future climate change vulnerability on the agricultural ... · deliverable c4.2: report on the...

Deliverable C4.2: Report on the estimation of future climate change vulnerability on the agricultural sectors of Cyprus, Crete and Sicily (including relative database) Date: 30/09/2017

Report on Activity 4.2

'Use of crop models for assessing the vulnerability of agriculture to climate change'

(Final Version)

Lorenzo Brilli1*, Paolo Merante1, Luisa Leolini2, Camilla Dibari2, Marco Moriondo1

1 CNR-Ibimet, Florence, Via Caproni 8, 50145, Italy

2 University of Florence, P.le delle Cascine 18, 50144, Italy

* l.brilli@ ibimet.cnr.it, [email protected]

Executive summary This report focus on the climate change vulnerability assessment on the agricultural sectors of Cyprus, Crete and Sicily. Future impacts on crop systems were evaluated considering different general circulation models, climate scenarios, crop models and crop types…..to be finished

2 | ADAPT2CLIMA-D e l i v e r a b l e C 4 . 2

Table of Contents

Executive summary .......................................................................................... 2

1. Introduction ............................................................................................ 4

2. Description of the simulation models ...................................................... 5

3. Calibration and validation ....................................................................... 8

3.1. CropSyst .......................................................................................... 8

3.2. OLIVEmodel.CNR ......................................................................... 12

3.3. GrapeModel ................................................................................... 15

4. Simulation scheme and Vulnerability Assessment ............................... 20

4.1. Crop Simulations ........................................................................... 20

4.2. Vulnerability Assessment .............................................................. 20

4.3. Crop response to future climate variability ..................................... 22

4.4. Crop response to sowing seasons, precocity levels and future climate scenarios ..................................................................................... 24

4.5. Overall Crop Vulnerability .............................................................. 41

5. Conclusions ......................................................................................... 42

Acknowledgements .................................................................................... 43

References ................................................................................................. 44


1. Introduction The Action 4.2 (‘Report on the estimation of future climate change

vulnerability on the agricultural sectors of Cyprus, Crete and Sicily (including

relative database’) of the Adapt2Clima (A2C) project is triggered by the need to

understand the impacts of future climate on agricultural areas of three

European islands in the Mediterranean basin namely Crete (Greece), Sicily

(Italy) and Cyprus. The interaction of mid-latitude and tropical atmospheric

circulation processes makes this area of the world as one of the most sensitive

area to climate change (Giorgi, 2006). For instance in the last century the

Mediterranean basin experienced a generalized decrease of precipitation and

the related water availability (Sousa et al., 2011; NorranT and Douguédroit,

2006), a significant annual warming trend (+0.75°C) especially during winter

and summer (Vautard et al., 2007; EEA, 2012) and an increase in climate

extremes (i.e. day with minimum and maximum temperature, heavy

precipitation, etc.) (Kostopoulou and Jones, 2005; Zolina et al., 2008; Costa

and Soares, 2009; Kyselý, 2009; Durão et al., 2010; Rodda et al., 2010; Ulbrich

et al., 2013). These changes in climate conditions have negatively affected the

whole agricultural sector of the regions around and within the basin. A joint

combination between drought and heat stress can negatively affect the

physiological status of the plants (Barnabás et al., 2008; Alves and Setter,

2004), thus reducing crop growth and gross primary production of terrestrial

ecosystems (Ciais et al., 2005). A brief overview about the impacts of climate

extremes on European agriculture in 2003 and 2012, years with a climate

pattern comparable to that expected for the last decades of the 21st century, is

highlighted in Brilli et al. (2014), determining economic losses for the agriculture

and forestry sectors more than 13 billion euros (i.e. 2003) (Copa-Cogeca,

2003). In particular, for the regions around the Mediterranean basin, the main

losses were due to a strong decrease in quantity and quality of harvests, with

sensible damages for cereals (FAO, 2013; JRC/IES MARS Unit, 2012) and tree

crops. Despite future climate projections are affected by uncertainties, the

major part of the studies over the Mediterranean basin suggested a worsening

of climate pattern, which is expected to show a further increase of temperature


and a decrease of total amount of precipitation associated with an increase in

intensity.

In this context, the Activity 4.2 ('Use of crop models for assessing the

vulnerability of agriculture to climate change') is focused on assessing the

impacts of future climate on agricultural sector of the study areas by means of

crop models. This activity is therefore essential in order to provide indications

to reduce crop vulnerability. The use of crop models, indeed, can enable

evaluating benefits provided by the adoption of adaptation strategies compared

to the current management. In the specific, alternative practices can be tested

through modelling so as to prevent production losses, which might be expected

under unchanged management. On these bases, the report firstly (Section 2)

describes the structure of the models used for simulating the main crops

cultivated in the three islands (i.e. wheat and barley amongst sowing crops,

olive tree and grapevine amongst perennial crops together with potatoes and

tomatoes amongst vegetables). Then, in the section 3, the most important

parameters used for crop calibration and statistics of calibration processes are

reported. In section 4, vulnerability of the crop systems has been assessed by

comparing phenology and yields using the same management under future and

current climate conditions. Also, some adaptation practices have been

simulated in order to assess those practices which may reduce crop

vulnerability under future climate. Finally, in section 5 discussions and brief

conclusions were reported.

2. Description of the simulation models

The simulation models used in this activity are the following:: i) CropSyst

(for wheat, barley, tomato and potato); ii) Olive model (for olive tree); iii)

UNIFI.GrapeML (for grapevine).

i) CropSyst (CS) model (Stöckle et al., 2003) is a “multi-year, multi-crop,

daily time step cropping systems simulation model”. CS has been developed in

order to encompass procedures and functions “to simulate productivity of crops

and crop rotations in response to weather, soil and management”. Based on

detailed information about climate, soils, crops and management, CS can


assess both the climate impact on crop performances and the environmental

impacts determined by crop rotations and by different cropping (or farming)

system management (Dalla Marta et al., 2011; Jalota et al., 2011; Moriondo et

al., 2009). Furthermore, CropSyst simulation model may provide a wide array

of simulation’s outputs ranging from the soil water budget, soil plant budget and

soil erosion to specific crop-related outputs as phenology, canopy and root

growth, yield, and biomass production (Stöckle et al., 2003). CS simulates a

“land block fragment” which is an area with uniform soil, weather, crop rotation

and management (i.e. the case study’s sites). CS has been used for various

applications with different purposes as, for instance: a) to identify the efficient

best management practice (BMP) with regard to water and nitrogen use, the

CS model was used in a combined approach of field experimentation and

simulation (Jalota et al., 2011); b) to include environmental impacts in the

assessment of sustainability of farming system (Merante et al., 2014) in a

changing climate (Moriondo et al., 2009); c) to simulate crop production, water

requirement, and cultivation techniques to appraise the energy and water use

related to the cultivation of energy crops (Marta et al., 2011); d) to assess the

climate change impact on crop performances (Stöckle et al., 2009).

ii) The OLIVEmodel.CNR (Moriondo et al., under revision) simulates the

growth and development of olive agroecosystem at daily time step. Growth of

both olive tree and grass cover is also simulated by the model, considering the

competition for water between the two layers (Fig.1). A phenological sub model

reproduces changes in biomass allocation and the final yield, which is

calculated at the end of the growing season as a fraction of total olive tree

biomass accumulation (harvest index, HI). The key process of the model is the

simulation of daily potential biomass increase (g dry matter m-2) for both layers

as dependent on the relevant intercepted radiation (INT.RAD, %), daily

photosynthetic active radiation (RAD, MJ m-2), and Radiation Use efficiency

(RUE, g MJ-1). The fraction of transpirable soil water, (FTSW) is used as index

to rescale potential growth and leaf area of olive tree and grass cover to their

actual values.


iii) UNIFI.GrapeML (Leolini et al., submitted) is a BioMA

(http://www.biomamodelling.org/) software model library jointly developed by

UNIFI and CREA-AA and used for simulating vine development and growth

under different pedo-climatic conditions. The model architecture is an

implementation of the original model of Bindi et al. (1997a, 1997b) and it is

based on simple or composite strategies (model units) that allow a fine

granularity and an easier implementation and maintenance of the code. This

aspect assumes a relevant importance when the collection of the main plant

processes must be implemented in alternative ways maintaining the original

modelling approach. The sharing of the new/alternative modelling approaches

is a prerogative of the BioMA software environment in which the model is

developed (Bregaglio and Donatelli, 2015; Cappelli et al., 2014; Donatelli et al.,

2014; Stella et al., 2014). UNIFI.GrapeML takes in account eight main plant

processes: i) Phenological development, which estimates the main phenology

stages over the grape growing season (i.e. Bud-break, Flowering, Veraison and

Maturity Day Of Year); ii) Leaf Area Growth, which reproduces the plant leaf

are growth and leaf area index (i.e. shoot leaf number, shoot leaf area, plant

leaf area and leaf area index); iii) Biomass accumulation and light interception

(i.e. daily photosynthesis); iv) Extreme event impacts of high/low maximum

temperature around flowering; v) Biomass partitioning between single plant

organs (i.e. fruit biomass, vegetative and total biomass, root deepening); vi)

Evapotranspiration process (i.e. soil evaporation and plant transpiration); vii)

Grape quality (i.e. sugar accumulation); viii) AgroManagement events. The

main model inputs are climate (Maximum and minimum air temperature,

rainfall, wind global solar Radiation), soil water content, management, and

specific phenological data (i.e. number of shoot per plant, bud-break, flowering,

veraison and maturity stages).


