lable at ScienceDirect

Environmental Modelling & Software xxx (2015) 1e15

Contents lists avai

Environmental Modelling & Software

journal homepage: www.elsevier .com/locate/envsoft

Modelling olive trees and grapevines in a changing climate*

Marco Moriondo a, Roberto Ferrise b, Giacomo Trombi b, Lorenzo Brilli b, Camilla Dibari b,Marco Bindi b, *

a CNR-IBIMET, via G. Caproni 8, 50145 Florence, Italyb Department of Agri-food Production and Environmental Sciences, University of Florence, P.le delle Cascine 18, 50144 Florence, Italy

a r t i c l e i n f o

Article history:Received 31 October 2013Received in revised form10 December 2014Accepted 19 December 2014Available online xxx

Keywords:Tree cropsClimate changeSimulation modelsCrop yield

* Thematic Issue on Agricultural Systems Modeling* Corresponding author. Tel.: þ390553288235; fax:

E-mail address: marco.bindi@unifi.it (M. Bindi).

http://dx.doi.org/10.1016/j.envsoft.2014.12.0161364-8152/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: MoriondoSoftware (2015), http://dx.doi.org/10.1016/j.

a b s t r a c t

The models developed for simulating olive tree and grapevine yields were reviewed by focussing on themajor limitations of these models for their application in a changing climate. Empirical models, whichexploit the statistical relationship between climate and yield, and process based models, where cropbehaviour is defined by a range of relationships describing the main plant processes, were considered.The results highlighted that the application of empirical models to future climatic conditions (i.e. futureclimate scenarios) is unreliable since important statistical approaches and predictors are still lacking.While process-based models have the potential for application in climate-change impact assessments,our analysis demonstrated how the simulation of many processes affected by warmer and CO2-enrichedconditions may give rise to important biases. Conversely, some crop model improvements could beapplied at this stage since specific sub-models accounting for the effect of elevated temperatures and CO2

concentration were already developed.© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Grapevine and olive tree represent worldwide key economicactivities with Europe representing the largest vineyard area in theworld (38%, Fraga et al., 2012) and olive tree widespread cultivatedaround the Mediterranean basin (Vossen, 2007). Besides theireconomic importance, the cultivation of these species provides anumber of services for the society, including landscape mainte-nance and improvement, enhancement of the quality of life, andecological and environmental services (Loumou and Giourga, 2003;Nieto et al., 2010).

Centuries of experience in the cultivation of these crops haveresulted in a strong association of both olive and grapevine growingwith geographically distinct regions in European countries (Joneset al., 2005; Moriondo et al., 2013a,b). Over these areas, olivetrees and grapevines are confined to specific climatic niches thatput them at greater risk from temperature and rainfall changes. Forinstance, rising temperatures, as predicted in future climate sce-narios, might alter the phenological responses of olive trees (i.e.timing of flowering, end of dormancy as well as minimum chilling

& Software.þ39055332472.

, M., et al., Modelling olive treenvsoft.2014.12.016

requirements) (Avolio et al., 2012), as well as a gradual northwardshift of current olive cultivation areas in the coming decades(Moriondo et al., 2013b; Tanasijevic et al., 2014). Extreme events(both in temperature and precipitation), as expected in futureclimate scenarios, could have a strong impact on grape growth andwine quality (Jones et al., 2005), especially if they occur during thegrowing season (Jones and Davis, 2000).

Olive grove and grapevine productivity, and consequentlyprofitability, depend on several factors including soil fertility,management practices, climate and meteorology. Since the profit-ability of a crop is a prerequisite for its cultivation, there is a sub-sequent need for a reliable assessment of the effects of a changingclimate on its yield and quality. This would help to define thepossible economic impact on the areas where they are currentlycultivated, as well as on the regions in which cultivation couldbecome viable in the future.

In this context, crop models are essential tools for investigatingthe effects of climate change on crop development and growth viathe integration of existing knowledge of crop physiology relating tochanging environmental conditions. This approach could provide aunique opportunity to assess possible effects of a warmer climateon crop yield and quality, as well as to evaluate different adaptationoptions for reducing or exploiting the effects of a changing climate.As such, crop models were extensively applied to assess the po-tential of various crops on farm, regional and national scales. In

es and grapevines in a changing climate, Environmental Modelling &

Fig. 1. Outline of a possible structure of a process-based model. Leaf intercepted ra-diation is converted into biomass that is allocated to different plant organs (leaves,stems, trunk and roots) according to the phenological stage. Abiotic stresses (nitrogenand water suboptimal conditions) affect this scheme e.g. by reducing photosyntheticefficiency and the pattern of biomass allocation.

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e152

particular, during the last 20 years, these tools were applied atdifferent time and spatial scales to assess the potential impact ofclimate change on crop productivity, thereby providing a reliablebenchmark to establish relevant adaptation options (White et al.,2011).

Despite their widespread use in studying crop productivity forboth the present and future scenarios, crop models have shownlimitations, whichmay simply be represented by the fact that whenapplied in similar conditions, different models often providedifferent results (Palosuo et al., 2011; R€otter et al., 2012; Assenget al., 2013). These differences are usually related to the specificapproach used to describe a certain process that may be eitherover-simplified or even incorrectly outlined (Eitzinger et al., 2013).

Based on these premises, this paper provide a review of differentmodels, either generically adapted to, or specifically developed forsimulating grapevine and olive tree growth and development byfocussing on the major limitations of these models, especially intheir application to climate change impact and adaptationassessments.

The analysis first considers the type of modelling approach, i.e.empirical or process-based by detailing the level of their parame-terization. Differences in the approaches adopted for the simulationof the main processes such as phenology, leaf-area growth,photosynthesis, biomass accumulation and partitioning, and waterand nutrient stress were further analysed for the process-basedmodel. The strengths and weaknesses of each simulated processwere evaluated and discussed in view of the recent advances incrop modelling. The analysis specifically outlined the following: i)the processes that are not yet simulated but may play an importantrole in climate-change impact assessment (e.g. the effect ofincreased CO2 on radiation and water-use efficiency, RUE, WUE)(Tubiello and Ewert, 2002); ii) the effects of extreme events on cropyield (Moriondo et al., 2011a); iii) the impact on yield quality.

