependence of the tortuosity and partial correlation between adjacent pores.

In SWAP, the condition at the bottom boundary can be set in several ways (e.g., pressure head, water table, fluxes, impermeable layer, unit gradient). The upper boundary condition is generally described by the potential evapotranspiration (ET_p), daily precipitation, and irrigation (if it is applied). ET_p can be calculated by the Penman–Monteith equation (Monteith 1965). At a large scale, due to the lack of detailed information required for its application (i.e., air temperature and humidity, net radiation, wind speed), simplified approaches can be used, i.e., Priestley and Taylor (1972) or Hargreaves and Samani (1985). Specifically, SWAP calculates potential evapotranspiration rate of a canopy (ET_p) from reference evapotranspiration (ET_o) and a crop factor. Then, the potential evapotranspiration is partitioned in soil potential evaporation (E_p) and crop potential transpiration (T_p) according to the LAI evolution, following the approach of Ritchie (1972).

SWAP simulates water uptake and actual transpiration according to the model proposed by Feddes et al. (1978) where actual root water uptake (S) is described as a function of the soil water pressure head (h):

$$ S(h)=\alpha (h)\cdot {S}_{\mathrm{max}}=\alpha (h)\cdot \frac{T_{\mathrm{p}}}{\left|{z}_r\right|} $$

(12)

where S_max is the maximum root water extraction rate; α(h) is a semi-empirical function of pressure head h, varying between 0 and 1; and z_r (cm) is the thickness of the root layer. The function α(h) defines four critical values of h, which are related to crop type and to potential transpiration rates. The actual transpiration rate (T_a) is computed by the integration of S(h) over the root layer. The root depth is specified by the user as a function of crop development stage.

1.3 Appendix 3: A generic yield response function

Fig. 10

Relative yield of a cultivar as a function of water availability indicators: observed relative soil water deficit (RSWD_obs) or observed relative evapotranspiration deficit (RETD_obs). The relative yield desirable for adaptation (Yr_target) and the corresponding values of cultivar-specific hydrological requirements (RSWD_req or RETD_req) are shown

1.4 Appendix 4: Description of the hydrological behavior of two soil typological units (STU 52 and 91): physical properties and simulated data

Figure 11a, b shows the time series of daily values of calculated relative transpiration (i.e., T_calc/TM_calc) and soil water pressure head (expressed as log₁₀|h|) between − 0.4 and − 0.6 m (representing the interval of maximum root extraction in the soil layer)in STU 52 (Fig. 11a) and in STU 91 (Fig. 11b). Daily values were averaged within the reference climate. Close to day of year (DoY) 160, the increase of log₁₀|h| and the decrease of relative transpiration indicated the start of water stress. Close to DoY 240, the maximum level of water stress was reached; the variability of log₁₀|h| values in the maximum root extraction layer was negligible and log₁₀|h| values reached approximately the wilting point. This indicated that water-stressed olive tree was able to extract all the available water in the maximum root extraction layer. On DoY 240, relative transpiration was higher in STU 52 (Fig. 11a) than in STU 91 (Fig. 11b) suggesting a lower plant water deficit. Moreover, in STU 52, the water stress period (T_calc/TM_calc ≤ 0.8) was shorter than in STU 91. Figure 12a, b shows respectively soil hydraulic conductivity and water retention curves, based on the Mualem-van Genuchten model (Van Genuchten 1980), both in STU 52 and 91. In each soil, the variability among the horizons was low both in hydraulic conductivity and water retention, but differences between the two STUs were larger. The soil water retention curves showed a similar pattern close to saturation but they differentiated as |h| increased, with large differences (more than 0.10 cm³ cm⁻³) at very dry conditions (i.e., |h| > 10,000 cm). It should be noted that STU 52 had a higher available water capacity (AWC = 205 mm) than STU 91 (AWC = 172 mm). The vertical profile of log₁₀|h| on DoY 240 shows the different patterns of soil water pressure head in STU 52 (Fig. 11c) and in STU 91 (Fig. 11d), caused by different soil physical properties and profile layering. In STU 91, the average of log₁₀|h| was higher than 3.6 in the entire root layer, whereas lower log₁₀|h| values occurred in the upper and lower soil layers of STU 52. Therefore, a larger amount of water was available for root uptake in the upper and lower soil layers of STU 52.