3. Calibration and validation 3.1. CropSyst

CropSyst has been calibrated by adjusting the cultivar parameters of

different crops. The calibration process has invested barley, wheat, potato and

tomato. In particular, the calibration was performed for two barley cultivars (i.e.

Mattina and Aliseo) and seven wheat cultivars (i.e. Bronte, Ciccio, Claudio,

Duilio, Iride, Platani and Simeto). By contrast, no specific cultivars have been

used for potato and tomato.

The phenological and biomass data used to calibrate the modelled crops

were retrieved from several Long Term Experiment (LTE): i) Foggia (Apulia,

Italy) provided data for barley and tomato; ii) Caltagirone (Sicily, Italy) provided

data for several wheat cultivars; iii) potato was calibrated using data retrieved

from four different locations: Chinoli (Bolivia), Gisozi (Burundi), Jyndevad

(Denmark), Washington (USA). For tomato has been used the default

calibration.

Below are reported the calibrated parameters and the related statistics for

Barley, Wheat and Potatoes.

a) Barley:

BARLEY Variety Default ALISEO MATTINA Above ground biomass-transpiration 4.5 6 6 Light to above ground biomass conversion 3 4 5.5 Degree days Emergence (°days) 100 100 100 Peak LAI (°days) 600 500 500 Begin flowering (°days) 632 520 530 Begin grain filling (°days) 732 732 732 Physiological maturity (°days) 1100 1100 1075 Unstressed harvest index 0.48 0.5 0.5

Table 1 – Calibrated parameters for Barley

BARLEY Variety Parameter Obs Sim MBE nMBE RMSE nRMSE pearson R2 ALISEO Flowering 125 125 0.6 0.5 3.0 2.4 0.8 0.7


Yield 5820 5300 -520.0 -8.9 1473.0 25.3 0.5 0.2

MATTINA Flowering 127 126 -0.5 -0.4 3.7 2.9 0.8 0.7 Yield 5750 5940 190.2 3.3 1172.9 20.4 0.6 0.4

Table 2 – Overview of statistics of the calibrated barley cultivars Aliseo and Mattina

The statistics regarding the phenology of Aliseo cv. showed contrasting

performances (i.e. flowering date R2: 0.7.; RMSE: 3.0;), whilst overall lower

performances resulted for yield (R2: 0.2; RMSE: 1473).

Fig.1 – Statistical correspondence of the simulated values with the observed values: Flowering date and yield of barley cv Aliseo.


Similarly, Mattina cv. showed contrasting performances in phenology (i.e.

flowering date R2: 0.7; RMSE: 3.7), whilst yields gained good indicators (R2:

0.4; RMSE: 1172.9).

Fig.2 - Statistical correspondence of the simulated values with the observed values: Flowering date and yield of barley cv Mattina.

b) Wheat:

WHEAT

Cultivar Bronte Ciccio Claudio Duili

o Iride

Platani

Simeto

Abv biomass-transp. 4.5 4 5 3.5 3.5 3.5 4 5 Light to ABV biom.conv. 3 3 5 4.5 3.5 4

Initial green leaf area index 0.011 0.03

Peak LAI 600 400 420 420 400 400 400 410 10 | ADAPT2CLIMA-D e l i v e r a b l e C 4 . 2

(°days) Begin flowering (°days) 632 430 450 450 410 440 430 440

Table 3 - Calibrated parameters for the seven wheat cultivars.

WHEAT

Variety Parameter Obs Sim MBE nMBE RMSE nRMSE pearson R2

BRONTE Flowering 114 114 0.0 0.0 5.6 4.9 0.9 0.8 Yield 3073 3060 -12.6 -0.4 913.9 29.7 0.5 0.2

CICCIO Flowering 115 117 2.3 2.0 5.3 4.6 0.9 0.7 Yield1 3053 3364 311.3 10.2 1149.9 37.7 0.1 0.0 Yield2 306 3039 -14.1 0.5 1017.8 33.3 0.1 0.0

CLAUDIO Flowering 121 117 -3.7 -3.1 6.9 5.7 0.9 0.8 Yield 3258 3281 22.4 0.7 994.6 30.5 0.5 0.2

DULIO Flowering 115 111 -4.5 -3.9 6.8 5.9 0.9 0.8 Yield 3224 3196 -28.1 -0.9 1551.8 48.1 -0.1 0.0

IRIDE Flowering 117 116 -1.0 -0.9 5.3 4.5 0.9 0.8 Yield 2921 3219 298.7 10.2 817.0 28.0 0.4 0.2

PLATANI Flowering 113 114 0.4 0.4 4.2 3.7 0.9 0.8 Yield 3113 3261 148.5 4.8 950.3 30.5 0.6 0.3

SIMETO Flowering 117 116 -1.4 -1.2 4.8 4.1 0.9 0.8 Yield 3316 2295 -1021.3 -30.8 1771.8 53.4 0.4 0.2

Table 4 - Overview of statistics of the calibrated seven wheat cultivars.

Also simulations of wheat showed contrasting results (Fig. 3). The

phenological pattern of all varieties has been well reproduced by the model.

This has been confirmed by statistics (Tab .4). The lowest performances were

observed using Ciccio cv (R=0.84, RMSE= 5.3), which in contrast, has been

well simulated by the model.

Conversely, the model was unable to well simulate the final yields. The

best performance has been observed for Platani cv (R=0.55; RMSE= 950), the

worst for Ciccio and Duilio varieties.


Fig.3 - Statistical correspondence of the simulated values with the observed values: Flowering dates and yields of the seven wheat cultivars.

3.2. OLIVEmodel.CNR

The olive model has been calibrated and validated using experimental

data from two different sites located in Tuscany region (Italy): i) FTSW

measured in a 25 years old orchard located at Istituto Tecnico Agrario Statale

(ITAS) farm (Florence, 10.35 E 43.5 N); ii) Net Primary Production (NPP)

extrapolated from three years eddy covariance data (2010-2012) in a rainfed

olive orchard located in “S. Paolina” experimental farm of National Research

Council (Follonica. 42.55N, 10.45E).

Preliminary results showed the model reliability at reproducing both soil

water and growth dynamics. For the FTSW simulation over the soil profile 12 | ADAPT2CLIMA-D e l i v e r a b l e C 4 . 2

indicated that the model correctly simulated the daily course of FTSW in the

layer explored by both grasses and olive tree rooting system (0-30 cm,

r2=0.91). In particular, the model simulated correctly both the drought stress

experienced by the orchard during summer 2016 and the soil water recharge

after early autumnal rainfalls (i.e. DOY 262).

Fig.4 - Daily course of observed and simulated FTSW in 2016 at 30 cm.

The model was additionally tested against three years of eddy covariance

data (2010-2012) showing high performances also to reproduce the daily NPP.

In 2010 (Fig. 5a) the model was able to detect and reproduce the two main

peaks of dry matter production of the orchards (i.e. early and late springtime)

whilst, on yearly basis, it only slightly underestimated the whole olive orchard

biomass (745 g DM m-2) compared to the observed (827 DM g m-2). For the

following years (2011 and 2012, Fig. 5b, c) the model confirmed its ability to

reproduce the daily NPP, also considering the effect of prolonged drought

periods on NPP in both years. Overall, the model was able to simulate the

yearly trend of three contrasting years, capturing changes of carbon driven by

climatic variables, resulting to be a reliable tool for olive orchards biomass

prediction.


Fig.5 - Simulated and observed daily course of NPP for the three years of study.


3.3. GrapeModel

UNIFI.GrapeML has been calibrated for two grape cultivars, namely

Chardonnay and Sangiovese. For Chardonnay cv. the model was calibrated

with observed data of phenology, soil water and grape yield retrieved in a

vineyard located in Spain (lat. 41.53 N, long. 1.7 E, 340 m. a.s.l.). Climate, soil

and management practices were monitored during the period 1998-2012. For

Sangiovese cv. phenology and grape quality data from Susegana (45°51"N,

12°15E, 83 m a.s.l, Treviso, Italy) and Montalcino (43°03'N; 11°29'E; 326 m

a.s.l., Siena, Italy) were used for model calibration. More specifically, the

phenological pattern was calibrated considering bud-break, flowering, veraison

and maturity over the period 1964-2005, whilst the grape quality was calibrated

using data collected only at Montalcino during the period 1998-2015.