2. Modelling approaches

The models selected for this analysis can be divided into twocategories, namely (i) empirical and (ii) process-based. In the firstcategory, statistical methods are used to generally describe a linearrelationship between the dependent (e.g. crop yield) and predictorvariables accounting for the fact that climate (e.g. climate indicessuch as cumulated rainfall or average temperature of a certainmonth) is one of the key factors influencing yield quantity andquality (Jones et al., 2005). Conversely, other processes like plantdevelopment, Leaf Area Index (LAI) growth, and nitrogen dynamicsin the soil are not considered. These models may be summarizedaccording to the following linear equation:

Yj ¼ aTji þ bRji þ/þ d

where Y is the yield quantity or quality, T and R are predictingvariables aggregated at different time scales (i) (usually monthly,seasonal or yearly) in the year (j), a, b are the regression co-efficients, and d is the intercept.

As such, empirical models need a limited amount of input datato produce an output but causes-effects mechanisms betweenclimate and yield are not explicitly described and this limits theapplicability of this approach to the specific regions or environ-mental conditions for which the relationships were calibrated.

Conversely, in process based models, crop behaviour is definedby a range of relationships describing the main plant processes.Crop models working at different levels of detail basically includethe simulation of 1) crop development, which indicates the periodby which crop growth should get started, the priorities in drymatter partitioning, and the maturity; 2) biomass accumulation,

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

which is a function of the amount of Photosynthetically ActiveRadiation (PAR) intercepted by the canopy and the efficiency viawhich the intercepted PAR is converted into dry matter. Dependingon the model, potential biomass accumulation may then bereduced by water, nitrogen stress, or heat stress; 3) biomass parti-tioning to the different organs, (leaves, stem, fruit, and roots) ac-cording to the priorities during the plant life cycle, providing afeedback for updating the simulation of leaf area extension and thegrowth of stem, fruit and roots (Fig. 1).

A specific sub-class of these models include functional modelsthat are oriented to simulate specific plant processes such as dailyCO2 exchange or water dynamics in the soil while sketching orneglecting other processes such as phenology, biomass partitioningor abiotic stresses.

Further, process-based models, originally designed to work onfield scale (Sinclair and Seligman, 1996), explicitly includes non-climatic factors, such as management practices (e.g. sowing dates,fertilization) or soils type, as determinants of crop yield.

As such, process basedmodels, may be easily applied to simulatecrop yield on a local scale, provided that climatic and non-climaticinformation are available at the required scale.

Based on these premises, the different approaches used tosimulate plant growth and development were compared separatelyfor each category of model, emphasizing the relevant limitations inthe light of new findings in the specific approach.

2.1. Empirical models

The driving variables used to fit empirical models for yieldprediction are generally temperature (T) and rainfall (R) aggregatedon amonthly timescale (Table 1). Despite the large variability of theenvironments where these models were parameterized, there is agenerally good agreement regarding the variables selected as thebest predictors of final yield. This implies that thesemodels are ableto detect the importance of critical stages in determining final yield,even when there is a rough resolution of the input variables.

Under this context, Nemani et al. (2001) correlated the positivetrend in grapevine yield of the Napa valley (1963e1996) to the

es and grapevines in a changing climate, Environmental Modelling &

Table 1Predictors of yield and quality of grapevine and olive tree according to different empirical models.

Author Crop Region Meteorologicalvariables

Output Season CC

Winter Spring Summer Fall

d j f m a m j j a s o n

Santos et al. (2011) Grapevine Porto (PT) T Yield þ 1R þ �

Jones and Davis (2000) Bordeaux (FR) T Yield 0R � � � �

Nemani et al. (2001) Napa Valley (USA) T Yield þ 0

RLobell et al. (2006) California (USA) T* Yield þ 1

R* þ þQuiroga and Iglesias (2009) Mediterranean basin T Yield � 0

R þNemani et al. (2001) Napa Valley (USA T Qual. 0

R þJones and Davis (2000) Bordeaux (FR) T Qual. þ þ þ þ þ 0

R � �Jones et al. (2005) Global scale T* Qual. þ 1

RMoriondo et al. (2011b) Chianti (IT) T* Qual. þ þ 1

R � �Quiroga and Iglesias (2009) Olive tree Mediterranean basin T Yield 0

R þ þ þ þOrlandi et al. (2013) Italy T Yield � 0

R þ þMoriondo et al. (2001) Tuscany (IT) T Yield � � � 0

R þ þ þ

Signs þ and � refers to correlation sign of the relationship between predictor variables and quality/yield; T ¼ temperature, R ¼ cumulated rainfall. G ¼ grapevine, O ¼ olivetree; * ¼ non-linear model equation. The last column (CC) indicates whether the model was developed for climate change impact assessment (0 ¼ not; 1 ¼ yes).

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 3

increasing minimum temperature (Tmin) observed in spring.Santos et al. (2011) found that high precipitation in early spring aswell as relatively low precipitation together with high tempera-tures in late spring increased grapevine yield in the Porto wineregion (Portugal). These findings partially overlap those of Jonesand Davis (2000) that showed an increase in precipitation duringthe budburst-flowering and veraison-harvest periods to be asso-ciated with lower yields. These responses are generally attributedto favourable/non-favourable events during the critical stages ofthe grapevine developmental cycle, which in turn determine apositive/negative impact on crop yield. For example, high precipi-tation during anthesis were reported to have a detrimental impacton yield, due to the effect of a lower fruit set and inflorescencedifferentiation (Jones and Davis, 2000).

Other studies demonstrated that climate conditions of previousyear may play a role as well in determining the final yield of thefollowing year. Lobell et al. (2006) found that minimum tempera-tures in April and cumulated rainfall in June and September prior toharvest are positively correlated to grapevine yield anomalies inCalifornia. The influence of the previous year, also observed byQuiroga and Iglesias (2009) is consistent with the Mediterraneanenvironments and is related to the accumulation of soil moistureduring the wet period that assures vegetative vigour at the onset ofthe season. Interestingly, Lobell et al. (2006) introduced a corre-sponding negative quadratic term for each predictive variable, ac-counting for the non-linear response of grape yield toenvironmental variables. This can play a fundamental role, espe-cially for impact assessment studies, when these empirical modelsare coupled to climate scenarios in which temperature and rainfallexceed those generally observed in the area.