Fig. 11

a, b Mean (lines) and standard deviation (bars) of daily values, averaged over the 1961–1990 period, of calculated relative transpiration (T_calc/TM_calc) (black line) and soil water pressure head (expressed as log₁₀|h|, gray line) in the soil layer of maximum root extraction. Variables were simulated in soil typological unit (STU) 52 (a) and in STU 91 (b). c, d Mean (lines) and standard deviation (bars) of water pressure head (log₁₀|h|) along the soil profile, on DoY 240 in the 1961–1990 period. Variables were simulated in STU 52 (c) and STU 91 (d)

Fig. 12

Soil hydraulic conductivity (a) and water retention curves (b) of soil horizons in soil typological units 52 and 91

1.5 Appendix 5: Table of changes in fractional area of Valle Telesina from reference to future climate within the ranges of the distributions of hydrological indicators

Table 6 Transition matrix of the changes within the ranges of the distributions of calculated hydrological indicators relative evapotranspiration deficit (RETD_calc) and relative soil water deficit (RSWD_calc), from reference to future climate case. Data were expressed as percentage of the surface of Valle Telesina. The distribution of each hydrological indicator was set up by pooling indicator values in each soil and year in reference and future climate case; the ranges of the distributions were defined by the quartiles (Q)

1.6 Appendix 6: Symbols and abbreviations

AWC:

Available water capacity

BBC:

Bottom boundary condition

CO₂:

Carbon dioxide

cv.:

Cultivar

CWSI:

Crop water stress index

DoY:

Day of year

E _p :

Soil potential evaporation

ET_calc:

Calculated actual crop evapotranspiration (mm)

ET_o:

Reference evapotranspiration (mm)

ET_obs:

Observed actual crop evapotranspiration (mm)

ET_p:

Crop potential evapotranspiration (mm)

ETM_calc:

Calculated maximum crop evapotranspiration (mm)

ETM_obs:

Observed maximum crop evapotranspiration (mm)

FACE:

Free-air CO₂ enrichment

FAO:

Food and Agriculture Organization of the United Nations

h :

Soil water pressure head (cm)

k :

Soil hydraulic conductivity (cm day⁻¹)

LAI:

Leaf area index (m² m⁻²)

M :

Median of probability of adaptation (–)

p :

Soil water depletion fraction

P :

Probability of adaptation (–)

Q :

Quartile

R :

Yearly rainfall (mm)

R ² :

Coefficient of determination of regressions

RETD_calc:

Calculated relative evapotranspiration deficit ( ̶)

RETD_calc*:

Calculated relative evapotranspiration deficit averaged over time ( ̶)

RETD_calc⁺:

Calculated relative evapotranspiration deficit averaged over 1 year (–)

RETD_obs:

Observed relative evapotranspiration deficit (–)

RETD_req:

Cultivar-specific hydrological requirement calculated on relative evapotranspiration deficit (–)

RETD_TV:

Threshold value of relative evapotranspiration deficit (–)

RSWD_calc:

Calculated relative soil water deficit (–)

RSWD_calc*:

Calculated relative soil water deficit averaged over time (–)

RSWD_calc⁺:

Calculated relative soil water deficit averaged over 1 year (–)

RSWD_obs:

Observed relative soil water deficit (–)

RSWD_req:

Cultivar-specific hydrological requirement calculated on relative soil water deficit (–)

RSWD_TV:

Threshold value of relative soil water deficit (–)

RTD_calc:

Calculated relative transpiration deficit (–)

SE:

Standard error

SI:

Sensitivity index (mm⁻¹)

SMU:

Soil mapping unit

STU:

Soil typological unit

SWAP:

Soil-water-atmosphere-plant model

T _calc :

Calculated actual crop transpiration (mm)

T _p :

Crop potential transpiration (mm)

TM_calc:

Calculated maximum crop transpiration (mm)

Y _obs :

Observed actual yield (t)

Yr:

Relative yield (–)

Yr_target:

Target relative yield for adaptation (–)

YM_obs:

Observed maximum yield (t)

θ :

Volumetric soil water content (cm³ cm⁻³)