For assessing the most sensitive model parameters a sensitivity analysis

(SA) was performed. The SA evidenced that parameters driving the leaf area

appearance and expansion showed the most relevant effect on final fruit

biomass production (Table 5).


Parameter Description Default Calibration Units aParam Curve shape parameter 0.005 0.006 unitless cParam Optimal chilling temperature 2.8 1.79 °C ChillingReq Chilling requirement 78.692 59.28 CU db Slope of forcing unit eq.for bud-break -0.26 -0.23 unitless df Slope of forcing unit eq.for flowering -0.26 -0.13 unitless dv Slope of forcing unit eq. for veraison -0.26 -0.4 unitless dm Slope of forcing unit eq. for maturity -0.26 -0.4 unitless eb Optimal forcing temperature for bud-break 16.06 15.15 °C ef Optimal forcing temperature for flowering 16.06 9.43 °C ev Optimal forcing temperature for veraison 16.06 10.93 °C em Optimal forcing temperature for maturity 16.06 17.24 °C Col Curve shape parameter 176.26 108.36 unitless co2 Curve shape parameter -0.015 -0.014 unitless LimitForcingReq Last day of chilling effect on forcing requirement 234 200 day FloweringReq Forcing requirement for flowering 24.7 38.81 FU VeraisonReq Forcing requirement for veraison 51.14 57.31 FU MaturityReq Forcing requirement for maturity - 30.73 FU Leaf Area Growth ShootLeafAreaExp Curve shape parameter of Shoot Leaf Area equation 2.13 1.5 unitless ShootLeafNumberIntercept Curve shape parameter of Shoot Leaf Number equation -0.28 -0.22 n° of leaves d-1

LeafAppearanceRate1 Curve shape parameter of Shoot Leaf Number equation 0.04 0.06 n° of leaves d-1

°C-1 SLN1 Coefficient of water stress equation on leaf development 25.9 39.92 unitless SLN2 Coefficient of water stress equation on leaf development 17.3 39.16 unitless Light interception and biomass accumulation PHO1 Coefficient for water stress effect on photosynthesis 6.01 12.9 unitless PHO2 Coefficient for water stress effect on photosynthesis 8.59 14.1 unitless

Biomass partitioning HarvestIndex Slope of Fruit Biomass Index equation 0.0035 0.00443 d-1 Evapotranspiration WUE Water use efficiency 3.86 6.1 Pa Extreme events impact Tmax Maximum temperature for fruit-set at flowering 40 41 °C Tmin Minimum temperature for fruit-set at flowering 1 1 °C Topt Optimum temperature for fruit-set at flowering 25 25 °C q Curve shape parameter 1.9 1.9 unitless

Table 5 – Calibrated parameters of the UNIFI.GrapeML


Calibration results showed satisfactory performances for both cultivars.

Statistics related to different phases are hereafter reported:

i) Phenology: for Chardonnay cv. using the chilling-forcing approach of

Caffarra and Eccel (2010) the higher performances were found for

flowering (R: 0.64; RMSE: 4.39) and maturity (R: 0.69; RMSE: 4.45),

whilst little lower performances were obtained for bud-break (R: 0.52;

RMSE: 5.05) and veraison (R: 0.44; RMSE: 4.88; Fig. 6). For

Sangiovese cv. satisfactory performances were observed for all

phases considered (i.e. flowering, RMSE: 4.54; R: 0.78; bud-break,

RMSE: 5.96; R: 0.63; veraison, RMSE: 6.85; R: 0.57; and maturity,

RMSE: 10.84; R: 0.39).

Fig.6 – Phenology calibration of the Chardonnay grape variety

ii) Soil Water Content: for Chardonnay cv. the calibration carried out at

four different soil depth (0.3, 0.5 ,0.7 and 1 m) showed good

performances (Layer 10-30: R: 0.62 RRMSE: 26.8; Layer 30-50: R:

0.66 RRMSE: 12.9; Layer 50-70: R: 0.81, RRMSE: 8.34; Layer 70-

100: R: 0.72, RRMSE: 10.17)

iii) Fruit biomass: the calibration of fruit biomass showed satisfactory

performance (R: 0.59; RRMSE: 23.00; EF: 0.10) over the period

1998-2012 (Fig.7).

Fig.7 -Fruit Biomass (g/m2 dry matter) calibration of the Chardonnay variety


4. Simulation scheme and Vulnerability Assessment 4.1. Crop Simulations

After the calibration phase, the crop simulations of wheat, barley,

potato and tomato were performed with CropSyst (version 3.2), while

simulations of olive trees and grapevine were run with OliveModel.CNR

and UNIFI.GrapeML models, respectively. Simulations have been

performed under three different climate conditions: the Baseline-

Historical, which is based on a CO2 concentration of 360 ppm in the time

slice 1971-2000, the RCPs (Representative Concentration Pathways)

4.5 and 8.5 referring to a CO2 concentration of 485 and 540 ppm

respectively both, covering the temporal period 2031-2060. The

aforementioned scenarios were produced by NOA through a dynamic

downscaling of the climatic outcomes of two Global Climate Models, MPI

(Max Planck Institute for Meteorology model MPI-ESM-LR) and MOHC

(Met Office Hadley Centre) and applying a Regional Climate Model

SMHI (Swedish Meteorological and Hydrological Institute). Besides, the

Harmonized World Soil Database by FAO has been used to take into

consideration the main various soil textures and hydrologic properties

characterizing the soils of the three islands. In order to effectively

analyse the crop performances under different climatic conditions, six

performance indicators were selected: Flowering date and the Maturity

date regarding the crop phenology; the Actual Evapotranspiration (AE),

the Potential Evapotranspiration (PE), and the rate between the two

AE/PE, covering physiological aspects; the yield to analyse the biomass

changes. Furthermore, to appreciate the different sensitivity of the

selected crops in different climatic conditions, the simulations outcomes

were provided both in absolute and relative terms where the latter refers

to the difference between the performance value of the scenario RCP

4.5 or RCP 8.5 and the performance value of the Baseline (Historical).

This was done for each performance indicator.

4.2. Vulnerability Assessment

Assessing the degree to which a cropping system experience harm

because of specific hazards and or threats plays a crucial role in planning

effective response to changes of the surrounding environment. In order to

appreciate the variation of the crops’ vulnerability degree to climate change, the

crop vulnerability has been assessed by changes of four out of the six crop

performance indicators (i.e. flowering and maturity dates, yield, ETA/ETP ratio)

that were used in the simulation phase. Vulnerability was calculated by the

difference between the baseline (historical) and the two different future climate

scenarios (i.e. RCP4.5 and RCP8.5, time slice 2031-2060), further taking into

consideration different sowing seasons and precocity levels. Results were

following summarized for each island reporting the global average and the

related changes according to the climatic scenario and the sowing seasons (for

cereals and vegetables)/ precocity levels (for grapevine and olive tree) adopted

(Tab 6). It is underlined that, because of the considerable amount of simulations

and related outcomes, in this document the only simulations results under the

RCM-MOHC were discussed in terms of crop vulnerability.


Crop type

Sowing season/Precocity

Sowing date

Budbreak (DOY)

Flowering (DOY)

Harvest (DOY)

Barley - Wheat

Early Autumn 10 November - - - - - - - - - Late Autumn 30 November - - - - - - - - - Winter 31 January - - - - - - - - - Spring 15 February - - - - - - - - -

Tomato

Early Winter 30 January - - - - - - - - - Late Winter 28 February - - - - - - - - - Early Spring 30 March - - - - - - - - - Late Spring 31 April - - - - - - - - -

Potato Late Autumn 20 October - - - - - - - - - Early winter 15 November - - - - - - - - - Late Winter 25 December - - - - - - - - -

Grapevine Early - - - 80-90 - - - 235-240 Medium - - - 100-110 - - - 260-270 Late - - - 110-130 - - - 265-275

Olive tree Early - - - - - - 145-155 240-270 Late - - - - - - 160-170 300-330

Table 6 – Description of sowing dates and phenological phases of the simulated crops.

4.3. Crop response to future climate variability

The crops response to both future climate scenarios has been firstly

evaluated by averaging the different sowing seasons and precocity levels

adopted (Table 7). This allowed to have a first indication about the crop

response to only future climate variability, regardless sowing time and precocity

levels.