Quiroga and Iglesias (2009) introduced an empiric modelconsidering both climate and technological variables (consumptionof nitrogen fertilizers, farm-machinery power, irrigation) in esti-mating grapevine yield. Climatic datawere introduced in themodel

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

using both monthly and seasonal aggregation periods. Yearlydrought explained a significant component of the inter-annualvariability of grapevine yield, with high temperatures having ageneral negative effect.

For olive trees, the same study pointed out that both manage-ment practices (irrigation and fertilization) and precipitationaround anthesis (April) and during fruit development have a pos-itive impact on final yield.

In addition to the meteorological variables, other studies alsoused air pollen concentration of olive trees as an explicative vari-able of yield formation (e.g. Moriondo et al., 2001; Gal�an et al.,2004, 2008). The pollen quantities can in fact be considered aproxy variable of the potential productivity of olive groves locatedaround the pollen traps (Moriondo et al., 2001), and in many casespollen quantity acts as the most important predictive parameter ofolive yield. Additional climatic information, recorded during theolive tree growth cycle, acts either to reduce or increase yield, butwith a lower predictive capacity compared to the biological variableexpressed by the pollen emitted.

For grapevines, climate undoubtedly plays a major role indetermining berry composition, and simple relationships betweenyield quality and climate parameters were exploited in severalpapers. Both temperature and rainfall have demonstrated to beprimary determinants of berry composition and grape quality.Quality itself in modelling approaches was defined in differentways, ranging from vintage quality, i.e. a quality score scaled be-tween a minimum and a maximum value, to wine ratings, andsugar and acid concentrations (Barnuud et al., 2013).

Nemani et al. (2001) pointed out how the decreasing trend infrost events in the 1963e1996 period was strongly associated withincreasing wine ratings. For the Bordeaux wine region, Jones andDavis (2000) used a numeric vintage quality rating as a depen-dent variable versus meteorological variables (i.e. cumulated R,averaged T and radiation) over each phenological-phase duration

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e154

(e.g. from bud break to anthesis, from anthesis to veraison) alongwith the number of events where a variable was above or below athreshold. High radiation levels during the bud break interval had apositive impact on yield quality, as well as the number of eventswith temperatures higher than 30 �C during the anthesis-veraisonphase. Conversely, rainfall during veraison caused a reduction inquality. The positive effect of radiation (RAD) and temperatureduring the bud break-veraison stage may be ascribed to high levelsof photosynthetic activity that induce complete maturation, whilethe negative impact of rainfall may be due to berry dilution and/ormoisture-related problems.

Similar results were obtained by Jones et al. (2005) andMoriondo et al. (2011b), indicating that the average T of thegrowing season is a valid predictor of vintage quality in many partsof the world, and in many cases this relationship is not linear.Moreover, Moriondo et al. (2011b) refined the approach for theChianti region (Italy) emphasizing how the negative effect ofrainfall in the months prior to harvest is possibly related to thedilution of berries during the ripening phase and the onset of fungaldiseases triggered by high humidity.

2.2. Process-based models

Different kinds of process-based models were developed forgrapevines and olive trees, ranging from crop models, which aim atsimulating the entire crop growth cycle, to functional models,which focus on specific processes such as plant gas exchange andwater dynamics in the soil. For grapevines, we cited Gutierrez et al.(1985), Wermelinger et al. (1991), Bindi et al. (1997), Pallas et al.(2011), STICS (Brisson et al., 2009), Cropsyst (St€ockle et al., 2003)and SUCROS (as modified for grapevines in Nendel and Kersebaum,2004), VineLogic (Godwin et al., 2002) and Cola et al. (2014). Forolive trees, a fewmodels were found in literature. These range fromcrop models (Abdel-Razik, 1989; Gutierrez et al., 2009) which insome cases simply rely on the relationship between interceptedradiation and biomass accumulation (i.e. Villalobos et al., 2006;Maselli et al., 2012), or are unbalanced with regard to the func-tional approach (Viola et al., 2012) which describe in detail leaf gasexchange.

Despite the heterogeneity in themodel structure (Table 2), somecomments and specific recommendations for each process may bemade, especially in view of the increasing temperatures and at-mospheric CO2 concentrations.

2.2.1. Crop developmentThe simulation of crop development is a fundamental step for

correct simulation of plant growth, where the timing of eachdevelopmental stage corresponds to a change in carbon assimila-tion, transpiration, partitioning of carbohydrates and nutrients. Assuch, the simulation of phenology is included in all the crop modelsanalysed (Bindi et al., 1997; CropSyst; STICS; Abdel-Razik, 1989;Nendel and Kersebaum, 2004), for ensuring a reliable simulationof growth of single plant organs or merely for setting the start ofanthesis (Maselli et al., 2012). Conversely, it is often neglected inmodels focussing on single processes such as light interception(Villalobos et al., 2013), leaf gas exchange (Poni et al., 2006; Violaet al., 2012), or water dynamics in the soil (Lebon et al., 2003;Celette et al., 2010).

The occurrence of each phenological stage (basically bud burst,anthesis and maturation) is based on the accumulation of tem-peratures above a threshold calculated on a daily (degree dayaccumulation, DDA) or hourly time step (e.g. Normal Heat Hours,NHH) up to a fixed amount, which is variety dependent. In Bindiet al. (1997), Ben-Asher et al. (2006), Abdel-Razik (1989) and Colaet al. (2014) the thermal forcing of DDA to bud burst starts on a

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

fixed day of the year. In STICS, and SUCROS for grapevine andMaselli et al. (2012) for olive the accumulation of temperature onlystarts after the end of dormancy and is triggered when a specificnumber of chill units has accumulated during winter. Cropsyst usesthe same approach but the end of dormancy occurs after sevenconsecutive days with temperatures above 10 �C (Fig. 2).