Site Crop type Flowering (DOY) Maturity (DOY) Yield (Kg/ha) ETA/ETP (mm) - - - - - - Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5

Sicily

Wheat 114 -11 -13 185 -13 -15 5557.5 6% 9% 0.4 -1% 7% Barley 130 -14 -17 181 -15 -18 5675.4 7% 9% 0.4 1% 10% Tomato 153 -12 -15 226 -18 -20 8488.5 1% 3% 0.3 -7% -1% Potato 143 -21 -26 180 -21 -25 3410.2 15% 17% 0.3 3% 12% Grapevine 154 -6 -8 255 -10 -11 1444.3 -11% -3% 0.5 -11% -6% Olive tree 162 -15 -19 310 -41 -47 962.8 -1% 8% 0.6 -4% 0%

Cyprus

Wheat 99 -10 -12 169 -12 -13 5245.6 -2% 4% 0.4 0% -4% Barley 111 -13 -16 162 -14 -18 5277.2 -1% 5% 0.4 6% 4% Tomato 138 -12 -15 209 -16 -19 8718.4 -7% -6% 0.3 -8% -10% Potato 114 -22 -27 153 -21 -26 3440.8 17% 22% 0.4 9% 8% Grapevine 148 -4 -4 247 -6 -7 990.4 -7% -17% 0.3 -6% -16% Olive tree 144 -16 -20 271 -33 -40 926.3 0% 3% 0.5 4% 0%

Crete

Wheat 105 -9 -10 178 -12 -14 5053.7 -7% 8% 0.3 0% 3% Barley 120 -13 -16 172 -15 -18 5108.1 -6% 9% 0.4 4% 9% Tomato 149 -13 -16 222 -19 -22 11240.0 -2% -1% 0.3 -8% -1% Potato 129 -22 -26 168 -21 -26 3051.8 11% 22% 0.3 5% 10% Grapevine 152 -7 -8 254 -10 -12 1190.3 -9% 0% 0.4 -11% -7% Olive tree 156 -16 -20 299 -40 -48 976.7 2% 10% 0.5 2% 6%

Table 7 – Crops' response to the only future climate variability for each test site. Values were obtained by averaging results from the

different sowing seasons and precocity levels adopted.

Looking at Table 7, an advancement of the phenological pattern for all

crops, which is higher under RCP8.5 than RCP4.5, was observed in the all

three sites compared to the baseline.

In Sicily, cereals and vegetables showed a yield increase compared to the

baseline. The highest increase was found for potato (+17%), whilst the lowest

for tomato (+3%). Both wheat and barley showed a general increase of 9%

compared to the baseline. For the woody crops, olive tree showed low variation

under RCP4.5 and an increase of 8% under RCP8.5, compared to the baseline.

By contrast, grapevine showed a production decrease under both scenarios,

with the greatest losses under RCP4.5 (-11%).

In Cyprus, wheat and barley showed a slight yield decrease under RCP4.5

(-2% and -1%, respectively) and a more pronounced yield increase under

RCP8.5 (+4% and +5%, respectively) compared to the baseline. Vegetables

showed an opposite pattern: whilst tomato showed a similar yield decrease for

both RCPs (-6.5%, on average), potato showed a considerable yield increase

both under RCP4.5 (+17%) and RCP8.5 (+22%). For the woody crops, olive

tree showed no variation under RCP4.5 but a slight yield increase under

RCP8.5 (+3%) compared to the baseline. By contrast, grapevine showed the

greatest yield decrease among all crops, with production reduction of 7% and

17% under RCP4.5 and RCP8.5, respectively.

In Crete, the general yield pattern reflected that observed at Cyprus.

Wheat and barley showed a yield decrease under RCP4.5 (-7% and -6%,

respectively) and a considerable yield increase under RCP8.5 (+8% and +9%,

respectively) compared to the baseline. Concerning vegetables, tomato

showed a slight yield decrease for both RCP (-1.5%, on average), whilst potato

showed a considerable yield increase both under RCP4.5 (+11%) and RCP8.5

(+22%). For the woody crops, grapevine showed a strong yield decrease under

RCP4.5 (-9%), whilst no changes where found under RCP8.5, compared to the

baseline. The olive tree showed a slight yield increase under RCP4.5 (+2%)

and a stronger yield increase under RCP8.5 (+10%) compared to the baseline.

4.4. Crop response to sowing seasons, precocity levels and future climate

scenarios

The response of the crops to the interaction among sowing seasons or

crop precocity levels (henceforth identified as early, medium and late precocity)

and future climate scenarios has been reported for each site (Tables 8, 9, 10,

11,12). Results allowed to evaluate the best management practices and

genotype to adopt to limit the impacts due to the expected future climate

variability in each area analyzed.

4.4.1. Wheat

In Sicily, the phenological pattern indicated an earlier flowering and

maturity time under warmer scenarios compared to the baseline. In particular,

the maximum advance was found using the RCP8.5 in all possible scenarios

(management + crop type + site). This advance is found in a range of 1-3 weeks

for both flowering and maturity date by progressively anticipating the sowing

time (Table 8). Concerning crop production, under the baseline scenario the

early autumn provided the highest yield (6326.3 Kg/ha). Delaying the sowing

date production tended to decrease, reaching the minimum production using

the Spring sowing (-24.6%). Looking at future projections, the highest yield

increase was found under RCP8.5 using the Early Autumn sowing (+12.0%),

whilst the highest decrease was found under RCP4.5 using spring sown (-

25.3%) (Table 11). Under the baseline, the Early Autumn sowing provided the

highest ETA/ETP ratio (0.0.47 mm). Delaying the sowing date, ETA/ETP ratio

tended to decrease, reaching the minimum using spring sowing (-33.2%).

Looking at future projections, the highest ETA/ETP ratio increase was found

under RCP8.5 using the Early Autumn sowing (+9.9%), whilst the highest

ETA/ETP ratio decrease was found under RCP4.5 using spring sowing (-

36.7%) (Table 12).

In Cyprus, the phenological pattern indicated an earlier flowering and





time (Table 9). Concerning crop production, under the baseline the early


autumn sowing provided the highest yield (6017.1 Kg/ha). Delaying the sowing

date, production tended to decrease, reaching the minimum production using

spring sowing (-28.2%). Looking at future projections, the highest yield increase

was found using the early autumn sowing under RCP8.5 (+5.3%), whilst the

highest decrease was found under RCP4.5 using spring sowing (-31.1%)

(Table 11). Under the baseline, sowing at early autumn provided the highest

ETA/ETP ratio (0.49 mm). Delaying the sowing date, ETA/ETP ratio tended to

decrease, reaching the minimum using spring sowing (-45.3%). Looking at

future projections, the highest ETA/ETP ratio increase was found using the

early autumn sowing under RCP8.5 (+0.9%), whilst the highest ETA/ETP ratio

decrease was found under RCP8.5 using spring sowing (-50.4%) (Table 12).

In Crete, the phenological pattern indicated an earlier flowering and







date, production tended to decrease, reaching the minimum production using




(Table 11). Under the baseline, sowing at early autumn showed the highest


decrease, reaching the minimum using spring sowing (-41.7%). Looking at




4.4.2. Barley

In Sicily the phenological pattern indicated an earlier flowering and



the highest advance was found using the RCP8.5 in all possible scenarios



time (Table 8). Concerning crop production, under the baseline the highest

yield were obtained using the early autumn sowing (6515.8 Kg/ha). Delaying

the sowing date the production tended to decrease, reaching the minimum

using spring sowing (-27%). Looking at future projections, the highest yield

increase was found using the early autumn sowing under RCP8.5 (+11.5%),

whilst the highest decrease was found using spring sowing under RCP4.5 (-

25.9%) (Table 11). Under the baseline, the early autumn sowing provided the

highest ETA/ETP ratio (0.50 mm). Delaying the sowing date, ETA/ETP ratio

tended to decrease, reaching the minimum using the spring sowing (-35.4%).


using the early autumn sowing under RCP8.5 (+13.1%), whilst the highest

ETA/ETP ratio decrease was found under RCP4.5 using the spring sowing (-

37.5%) (Table 12).

In Cyprus the phenological pattern indicated an earlier flowering and


the maximum advance is found using the RCP8.5 in all possible scenarios




autumn sowing provided the highest yield (6339 Kg/ha). Delaying the sowing

date, production tend to decrease, reaching the minimum production using the




(Table 11). Under baseline, the early autumn sowing provided the highest


decrease, reaching the minimum using the spring sowing (-45.6%). Looking at





In Crete the phenological pattern indicated an earlier flowering and







date, production tend to decrease, reaching the minimum production using the


was found using the early autumn sowing under RCP8.5 (+8%), whilst the


(Table 11). Under the baseline, the early autumn sowing provided the highest

ETA/ETP ratio (0.47 mm). Delaying the sowing date, the ETA/ETP ratio tended

to decrease, reaching the minimum using spring sowing (-43.5%). Looking at



decrease was found under RCP4.5 using the spring sowing (-44.0%) (Table

12).

4.4.3. Tomato






time (Table 8). Concerning crop production, under the baseline the late winter

sowing provided the highest yield (9524 Kg/ha). Delaying the sowing date,

production tended to decrease, reaching the minimum production using the late


was found under RCP8.5 using the early winter sowing (+12.7%, on average),


whilst the highest decrease was found using the late spring sowing under both

RCPs (-47.4%, on average) (Table 11). Under the baseline, the early winter

sowing provided the highest ETA/ETP ratio (0.34 mm). Delaying the sowing

date, ETA/ETP ratio tended to decrease, reaching the minimum using the late

spring sowing (-30.5%). Looking at future projections, the highest ETA/ETP

ratio increase was found using the Early Winter sowing under RCP8.5 (+7.2%),

whilst the highest ETA/ETP ratio decrease was found under RCP4.5 using late

spring sowing (-41.35%) (Table 12).

In Cyprus the phenological pattern indicated an earlier flowering and





time (Table 9). Concerning crop production, under the baseline the early winter

sowing provided the highest yield (10781.2 Kg/ha). Delaying the sowing date,


spring sowing (-50%). Looking at future projections, the highest yield increase

was found under RCP8.5 using the early winter sowing (+9.2%), whilst the

highest decrease was found using the late spring sowing under both RCPs (-

63.4%, on average) (Table 11). Under the baseline, the early winter sowing

provided the highest ETA/ETP ratio (0.34 mm). Delaying the sowing date, the

ETA/ETP ratio tended to decrease, reaching the minimum using the late spring

sowing (-45.2%). Looking at future projections, the highest ETA/ETP ratio

increase was found using the early winter sowing under RCP8.5 (+4.5%),



In Crete the phenological pattern indicated an earlier flowering and





time (Table 10). Concerning crop production, under the baseline the late winter


sowing provided the highest yield (12070.3 Kg/ha). Delaying the sowing date,



was found under RCP8.5 using the early winter sowing (+8.1%, on average),

whilst the highest decrease was found using the late spring sowing under both

RCPs (-39.3%, on average) (Table 11). Under the baseline, the early winter

sowing provided the highest ETA/ETP ratio (0.33 mm). Delaying the sowing

date, the ETA/ETP ratio tended to decrease, reaching the minimum using the

late spring sowing (-23.4%). Looking at future projections, the highest ETA/ETP

ratio increase was found using early winter sowing under the RCP8.5 (+8.4%),



4.4.4. Potato



the maximum advance was found under the RCP8.5 in all possible scenarios


for flowering and in a range of 2-3 weeks for maturity date by progressively

anticipating the sowing time (Table 8). Concerning crop production, under the

baseline the late winter sowing provided the highest yield (3451.8 Kg/ha), which

tended to slight decrease by anticipating the sowing time until to reach the

minimum production using the late autumn sowing (-3.6%). Looking at future

projections, all scenarios (i.e. sowing dates + RCPs) showed a yield increase

compared to the baseline, varying in a range between 10.0% to 18.2%. This

maximum yield increase was found under RCP8.5 using the early winter sowing

(Table 11). Under the baseline, the late autumn sowing provided the highest


decrease, reaching the minimum using the late winter sowing (-27.6%). Looking

at future projections, the highest ETA/ETP ratio increase was found under

RCP8.5 using the late autumn sowing (+16.7%), whilst the highest ETA/ETP


ratio decrease was found under RCP4.5 using the late winter sowing (-22.2%)

(Table 12).

At Cyprus the phenological pattern indicated an earlier flowering and






baseline the late winter sowing provided the highest yield (3582.7 Kg/ha).

Yields tended to decrease by advancing the sowing time, reaching the minimum

production using the late autumn sowing (-11.5%). Looking at future

projections, all scenarios (i.e. sowing dates + RCPs) showed a yield increase

compared to the baseline, which varied in a range between 2.9% and 21.8%.

The highest yield increase was found under RCP8.5 using the early winter

sowing. Under the baseline, the late autumn sowing provided the highest

ETA/ETP ratio (0.44 mm) (Table 11). Delaying the sowing date, ETA/ETP ratio

tended to decrease, reaching the minimum using late winter sowing (-36.7%).


under both RCPs using the late autumn sowing (+7.2%, on average), whilst the

highest ETA/ETP ratio decrease was found under both RCPs using the late

winter sowing (-32.1%) (Table 12).

At Crete the phenological pattern indicated an earlier flowering and






baseline the late winter sowing provided the highest yield (3159.1 Kg/ha), which

slightly decreased by anticipating the sowing time, with the minimum production

found using the late autumn sowing (-7.0 %). Looking at future projections, all

scenarios (i.e. sowing dates + RCPs) showed a yield increase compared to the

baseline which varied in a range between 4.4-23.1%. The highest yield increase


was found under RCP8.5 using the late winter sowing (Table 11). Under the

baseline, the late autumn sowing provided the highest ETA/ETP ratio (0.38

mm). Delaying the sowing date, the ETA/ETP ratio tended to decrease,

reaching the minimum using the late winter sowing (-31.4%). Looking at future

projections, the highest ETA/ETP ratio increase was found under RCP8.5 using

the late autumn sowing (+10.4%), whilst the greatest decrease of ETA/ETP

ratio was found under RCP4.5 using the late winter sowing (-29.7%) (Table 12).

4.4.5. Grapevine




(management + crop type + site). This maximum advance was in a range of 1-

2 weeks for both flowering and maturity using the medium precocity (Table 8).

Concerning crop production, under the baseline the grapevine showed highest

yield using the late precocity (2094.9 Kg/ha), whilst strong yield decrease was

observed using both early (-51.1%) and medium (-42.1%) precocity levels.

Looking at future projections, all scenarios (i.e. sowing dates + RCPs) showed

a strong yield decrease compared to the baseline, with the highest losses (-

54.3%) using the early precocity under RCP4.5. The lower decrease was

observed under RCP8.5 (-8.3%) using the late precocity (Table 11). Under the

baseline, the early precocity provided the highest ETA/ETP ratio (0.49 mm),

whilst the minimum was found using the medium precocity (-15.0%). Looking

at future projections, all scenarios (i.e. sowing dates + RCPs) showed an

ETA/ETP ratio decrease compared to the baseline, with the highest decrease

(-26.0%) using the medium precocity under RCP4.5 and the lowest decrease

(-2.8%) using the early precocity under RCP8.5 (Table 12).









observed using both early (-45.2%) and medium (-45.1%) precocity. Looking at

future projections, all scenarios (i.e. sowing dates + RCPs) showed a strong

yield decrease compared to the baseline, with the highest losses (-55.5%) using

the medium precocity under RCP8.5. The lower decrease was observed under

RCP4.5 (-8.3%) using the late precocity (Table 11). Under the baseline, the

early precocity provided the highest ETA/ETP ratio (0.36 mm), whilst the

minimum was found using the medium precocity (-18.6%). Looking at future

projections, all scenarios (i.e. sowing dates + RCPs) showed an ETA/ETP ratio

decrease compared to the baseline, with the highest decrease (-33.0%) using

the medium precocity under RCP8.5 and the lowest decrease (-0.8%) using the

early precocity under RCP4.5 (Table 12).








observed using both early (-55.2%) and medium (-41.4%) precocity. Looking at

future projections, all scenarios (i.e. sowing dates + RCPs) showed a strong

yield decrease compared to the baseline, with the highest losses (-56.8%) using

the early precocity under RCP4.5. The lower decrease was observed under

RCP8.5 (-6.2%) using the late precocity (Table 11). Under baseline, the late

precocity provided the highest ETA/ETP ratio (0.38 mm), whilst the minimum

was found using the medium precocity (-7.5%). Looking at future projections,

all scenarios (i.e. sowing dates + RCPs) showed an ETA/ETP ratio decrease

compared to the baseline, with the highest decrease (-18.5%) using the

medium precocity under RCP4.5 and the lowest decrease (-7.9%) using the

early precocity under RCP8.5 (Table 12).