2.2.1.1. Issues related to climate change. Reviewed literature in-dicates that robust phenological models are required to give morecertainty to extrapolations of global-change scenarios (Chuine,2000; Caffarra and Eccel, 2010) since adaptation relies also onphenological performances in response to a warmer climate(Duchene et al., 2010). Until now, great effortswere taken to developand test models to predict bud-break for both grapevines and olivetrees. This stage, while representing the beginning of the growingseason and hence the accumulation of biomass, is also important asan indicator of variety earliness and therefore, adaptability todifferent climates (Jones et al., 2005; Garcia de Cortazar-Atauri et al.,2009; Webb et al., 2007). For grapevines, the results of Caffarra andEccel (2010) which compare the performance of a number ofphenologicalmodels inpredicting budburst, indicated that process-based models accounting for the effect of the chilling unit ondormancy release (Unified Model for Bud Burst, Chuine, 2000)provide the highest performance for the Chardonnay variety inNorthern Italy. In particular, rises in temperature after fulfilment ofthe critical chill requirement help delay the bud-burst stage byincreasing the sum of accumulated forcing units.

These results partially overlap those of Garcia de Cortazar-Atauriet al. (2009, 2010), which conclude that even though the calculationof dormancy release using chill units is not a critical factor forimproving model performance in this period, it may assume moreimportance given the expected rise in temperature in the future. Asimilar approach based on chill days and a subsequent forcing stagewas described as effective for olive trees in predicting both budburst (Cesaraccio et al., 2004) and bud burst and anthesis (DeMelo-Abreu et al., 2004).

According to thesefindings, awarmer climatewould give rise toapostponing rather than a advancingof the budburst date. In anycasethe advance of this phase would not be proportional to the increasein temperature. Conversely, phenological models based solely on athermal time accumulation (i.e. Bindi et al., 1997; Ben-Asher et al.,2006; Abdel-Razik, 1989), would undoubtedly overestimate thiseffect in a warmer climate due to being linearly related to temper-ature. When applying these models in future climate scenarios,increasing temperatures would gradually bring the occurrence ofthe phenological stages forward to unrealistic dates (Fig. 2). This, inturn, would further reduce the time for biomass accumulation,overestimating the expected impact of climate change on yield.

2.2.2. Leaf-area growth and intercepted radiationIn many cases, leaf area is empirically simulated as dependent

on temperature. Bindi et al. (1997) calculates leaf emission rate asrelated to average daily temperature and finally simulates total leafarea using an empirical relationship between the number ofappeared leaves and the shoot-leaf area. Similarly, in STICS, theleaf-growth rate is dependent on the temperature and phenologicalstages. Lebon et al. (2003) used a simplified grapevine canopymodel consisting of a parallelepiped, the vertical sides of which canbe penetrated by the light, where the increase in canopy di-mensions and the relevant intercepted radiation are linearly relatedto cumulative thermal time. Cola et al. (2014) used the sameapproach as modified in Oyarzun et al. (2007). In Poni et al. (2006),the leaf-area expansion rate is dependent on the degree days andaccounts for the effect of the management practices (trimmedshoots). Gutierrez et al. (1985), Wermelinger et al. (1991), Nendel

es and grapevines in a changing climate, Environmental Modelling &

Table 2List of models under analysis for olive tree and grapevine.

Model Literature Light int.a Light useb Yieldc Biomasspart.d

Phenol.e Root distr.f Environ.stressg

Waterdyn.h

ETi SoilCN-modelj

Effect of highCO2

kClim. inputvariablesl

Cropmanag.m

Varn Applico Pr p

Grapevine Vinelogic Godwin et al. (2002) S RUE B/Gn S/L/F/R T/DL/CU EXP W/N/S/Trp C PT/PM CN/P(4)/B RUE/WUE R/Tx/Tn/Rd P/TS P 1STICS Brisson et al., 2009 S/3D RUE Prt/B S/F/L/R T/DL/CU/O LIN W/N/Trp C P/PM/

PT/SWN/P/B RUE/WUE R/Tx/Tn/Rd/RH/W P/D/TS/I 8 2 F

Bindi et al. (1997) S RUE Prt/B F T LIN W/Trp C P RUE R/Tx/Tn/Rd D 2 FCola et al. (2014) 3D P-R Prt/B S/L/F/R T LIN W/Trp/CC C P P R/Tx/Tn/Rd/CC TS,/T 1

NVINE Nendel and Kersebaum (2004) D P-R Prt/B S/L/F/R T/DL/CU EXP W/N/Trp C P N R/Tx/Tn/Rd P 1 FCropsyst St€ockle et al. (2003) S RUE/WUE HI_mw S/L T/DL/O LIN W/N/Trp C PM/PT N RUE/WUE R/Tx/Tn/Rd/RH/W 1 C

Pallas et al. (2011) S RUE Prt/B S/L/F/R T LIN W C R/Tx/Tn/Rd P 1 GSWAP Ben-Asher et al. (2006) S P-R Prt/B S/L/R T Call W/S R PM R/Tx/Tn/Rd 1 FVIMO Wermelinger and

Baumg€artner (1990)S RUE Prt/B S/L/F/R T N/Trp N Tx/Tn/Rd TS 1 3

Gutierrez et al. (1985) S RUE Prt/B S/L/F/R T Trp Tx/Tn/Rd TS 6 4Lebon et al. (2003) 3D LIN C P R/Tx/Tn/Rd/RH/W TS 1

WALIS Celette et al. (2010) 3D LIN C P R/Tx/Tn/Rd/RH/W TS/I 1 EPoni et al. (2006) S P-R Trp Tx/Tn/Rd T,/TS 3 5 S