4.4.6. Olive tree




(management + crop type + site). The maximum advance was in a range of 2-

3 weeks for flowering and in a range of 5-8 weeks for maturity using the late

precocity (Table 8). Concerning crop production, under the baseline the olive

showed highest yield using the late precocity (1048.8 Kg/ha), whilst strong yield

decrease was observed using early precocity (-16.4%). Looking at future

projections, the highest yield decrease (-17.8%) was found using the early

precocity under RCP4.5, whilst the highest yield increase (+7.9%) was found

using the late precocity under RCP8.5 (Table 11). Under baseline, the late

precocity provided the highest ETA/ETP ratio (0.56 mm), whilst the early

precocity showed an ETA/ETP ratio reduction of 1.7%. Looking at future

projections, the highest ETA/ETP ratio decrease was found using the late

precocity under RCP4.5 (-7.14%) whilst the highest increase (+3.07) was

observed using the early precocity under RCP8.5 (Table 12).






growing cycle (Table 9). Concerning crop production, under the baseline the

olive showed highest yield using the late precocity (999.2 Kg/ha), whilst strong

yield decrease was observed using early precocity (-14.6%). Looking at future



using the late precocity under RCP8.5 (Table 11). Under baseline, the early

precocity provided the highest ETA/ETP ratio (0.54 mm), whilst the late

precocity showed an considerable ETA/ETP ratio reduction (-11.4%). Looking

at future projections, the highest ETA/ETP ratio decrease was found using the


late precocity under RCP8.5 (-13.4%) whilst the highest increase (+6.4) was

observed using the early precocity under RCP4.5 (Table 12).






precocity (Table 10). Concerning crop production, under the baseline the olive

showed highest yield using the late precocity (1055 Kg/ha), whilst strong yield

decrease was observed using early precocity (-14.9%). Looking at future



using the late precocity under RCP8.5 (Table 11). Under baseline, late

precocity provided the highest ETA/ETP ratio (0.51 mm), whilst the early

precocity showed an ETA/ETP ratio reduction of 7.4%. Looking at future

projections, the highest ETA/ETP ratio decrease was found using the late

precocity under RCP4.5 (-10.2%) whilst the highest increase (+10.3) was

observed using the early precocity under RCP8 (Table 12).


Site Crop type Sowing season- Precocity Flowering (DOY) Maturity (DOY) Yield (Kg/ha) ETA/ETP (mm)

- - - - - - - - - Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Sicily

Wheat

Early autumn 79 -16 -20 164 -16 -20 6326.3 10.0% 12.0% 0.47 3.5% 9.9% Late autumn 98 -14 -17 174 -15 -18 6093.0 11.3% 12.1% 0.44 1.8% 9.0% Winter 135 -8 -10 198 -11 -12 5039.7 3.2% 7.1% 0.34 -3.2% 6.1% Spring 143 -7 -6 203 -10 -10 4771.0 -0.9% 3.9% 0.32 -5.2% 2.3%

Barley

Early autumn 97 -18 -22 158 -18 -21 6515.8 10.5% 11.5% 0.50 5.5% 13.1% Late autumn 115 -16 -20 170 -16 -20 6233.2 11.9% 12.9% 0.46 3.3% 11.7% Winter 150 -11 -13 195 -13 -15 5195.5 4.9% 8.2% 0.35 -1.7% 9.6% Spring 156 -10 -12 200 -12 -14 4757.1 1.4% 4.5% 0.33 -3.2% 7.5%

Tomato

Early winter 148 -17 -21 222 -20 -24 9487.4 11.7% 13.6% 0.34 -2.1% 7.2% Late winter 150 -14 -17 223 -19 -22 9524.0 8.5% 12.3% 0.32 -4.7% 3.4% Early spring 153 -11 -13 226 -18 -20 8861.1 1.4% 3.6% 0.29 -3.5% 1.1% Late Spring 162 -7 -8 232 -15 -16 6081.6 -18.0% -17.1% 0.24 -15.6% -14.2%

Potato Late autumn 123 -27 -32 165 -25 -30 3327.3 14.8% 18.5% 0.40 8.2% 16.7% Early winter 144 -21 -25 180 -21 -25 3451.6 16.6% 18.2% 0.35 3.9% 13.2% Late winter 163 -17 -20 195 -18 -21 3451.8 12.2% 15.8% 0.29 -2.2% 7.5%

Grapevine Early 143 -5 -6 237 -9 -10 1024.3 -6.4% 2.8% 0.49 -9.2% -2.8% Medium 160 -8 -10 266 -11 -13 1213.7 -11.6% -4.6% 0.42 -12.9% -7.2% Late 158 -6 -8 263 -10 -11 2094.9 -14.2% -8.3% 0.47 -12.3% -6.8%

Olive tree Early 148 -15 -19 271 -34 -39 876.7 -1.6% 8.9% 0.55 -0.3% 4.8% Late 177 -15 -18 348 -47 -56 1048.8 -0.8% 7.9% 0.56 -7.1% -4.5%

Table 8 – Crops’ response to the interaction among sowing seasons, precocity levels and future climate scenarios for Sicily.


- - - - - - - - - Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Cyprus

Wheat

Early autumn 56 -17 -20 142 -15 -20 6017.1 -1.2% 5.3% 0.49 0.8% 0.9% Late autumn 79 -13 -16 155 -11 -12 5997.7 -1.4% 4.3% 0.45 1.5% -1.8% Winter 125 -3 -4 186 -10 -12 4650.1 -2.3% 3.1% 0.30 -0.9% -7.6% Spring 137 -5 -6 194 -12 -9 4317.5 -4.0% 3.2% 0.27 -1.8% -9.2%

Barley

Early autumn 72 -18 -22 134 -18 -23 6339.0 -2.2% 3.9% 0.54 5.1% 6.7% Late autumn 94 -15 -19 148 -16 -20 6188.7 -1.4% 4.3% 0.48 7.7% 7.8% Winter 135 -10 -13 179 -12 -15 4468.3 0.6% 5.5% 0.32 7.0% 1.9% Spring 143 -9 -12 186 -11 -14 4112.8 0.5% 7.5% 0.29 5.2% -0.3%

Tomato

Early winter 127 -18 -23 202 -20 -24 10781.2 5.3% 9.2% 0.34 5.1% 4.5% Late winter 132 -13 -17 205 -18 -21 10228.9 -0.6% 3.3% 0.31 -1.0% -2.1% Early spring 140 -10 -12 210 -15 -18 8467.4 -6.2% -9.9% 0.27 -12.1% -17.4% Late Spring 155 -7 -8 220 -12 -13 5396.0 -26.3% -27.4% 0.19 -22.6% -23.9%

Potato Late autumn 86 -27 -33 131 -25 -31 3169.9 16.3% 23.7% 0.44 7.1% 7.2% Early winter 116 -21 -26 153 -21 -26 3569.7 18.5% 22.3% 0.37 10.0% 10.2% Late winter 142 -17 -21 174 -17 -22 3582.7 15.3% 20.1% 0.28 8.7% 5.9%

Grapevine Early 139 -5 -6 229 -7 -8 777.0 4.1% -1.9% 0.36 -0.8% -10.0% Medium 153 -6 -7 257 -8 -9 777.3 -7.9% -18.8% 0.30 -7.0% -17.6% Late 152 -2 -1 255 -4 -3 1416.8 -18.0% -29.6% 0.35 -10.1% -20.2%

Olive tree Early 129 -15 -19 239 -26 -32 853.4 1.4% 5.0% 0.54 6.4% 1.9% Late 159 -16 -20 303 -40 -49 999.2 -0.8% 1.8% 0.47 2.3% -2.2%

Table 9 – Crops’ response to the interaction among sowing seasons, precocity levels and future climate scenarios for Cyprus.



- - - - - - - - - Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Baseline

RCP4.5

RCP8.5

Crete

Wheat

Early autumn 66 -16 -18 154 -16 -19 5881.5 -6.9% 7.8% 0.44 3.8% 6.0% Late autumn 87 -13 -15 166 -14 -17 5513.2 -6.7% 8.2% 0.40 4.6% 6.9% Winter 130 -6 -7 192 -9 -9 4454.9 -6.8% 9.3% 0.28 -1.4% 2.3% Spring 139 -3 1 199 -7 -9 4365.5 -7.8% 7.9% 0.25 -5.5% -2.9%

Barley

Early autumn 83 -18 -21 147 -18 -22 6011.5 -6.1% 8.0% 0.47 7.0% 10.0% Late autumn 103 -16 -19 160 -16 -20 5659.5 -6.6% 8.7% 0.42 7.1% 10.2% Winter 144 -10 -13 189 -12 -15 4483.2 -5.0% 9.7% 0.29 1.7% 8.6% Spring 151 -10 -12 195 -12 -14 4278.3 -5.4% 9.3% 0.26 -0.9% 6.4%

Tomato

Early winter 140 -18 -23 217 -22 -26 11932.4 7.6% 9.3% 0.33 1.5% 8.4% Late winter 144 -14 -18 219 -20 -23 12070.3 6.2% 8.1% 0.30 -2.4% 4.7% Early spring 149 -11 -14 222 -18 -21 11528.6 1.6% 0.7% 0.28 -6.8% 0.1% Late Spring 161 -8 -9 231 -16 -18 9428.6 -23.6% -21.1% 0.25 -25.0% -16.2%