Olive tree Viola et al. (2012) S P-R HI_mw F LIN W/Trp C PM P R/Tx/Tn/Rd/RH/W 2Maselli et al. (2012) S RUE HI F T/CU W/Trp P RUE R/Tx/Tn/Rd EAbdel-Razik (1989) D P-R Prt/B S/L/F/R T Trp R/Tx/Tn/Rd D 1Villalobos et al. (2006) S RUE HI F Tx/Tn/Rd DGutierrez et al. (2009) S RUE Prt/B S/L/F/R T/CU Trp C R R/Tx/Tn/Rd/RH/W 6

a Light interception model. S, simple approach (e.g. LAI); D, detailed approach (e.g. canopy layers); 3D three dimensions model.b Conversion of intercepted radiation into dry matter. RUE, radiation use efficiency approach; P-R, gross photosynthesis e respiration; WUE, water use efficiency.c Conversion of biomass into final yield. HI, fixed harvest index; B, total (above-ground) biomass; Prt, partitioning during reproductive stages; HI_mw, harvest index modified by water stress.d Biomass partitioning. Dry matter is allocated to: S, stem; L, leaves, F, fruit; R, storage organs.e Phenology driving variables. T, temperature; DL, photoperiod (day length); CU, chill units to the end of dormancy; O, other water/nutrient stress effects considered.f Model describing root growth. LIN, linear, EXP, exponential, SIG, sigmoidal, Call, carbon allocation; O, other approaches.g Stress factors reducing potential biomass. W, water limitation; N, nitrogen limitation; A, aeration deficit stress; H, heat stress; Trp, temperature response of photosynthesis (or RUE) and respiration; S, salinity; CC, cloud

coverage.h Description of water dynamics in the soil. C, capacity approach; R, Richards approach.i ET models. P, Penman; PM, Penman-Monteith; PT, PriestleyeTaylor; R, Ritchie, SW, Shuttleworth and Wallace.j CN, CN model; N, N model; P(x), x number of organic matter pools; B, microbial biomass pool.k Effect of high CO2 on physiological parameters. RUE, radiation use efficiency; WUE, water use efficiency; P, Photosynthesis model.l Climate variables driving the model. Cl, cloudiness; R, rainfall; Tx, maximum daily temperature; Tn, minimum daily temperature; Ta, average daily temperature; Td, dew point temperature; Rd, radiation; e, vapour pressure;

RH, relative humidity; W, wind speed; CC, cloud coverage.m Crop management. D planting density; P, pruning; T shoot trimming with a feedback to leaf area growth; TS, trellis system; I, intercropping.n Number of varieties calibrated.o Application of the model elsewhere than in the calibration site: 1) Webb et al. (2007), Zhang et al. (2002); 2) Garcia de Cort�azar Atauri (2006), Vald�es-G�omez et al. (2009); 3) Wermelinger et al. (1991); 4) Williams et al.

(1985); 5) Orlandini et al. (2008); 6 Ponti et al. (2014).p Programming language used (when applicable). F, Fortran, C, Cþþ; S, Stella environment; E, Excel; G, GreenLab environment.

M.M

oriondoet

al./Environm

entalModelling

&Softw

arexxx

(2015)1e15

5

Pleasecite

thisarticle

inpress

as:Moriondo,M

.,etal.,M

odellingolive

treesand

grapevinesin

achanging

climate,Environm

entalModelling

&Softw

are(2015),http://dx.doi.org/10.1016/j.envsoft.2014.12.016

Fig. 2. a. Change in the end-of-dormancy day (EOD) in response to increasing tem-peratures using a chill unit (CU) approach (De Melo-Abreu et al., 2004; modified),compared to Cropsyst for grapevines, where EOD is set when 7 consecutive days withTavg>10 occur. When using CU, EOD is progressively delayed as higher temperatureslower the CU accumulation process. When using thermal forcing as in Cropsyst, EODmay only be advanced. Calculations were made for Florence (43.15 Lat N, 10.5 Lon E,80 m asl). b. Change in the anthesis day in response to increasing temperatures usingdegree day accumulation (DDA) after the EOD based on the CU for olive trees (De Melo-Abreu et al., 2004; modified), DDA for olive trees (Moriondo et al., 2001) and grape-vines (Cropsyst) and average temperature for grapevines (Bindi et al., 1997) as drivingvariables. Models only considering the thermal forcing (Bindi et al., 1997; Moriondoet al., 2001; Cropsyst) show how the anthesis day is progressively advanced inresponse to higher temperatures. The progressive delay of the EOD in model using CU(De Melo-Abreu et al., 2004; modified) reduces the advancing of the anthesis bydelaying the EOD as in A. Calculations were made for Florence (43.15 Lat N, 10.5 Lon E,80 m asl).

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e156

and Kersebaum (2004) CropSyst, Abdel-Razik (1989) and Ben-Asher et al. (2006) use a process-based approach where the LAI iscalculated as the product of biomass partitioned to the leaves(g d�1) and the specific leaf-areaweight (SLA cm2 g�1). In Villaloboset al. (2006), Viola et al. (2012) and Maselli et al. (2012) the LAI andintercepted radiation are used respectively as model's inputs todrive the simulation.

In all models, except those accounting for the three-dimensionalshape of the crop canopy, intercepted radiation is implementedusing the BeereLambert production function corrected by theprojected leaf-area in order to account for non-continuous distri-bution of the grapevine canopy.

2.2.2.1. Issues related to climate change. Literature indicates thatprimordial initiation and leaf production of grapevine are linearlyrelated to temperatures (Lebon et al., 2004; Buttrose, 1969) andDDA or temperature seem a simple but rather robust approach tomodel leaf area growth, in consideration of the fact that this rela-tionship is valid under a wide range of temperatures. However,Lebon et al. (2004) suggest that this rate may decrease during theseason in response to water stress or to a strong competition

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

between vegetative and reproductive parts of the shoot. Accord-ingly, area growth reducing factors were introduced to account forstresses (Nitrogen or water) on leaf area growth (STICS, Bindi et al.,1997). In this perspective, the use of biomass partition coefficientsto leaves would provide a more effective tool to represent the feed-back between reduced biomass accumulation triggered by abioticstresses and reduced leaf area growth.

There are major concerns regarding the simulation of the effectsof extreme weather events such as prolonged heat waves ordrought periods that are likely to increase both in frequency andintensity in the near future. Until now, models have never simu-lated the effects of extreme events on plants (e.g. leaf loss) nor thefollowing recovery process. After a reduction in the assimilatorysurface, recovery of grapevines was triggered by increasing thephotosynthetic activity of the remaining leaves and redirecting theassimilate flow towards the defoliation zone (Chanishvili et al.,2005), whereas carbon stored in perennial organs supported fruitgrowth (Candolfi-Vasconcelos et al., 1994).

For olive tree, LAI is empirically approximated according theplanting density and canopy volume (Villalobos et al., 2006), orforced into the model using observed data (i.e. NDVI) while onlyAbdel-Razik (1989) and Gutierrez et al. (2009) use a process-basedapproach based on biomass partitioning. The empirical approaches,as in the previous case, limit to account for the possible feed backsof adverse weather conditions on leaf area growth.

2.2.3. Biomass accumulationBiomass accumulation is calculated according to the amount of

active photosynthetic radiation (PAR) intercepted by the canopythat is translated into biomass (B) using a parameter or functiondescribing the efficiency of this process. The light or Radiation UseEfficiency (RUE, g�1 MJ�1 m�2) is a simple empirical coefficientdescribing the linear dependence of B on PAR, providing an efficientway to estimate the net dry matter production per unit of absorbedenergy [i.e. gross primary production e (maintenance and growthrespirations)]. This approach was extensively applied in semi-empirical crop modelling approaches, including grapevines (Bindiet al., 1996; Pallas et al., 2011; Gutierrez et al., 1985; Wermelingeret al., 1991) and olive trees (Villalobos et al., 2006; Maselli et al.,2012).

RUE is also used in complex models such as STICS (Brisson et al.,2003). Other approaches, such as CropSyst (St€ockle et al., 2003)relies on the use of Water Use Efficiency, i.e. the amount of biomassassimilated per unit of water transpirated (WUE, g�1 m�3).

Poni et al. (2006) calculated the net photosynthetic rate as thedifference between the gross daily integral of canopy photosyn-thesis and the respiration rates for leaves, stems and clusters. Thesame scheme is applied in Nendel and Kersebaum (2004) forgrapevines and Abdel-Razik (1989) for olive trees. In Viola et al.(2012) biomass accumulation is obtained by applying the photo-synthetic model proposed by Farquhar et al. (1980).

2.2.3.1. Issues related to climate change. The use of RUE wasextensively applied in the models under analysis. This coefficient isconservative in non-limiting conditions but when the plants areoutside this range, the RUE was found to be reduced: water stress,causing a reduction in leaf gas exchange and also in CO2 exchange,results in reduced RUE. Similarly, nitrogen limiting conditions mayalso represent an important source of variation in RUE (Sinclair andMuchow,1999). These effects must be taken into account especiallywhen designing the cultivation of olive trees and grapevines in thefuture, where increasing temperatures and reduced rainfall arepredicted over most of the grapevine and olive growing areas(Moriondo et al., 2013a,b).

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 7

At the same time, higher CO2, as expected in the future, wasobserved to increase both RUE andWUE (Tognetti et al., 2001; Bindiet al., 2001) (Table 2). Therefore, the interaction between the pos-itive and negative effects of a changing climate should be consid-ered for reliable simulations.

Simple models partially reproduce the balance between thepositive and negative effects of climate change on photosyntheticefficiency. For grapevines, Bindi et al. (1997) found that RUE de-creases for suboptimal temperatures and water in the soil, whereasthe RUE is set to þ30% due to a doubling of the CO2 concentration(700 ppm) (Fig. 3a). In non-limiting water conditions, Poni et al.(2006) only set a reduction in assimilation starting from 31 �Cwhereas the effect of higher CO2 was not included. The effect ofwater stress was considered in Lebon et al. (2003), Bindi et al.(1997) and Viola et al. (2012) where the assimilation efficiency isdependent on the water available in the soil according to a bilinearresponse function. A similar approach was used in Maselli et al.(2012) based on a simplified water balance for olive trees, wherea CO2 fertilization factor (Veroustraete et al., 2004) was used toaccount for the effect of increased CO2 on the RUE. In all these cases,no specific adjustment was found for nutrient-limiting conditions.

Complex crop models have different approaches where theinteraction between high CO2 and water/nitrogen stress is orga-nized more efficiently. In CropSyst, unstressed biomass accumula-tion previously calculated using increased RUE/WUE is thencorrected for water or Nitrogen limitation. In STICS, the effect ofstress conditions directly affects plant growth by reducing the

Fig. 3. a. Effects of a rise in temperature on the photosynthetic efficiency according todifferent models (see Table 1 for references). b. Effects of increasing water stress on leaftranspiration as in Lebon et al. (2003), Bindi et al. (1997) and STICS (Brisson et al.,2009) assuming transpiration to be proportional to the gas exchange capability ofthe plant, used the same model to modify RUE in relation to water stress.FTSW ¼ fraction of total soil water, i.e. the amount of water in the soil with respect tothe water available between the filed capacity and the wilting point.

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

maximum RUE (Brisson et al., 2003) whereas the effect of increasedCO2 is accounted for by increasing the RUEwith specific coefficients(Tubiello et al., 2000; Brisson et al., 2003) (Fig. 3b). In Nendel andKersebaum (2004), photo-assimilation can be restricted either bywater stress, determined by the ratio of actual to potential tran-spiration, or by nitrogen deficiency.

The interaction between rising temperatures combined withhigh CO2 adds to the complexity in simulating the plant responsesin future scenarios (Schultz and Stoll, 2010) since these variableswere shown to have a synergistic effect on photosynthesis (Sageand Kubien, 2007). This was confirmed for grapevines in Schultz(2000) and Salazar-Parra et al. (2012a, b), where the optimumtemperature for photosynthesis shifted in response to higher CO2concentrations (Fig. 4 and Table 3).

2.2.4. Biomass partitioningIn semi-empirical models, biomass partitioning to different

plant organs (root, stems, leaf and fruit) is generally poorly simu-lated or even neglected. For instance, for grapevines Bindi et al.(1997) only considered biomass partitioned daily to fruits, whilefor olive trees Villalobos et al. (2006), Maselli et al. (2012) and Violaet al. (2012) only used a fixed HI to be applied to the total cumu-lated biomass at the end of the season.