Potato Late autumn 104 -27 -31 150 -25 -30 2936.6 12.3% 22.2% 0.38 6.7% 10.4% Early winter 130 -21 -26 168 -21 -25 3059.6 10.3% 21.8% 0.33 6.6% 11.1% Late winter 154 -17 -21 186 -18 -22 3159.1 11.5% 23.1% 0.26 3.2% 8.1%

Grapevine Early 141 -5 -6 234 -9 -11 786.6 -3.7% 6.2% 0.37 -7.2% -4.1% Medium 159 -9 -11 265 -12 -15 1028.6 -9.5% -0.5% 0.36 -11.9% -7.9% Late 157 -6 -8 262 -9 -11 1755.7 -14.5% -6.2% 0.38 -13.3% -9.8%

Olive tree Early 141 -15 -20 261 -32 -38 898.3 2.5% 11.3% 0.51 6.9% 10.3% Late 171 -16 -20 337 -48 -58 1055.0 1.2% 9.6% 0.47 -3.1% 0.7%

Table 10 – Crops’ response to the interaction among sowing seasons, precocity levels and future climate scenarios for Crete


Yield Sowing season - Precocity Sicily Cyprus Crete (Kg/ha) - - - Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5

Wheat

Early autumn 0.0% 10.0% 12.0% 0.0% -1.2% 5.3% 0.0% -6.9% 7.8% Late autumn -3.7% 7.2% 7.9% -0.3% -1.7% 4.0% -6.3% -12.5% 1.4% Winter -20.3% -17.8% -14.7% -22.7% -24.5% -20.3% -24.3% -29.4% -17.2% Spring -24.6% -25.3% -21.6% -28.2% -31.1% -26.0% -25.8% -31.6% -19.9%

Barley

Early autumn 0.0% 10.5% 11.5% 0.0% -2.2% 3.9% 0.0% -6.1% 8.0% Late autumn -4.3% 7.0% 8.0% -2.4% -3.7% 1.9% -5.9% -12.0% 2.3% Winter -20.3% -16.3% -13.7% -29.5% -29.1% -25.6% -25.4% -29.2% -18.2% Spring -27.0% -25.9% -23.7% -35.1% -34.8% -30.3% -28.8% -32.7% -22.2%

Tomato

Early winter -0.4% 11.3% 13.2% 0.0% 5.3% 9.2% -1.1% 6.3% 8.1% Late winter 0.0% 8.5% 12.3% -5.1% -5.7% 3.3% 0.0% 6.2% 8.1% Early spring -7.0% -5.7% -3.6% -21.5% -26.3% -29.2% -4.5% -2.9% -3.8% Late Spring -36.1% -47.6% -47.1% -50.0% -63.1% -63.7% -21.9% -40.3% -38.4%

Potato Late autumn -3.6% 10.7% 14.2% -11.5% 2.9% 9.5% -7.0% 4.4% 13.6% Early winter -0.01% 16.6% 18.2% -0.4% 18.0% 21.8% -3.1% 6.8% 17.9% Late winter 0.00% 12.2% 15.8% 0.0% 15.3% 20.1% 0.0% 11.5% 23.1%

Grapevine Early -51.1% -54.3% -49.7% -45.2% -42.9% -46.2% -55.2% -56.8% -52.4% Medium -42.1% -48.8% -44.7% -45.1% -49.5% -55.5% -41.4% -47.0% -41.7% Late 0.0% -14.2% -8.3% 0.0% -18.0% -29.6% 0.0% -14.5% -6.2%

Olive tree Early -16.4% -17.8% -8.9% -14.6% -13.4% -10.3% -14.9% -12.7% -5.3% Late 0.0% -0.8% 7.9% 0.0% -0.8% 1.8% 0.0% 1.2% 9.6%

Table 11 – Yield variations (%) under current and future climate conditions using different sowing dates and varieties compared to the current maximum production.

ETA/ETP Sowing season - Precocity Sicily Cyprus Crete (mm) - - - Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5 Baseline RCP4.5 RCP8.5

Wheat

Early autumn 0.00% 3.47% 9.88% 0.00% 0.78% 0.92% 0.00% 3.81% 5.96% Late autumn -7.26% -5.55% 1.10% -9.15% -7.82% -10.83% -8.70% -4.46% -2.38% Winter -28.80% -31.05% -24.45% -38.01% -38.58% -42.71% -35.80% -36.69% -34.33% Spring -33.21% -36.67% -31.70% -45.33% -46.34% -50.38% -41.72% -44.91% -43.39%

Barley

Early autumn 0.00% 5.47% 13.06% 0.00% 5.09% 6.71% 0.00% 7.00% 10.01% Late autumn -8.73% -5.71% 1.90% -11.27% -4.39% -4.31% -10.20% -3.83% -1.02% Winter -31.17% -32.33% -24.54% -39.73% -35.50% -38.58% -37.93% -36.89% -32.61% Spring -35.38% -37.48% -30.56% -45.58% -42.77% -45.76% -43.52% -44.05% -39.89%

Tomato

Early winter 0.00% -2.14% 7.22% 0.00% 5.14% 4.54% 0.00% 1.51% 8.41% Late winter -5.16% -9.64% -1.91% -10.22% -11.15% -12.07% -8.87% -11.07% -4.60% Early spring -14.25% -17.28% -13.30% -22.36% -31.74% -35.88% -15.21% -20.97% -15.12% Late Spring -30.49% -41.35% -40.38% -45.17% -57.55% -58.26% -23.34% -42.52% -35.76%

Potato Late autumn 0.00% 8.21% 16.67% 0.00% 7.11% 7.20% 0.00% 6.67% 10.41% Early winter -12.67% -9.23% -1.18% -16.73% -8.43% -8.26% -13.53% -7.83% -3.91% Late winter -27.59% -29.17% -22.16% -36.68% -31.18% -32.97% -31.93% -29.74% -26.39%

Grapevine Early 0.00% -9.25% -2.84% 0.00% -0.83% -9.98% -4.04% -10.98% -7.94% Medium -14.99% -25.99% -21.07% -18.62% -24.36% -32.97% -7.50% -18.48% -14.84% Late -4.24% -16.03% -10.79% -5.05% -14.65% -24.21% 0.00% -13.27% -9.85%

Olive tree Early -1.69% -1.99% 3.07% 0.00% 6.38% 1.89% 0.00% 6.92% 10.28% Late 0.00% -7.14% -4.49% -11.42% -9.34% -13.36% -7.36% -10.20% -6.69%

Table 12 – ETA/ETP ratio variations (%) under current and future climate conditions using different sowing dates and precocity levels compared to the current maximum production.

4.5. Overall Crop Vulnerability

Crop vulnerability is measured by taking into consideration only the

yield fluctuations that will occur in the two scenarios RCP 4.5 and RCP 8.5

compared to the Baseline, as summarized in the following equations:

Performance Deficit = Performance baseline - Performance future

climate scenario (RCP 4.5 and RCP 8.5)

(1)

Vulnerability category (%) = Deficit/Performance baseline * 100 (2)

Examples:

1) Yield loss barley 4.5 = Yield baseline (sowing date 10 November) –

Yield at RCP 4.5 (sowing date 10 November);

Vulnerability category (%) = Yield loss 4.5 / Yield baseline (sowing

date 10 November) * 100

2) Yield loss tomato 8.5 = Yield baseline (sowing date 30 March) –

Yield at RCP 8.5 (sowing date 30 March)

Vulnerability category (%) = Yield loss 8.5 / Yield baseline (sowing

date 30 March) * 100

The categorization will refer to different level of performances (i.e. losses or

gains) and will include six categories: very high, high, medium, low and very

low. A first attempt of categorization is following proposed:

• Very high ≥ 80%

• 80 > High ≥ 60

• 60 > Medium ≥ 40

• 40 > Low ≥ 20

• 20 > Very low ≥ 0.

This scale will then be normalized to a numerical range i.e. 0 to 5, where 0

corresponds to no increase of vulnerability while 5 corresponds to the maximum

increase of vulnerability. The rationale is summarized in Table 8. This

procedure will be applied for each of the four sowing dates in the annual crops

and precocity levels for perennial crops.

Category Very high High Medium Low Very low Performance Deficit (X) X ≥ 80% 80 > X ≥ 60 60 > X ≥ 40 40> X ≥ 20 20 > X ≥ 0 Vulnerability level 5 4 3 2 1

Table 8. Vulnerability categories definition (Table 2).

5. Conclusions The Activity 4.2 ('Use of crop models for assessing the vulnerability of

agriculture to climate change'), focused on assessing the impacts of future

climate on agricultural sector over the three Mediterranean islands of Crete,

Cyprus and Sicily, has provided important indication about the main strategies

which could be used to cope with the forecasted impacts of climate on crop

productivity.