In functional models that focus more on the simulation of aparticular process rather than on the entire plant growth (i.e. gasexchange in Poni et al., 2006; or water dynamics in the soil in Ben-Asher et al., 2006; Celette et al., 2010; Lebon et al., 2003), only thedaily biomass accumulation is calculated and the growth of singleplant organs is not considered.

In complex models, partitioning is simulated with different ap-proaches. In Cropsyst andNendel andKersebaum (2004), drymatteris partitioned amongst the various organs according fix partitioningfactors that are function of phenological development stage of crop.Gutierrez et al. (1985), Wermelinger et al. (1991) and Abdel-Razik(1989) STICS and Pallas et al. (2011) used source/sink approachwhere biomass is partitioned to different organs according to therelevant ability to attract available assimilated from source.

2.2.4.1. Issues related to climate change. Biomass partitioning tosingle organs may be implemented via fixed proportions that only

Fig. 4. Effects of increasing leaf temperature coupled with a different CO2 concen-tration on the grapevine leaf photosynthetic rate (Schultz, 2000). Data were fitted to asecond order polynomial equation. The analysis of the first derivative of these equa-tions indicated that the optimum temperature for photosynthesis rises from 31.4 �C(CO2 ¼ 364 ppm) to 33 �C (CO2 ¼ 605 ppm).

es and grapevines in a changing climate, Environmental Modelling &

Table 3Effect of elevated CO2 on photosynthetic efficiency in grapevines and olive trees.

Treat. Photosynthetic efficiency

Reference Crop CO2 ambient CO2 550 ppm CO2 700 ppm

Moutinho-Pereira et al. (2006). Grapevine (var. Touriga franca) 13.65 mmol CO2 m�2s�1 22.46a mmol CO2 m�2s�1

Salazar-Parra et al. (2012a, b). Grapevine (var. Tempranillo) WI 10.1 mmol CO2 m�2s�1 (þ4)a16.7 mmol CO2 m�2s�1

PI 5.2 mmol CO2 m�2s�1 (þ4)a12.9 mmol CO2 m�2s�1

Bindi et al. (2001). Grapevine (var Sangiovese) 1.0 g MJ�1 1.3a g MJ�1 1.3a g MJ�1

Tognetti et al. (2001). Olive (var. Moraiolo) 11.5 mmol CO2 m�2s�1 16.5a mmol CO2 m�2s�1

Olive (var. Frantoio) 13.2 mmol CO2 m�2s�1 17.3a mmol CO2 m�2s�1

Melgar et al. (2008). Olive (var Koroneiki) 17.84 mmol CO2 m�2s�1 27.04a mmol CO2 m�2s�1

Olive (var Picual) 21.08 mmol CO2 m�2s�1 26.06 mmol CO2 m�2s�1

WI ¼ well irrigated; PI ¼ partially irrigated, (4) ¼ þ4 �C with respect to control treatment at 350 ppm.a Significantly different from the control at 350 ppm. In Melgar et al. (2008) and Salazar Parra et al. (2012a,b) the maximum differences between treatments during the

season were reported.

Fig. 5. Leaf mass in relation to total above ground grapevine biomass grown at350 ppm and 700 ppm (Bindi et al., 1997; unpublished data). Treatment and controlfollow the same trend until the beginning of fruit growth. Afterwards, plants growingat 700 ppm partitioned more biomass to leaves compared to the control plant. Thismay be the effect of a reduced fruit sink strength in grapevine var. Sangiovese, that islimited to 2 grapes per shoot. The excess of carbon is then stored in leaves and, likely,in roots. The regression slopes are significantly different for P < 0.05.

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e158

vary according to the phenological stage. Conversely, the change indistribution of resources amongst different plant organs inresponse to stress plays a key role in the fitting capacity of plants(Grechi et al., 2007). As observed for many species (Keller andKoblet, 1995; Zerihun and Treeby, 2002), plants provide more re-sources to the organs responsible for the uptake of a limitingresource compared to non-limiting conditions. This was observedin limiting water and nitrogen conditions for both grapevines andolive trees. For grapevines, Bota et al. (2004) found that water stressmay alter the assimilate distribution pattern especially towards thetrunk and roots with a reduced contribution to clusters. Grechi et al.(2007) provided evidence that suboptimal nitrogen applicationaffected leaf canopy development while a larger portion of avail-able assimilates translocated from the leaves into the root system(i.e. the organ responsible for balancing the correct plant C/N ratio).Similar results were obtained in Rodríguez-Lovelle and Gaudill�ere(2002), who observed an increased root/shoot ratio in grapevinesin low nitrogen conditions.

In olive trees, Boussadia et al. (2010) observed that limiting ni-trogen conditions promote biomass allocation to favour root elon-gation. In drought conditions, Bongi (1994) concluded that watershortage increases carbon partitioning used for root formation,thereby reducing the carbon available for aboveground canopygrowth. These results are confirmed in Celano et al. (1997) and Sofoet al. (2008) who found an increasing root/canopy ratio that wasusually greater in non-irrigated plants as a mechanism to enhancewater uptake efficiency.

Crop growth simulation models should therefore be improvedusing dynamic allocation rules, especially in view of the expectedfuture climate. From this point of view, the models using a source-sink approach provide a better framework to implement the dy-namics in biomass partitioning according to the specific growingconditions.

The effect of CO2 concentration on biomass partitioning ingrapevines and olive trees is an additional issue. For instance,Norby et al. (2002) indicated that in FACE experiments on a forestecosystem, the biomass partition coefficients changed in responseto higher CO2. In these conditions, the extra carbon is mainlyallocated to leaves and fine roots, as also suggested by other authorsfor different ecosystems (e.g. Luo et al., 1996). The results of Bindiet al. (2001) highlight a significant effect of CO2 enrichment onbiomass partitioning of grapevines, where plants at 700 ppmshowed a higher efficiency of carbon partition to leaves comparedto control (350 ppm) conditions (Fig. 5).