Our results showed as an increase in warm conditions can lead to

contrasting results depending on the location of the area (Crete, Cyprus and

Sicily), the sowing period (early or late autumn and winter), and the crop

analyzed. In general, the RCP4.5 scenario resulted with higher yields with

respect to RCP8.5. This response was consistent for all the crops expected for

grape which showed a generalized decrease of yield when temperature

increased.

These results, providing an indication of the level of vulnerability of the

studied crops, should be considered as a starting point to cope with issues on

climate change as well as the agronomic requests of farmers in the study areas.


Acknowledgements

This report was produced under co-finance of the EC LIFE programme for the Environment and Climate Action (2014-2020), in the framework of Action E.3 “Informative material” of the project LIFE ADAPT2CLIMA (LIFE14 CCA/GR/000928) “Adaptation to Climate change Impacts on the Mediterranean islands' Agriculture”.

The project is being implemented by the following partners:

National Observatory of Athens - NOA

Agricultural Research Institute - ARI

Institute of Biometeorology - IBIMET

National Technical University of Athens - NTUA

Department of Agriculture, Rural Development and Mediterranean Fisheries, Region of Sicily - SICILY

Region of Crete - CRETE


References

Alves A.C. and Setter T.L.: Response of cassava leaf area expansion to water deficit: cell proliferation, cell expansion and delayed development. Ann. Bot. 94, 605-613 (2004).

Barnabás B., Jäger K. and Fehér A.: The effect of drought and heat stress on reproductive processes in cereals. Plant Cell Environ. 31, 11-38 (2008).

Bindi, M., Miglietta, F., Gozzini, B., Orlandini, S., Seghi, L.: A simple model for simulation of growth and development in grapevine (Vitis vinifera L.). I. Model Description. Vitis 36, 67–71(1997a).

Bindi, M., Miglietta, F., Gozzini, B., Orlandini, S., Seghi, L.: A simple model for simulation of growth and development in grapevine (Vitis vinifera L.). II. Model validation. Vitis 36, 73–76 (1997b).

Bregaglio, S., Donatelli, M.: A set of software components for the simulation of plant airborne diseases. Environ. Model. Softw. 72, 426–444. (2015). doi:10.1016/j.envsoft.2015.05.011

Brilli, L., Moriondo, M., Ferrise, R., Dibari, C., Bindi, M., Climate change and Mediterranean crops: 2003 and 2012, two possible examples of the near future. Agrochimica, Vol. LVIII – Special Issue

Caffarra, A., Eccel, E.: Increasing the robustness of phenological models for Vitis vinifera cv. Chardonnay. Int. J. Biometeorol. 54, 255–267. (2010). doi:10.1007/s00484-009-0277-5

Cappelli, G., Bregaglio, S., Romani, M., Feccia, S., Confalonieri, R.: A software component implementing a library of models for the simulation of pre-harvest rice grain quality. Comput. Electron. Agric. 104, 18–24. (2014). doi:10.1016/j.compag.2014.03.002

Ciais P., Reichstein M., Viovy N., Granier A., Ogee J., Allard V. et al.: Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437, 529-523 (2005).

Copa-Cogeca: Assessment of the impact of the heat wave and drought of the summer 2003 on agriculture and forestry. Committee of Agricultural Organizations in the European Union and the General Committee for Agricultural Cooperation in the European Union, Brussels (2003).


https://www.researchgate.net/publication/293172791_Climate_change_and_Mediterranean_crops_2003_and_2012_two_possible_examples_of_the_near_future



Costa A.C. and Soares A.: Trends in extreme precipitation indices derived from a daily rainfall database for the South of Portugal. Int. J. Climatol. 29, 1956-1975 (2009).

Dalla Marta, A., Natali, F., Mancini, M., Ferrise, R., Bindi, M., Orlandini, S.: Energy and water use related to the cultivation of energy crops: A case study in the Tuscany region. Ecology and Society, 16 (2):2. (2011).

Donatelli, M., Bregaglio, S., Confalonieri, R., De Mascellis, R., Acutis, M.: A generic framework for evaluating hybrid models by reuse and composition - A case study on soil temperature simulation. Environ. Model. Softw. 62, 478–486. (2014) doi:10.1016/j.envsoft.2014.04.011

Durao R.M., Pereira M.J., Costa A., Delgado J., Del Barrio G. and Soares A.: Spatial temporal dynamics of precipitation extremes in southern Portugal: a geostatistical assessment study. Int. J. Climatol. 30, 1526-1537 (2010).

EEA 2012: Climate change, impacts and vulnerability in Europe 2012. An indicator based report. Report No.12 (2012).

Giorgi F.: Climate change hot-spots. Geophys. Res. Lett. 33, L08707 (2006).

Jalota, S.K., Vashisht, B.B., Kaur, H., Arora, V.K., Vashist, K.K., Deol, K.S.: Water and nitrogen-balance and -use efficiency in a rice (Oryza sativa) wheat (Triticum aestivum) cropping system as influenced by management interventions: Field and simulation study. Experimental Agriculture, 47 (4) 609-628. (2011).

JRC/IES: Available at http://mars.jrc.ec.europa.eu/mars/Bulletins-Publications (2012).

Kyselý J.: Recent severe heat waves in central Europe: how to view them in a long term prospect?. Int. J. Climatol. 30, 89-109 (2010).

Kostopoulou E. and Jones P.D.: Assessment of climate extremes in the Eastern Mediterranean. Meteorol. Atmos. Phys. 89, 69-85 (2005).

Leolini, L., Bregaglio, S., Moriondo, M., Ramos, M.C., Bindi, M., Ginaldi, F. (submitted). A model library to simulate grapevine growth and development: software implementation, sensitivity analysis and field level application. Eur J Agron. (Submitted).

Merante, P., Van Passel, S., Pacini, C.: Using agro-environmental models to design a sustainable benchmark for the sustainable value method. Agric. Syst. 136, 1 – 13. (2015).


Moriondo M., Leolini L., Brilli L., Dibari C., Tognetti R., Giovannelli A., Rapi B., Battista P., Caruso G., Gucci R., Argenti G., Raschi A., Centritto M., Bindi M. (under revision) A simple model simulating development and growth of an olive grove. Agricultural and Forest Meteorology (Under revision).

Moriondo, M. , Pacini, C., Trombi G., Vazzana, C., Bindi, M.: Sustainability of dairy farming system in Tuscany in a changing climate. European Journal of Agronomy 32, 80–90(2010)

Norrant C. and Douguédroit A.: Monthly and daily precipitation trends in the Mediterranean (1950-2000). Theor. Appl. Climatol. 83, 89-106 (2006).

Rodda J.C., Little M.A., Harvey J.E. and McSharry P.E.: A comparative study of the magnitude, frequency and distribution of intense rainfall in the United Kingdom. Int. J. Climatol. 30, 1776 1783 (2010).

Sousa P.M., Trigo R.M., Aizpurua P., Nieto R., Gimeno L. and Garcia-Herrera R.: Trends and extremes of drought indices throughout the 20th century in the Mediterranean. Nat. Hazard. Earth Sys. 11, 33-51 (2011).

Stella, T., Frasso, N., Negrini, G., Bregaglio, S., Cappelli, G., Acutis, M., Confalonieri, R.: Model simplification and development via reuse, sensitivity analysis and composition: A case study in crop modelling. Environ. Model. Softw. 59, 44–58. (2014). doi:10.1016/j.envsoft.2014.05.007

Stöckle, C.O., Donatelli, M., Nelson, R.: CropSyst, a cropping systems simulation model. European Journal of Agronomy 18, 289–307. (2003).

Stöckle, C.O., Nelson, R.L., Higgins, S., Brunner, J., Grove, G., Boydston, R., Whiting, M., Kruger, C.: Assessment of climate change impact on Eastern Washington agriculture. Climatic Change, 102 (1-2), 77-102. (2009).

Ulbrich U., Xoplaki E., Dobricic S., García-Herrera R., Lionello P., Adani M. et al.: Past and current climate changes in the Mediterranean region. In: Regional Assessment of Climate Change in the Mediterranean (Navarra A. and Tubiana L., eds.). Springer Science Business Media, Dordrecht (2013).

Vautard R., Yiou P., D’andrea F., De Noblet N., Viovy N., Cassou C. et al.: Summertime European heat and drought waves induced by wintertime Mediterranean rainfall deficit. Geophys. Res. Lett. 34, L07711 (2007).


Zolina O., Simmer C., Gulev S.K. and Kollet S.: Changing structure of European precipitation: Longer wet periods leading to more abundant rainfalls. Geophys. Res. Lett. 37, L06704 (2010).


future climate change vulnerability on the agricultural ... · deliverable c4.2: report on the...

Documents