2.2.5. ETP and water stressPET is generally calculated according to the approaches of

Penman-Monteith as reported in Allen et al. (1998) that is con-verted into evapotranspiration in compliance with the coefficients

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

accounting for the crop development stages (e.g. cultural co-efficients Kc or the fraction of intercepted crop radiation; Fig. 6a). InSTICS, PET is calculated using the Penman-Monteith model sepa-rately applied for the soil and crop surfaces (Shuttleworth andWallace, 1985), accounting for the fact that grapevine do not pro-vide a close and uniform canopy. Similarly, in some model, PET isempirically partitioned into daily transpiration and soil evaporationwhich are converted into actual ET according to specific thresholdswhich reduces the process from the relevant compartments. (e.g.Lebon et al., 2006; Celette et al., 2010).

Bindi et al. (2001), assuming WUE as invariant for a wide rangeof environmental conditions (e.g. Amir and Sinclair, 1991), calcu-lated crop transpiration as the product of WUE and daily simulatedbiomass accumulation.

In those models considering the effect of water stress, crop ETPis converted into actual transpiration using reducing factors thatare dependent on soil moisture conditions relating to availableplant water (roughly the amount of water included between thefield capacity and the wilting point).

2.2.5.1. Issues related to climate change. Increasing CO2 concentra-tions and warmer temperatures have an opposite effect on ETP.While higher CO2 causes a decrease in stomatal conductance,transpiration per unit surface is decreased. By contrast, higher airtemperatures increase transpiration by increasing the vapourpressure deficit (VPD), i.e. the differences between the actual

es and grapevines in a changing climate, Environmental Modelling &

Fig. 6. a. Most crop ET models are based on the FAO approach that uses potentialevapotranspiration calculated for a reference grass surface (ET0) scaled during theseason to crop ET (ETc) using a cultural coefficient (Kc), which is the ratio between ETc/ET0. Kc is basically a function of leaf area development during the season (black line),and a shift in plant phenology in response to higher temperatures (i.e. advanced budbreak and delayed leaf loss) would result in a different Kc trend shape during theseason (dashed line). An increasing CO2 concentration is expected to further modifythe Kc during the season which could create imbalance in the ratio between ET0 andETc leading to differences in maximum Kc. In the graph we hypothesised that higherCO2 could reduce crop transpiration at a faster rate with respect to grass, leading to alower maximum Kc. b. Response of ETc to the joint effect of increasing temperaturesand CO2 concentrations according to different models: the FAOmodel and Cropsyst arebased on the same approach, but Cropsyst introduces a factor to account for the effectof CO2 on transpiration. Bindi et al. (1997), calculated ETc by assuming a constant WUEso that plant transpiration ¼ daily cumulated biomass (g ss m�2) * WUE (m3 H2O g�1

m�2). The model in Villalobos et al. (2013), labelled as olive tree considers the increaseconsiders the increase of both CO2 and temperature, and plant transpiration. Specif-ically CO2 at 500 and 700 ppm was set to increase RUE by 15% and 30%, respectivelyand crop intercepted radiation was set to 50% of incoming PAR. Calculations wereperformed for Follonica (Italy, 40� Lat N, 10.2 Lon E) experimental site.

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 9

moisture in the air and the moisture at saturation. While the effectof temperature is accounted for in commonly used ETP models,only STICS, Cropsyst and Viola et al. (2012) included algorithms toaccount for the effects of CO2 on stomatal conductance (Fig. 6b).

Additional issues are related to the use of the fixed crop co-efficients (KCs) to calculate crop ET (Allen et al., 1998). This impliesthat the response of crop ET and ETP (referred to a grass surface) toweather variables is always the same, whereas in a CO2-enrichedworld the response of a crop or a grass may not be proportional(Villalobos et al., 2013). Accordingly, new KCs should be calculateddepending on the ratio between crop and potential evapotranspi-ration measured at higher CO2 concentrations and changes in cropdevelopment should be considered as well (Lovelli et al., 2010)(Fig. 6a).

The approach in Bindi et al. (1997), which relies on a constantWUE and increased RUE, results in an overestimation of cropevapotranspiration (Fig. 6b).

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

The approach presented in Villalobos et al. (2013) allows forovercoming the missing effect of carbon dioxide on transpirationWUE and is more easily implementable due to a reduced number ofinput parameters compared to the FAO approach. This model pre-sents the transpiration rate as dependent on the ratio betweencanopy-intercepted radiation, radiation-use efficiency and CO2concentration in the boundary layer. Accordingly, the proposedmodel could consider both the changes in canopy conductance bymodifying the relevant coefficients in proportion to the atmo-spheric CO2 concentration and the expected changes in radiation-use efficiency (Fig. 6b).

2.2.6. Grape qualityThe general overview shows that while the simulation of yield

quality should be a highly desirable model output for grapevines,this is not considered in grapevine simulation models but only inSTICS, which exploits an empirical relationship between berry-water content and sugar concentration (Garcia de Cortazar-Atauriet al., 2009). Conversely, additional process-based approachesspecifically developed to simulate berry quality (i.e. sugar content,reviewed in Dai et al., 2010), were not yet dynamically included intoa crop growth simulation model. The general framework of thesemodels, focuses on the simulation of processes involved in sugarimporting, partitioning, and accumulation in berries.

2.2.6.1. Issues related to climate change. While the berry sugarcontent has a genetic fingerprint (Liu et al., 2007), the climaticenvironment and management practices (pruning, fruit thinning,green tipping, leaf removal, and fertilization) also play an importantrole and the interaction between these components is a challengefor grapevine growers, especially in the context of global change.For example, leaf removal during ripening may increase the tem-perature of the grape leading to early sugar ripeness, whereas acidsare depleted, causing unbalanced wines. This trend is likely to beexacerbated in a warmer climate where the effect of increased CO2should also be considered since it was demonstrated to have asignificant effect on the concentration of sugar and acid in the grape(Bindi et al., 2001). From this point of view, the empirical modelsuch as in Garcia de Cortazar-Atauri et al. (2009) (included inSTICS), that links sugar content directly to berry water status, mayfail to reproduce the effects of increased CO2 assimilation onassimilate partitioning to the fruit and the expected increase inberry sugar content.

The use of functional models, linking biological responses toclimate factors, is therefore recommended to obtain a reliableprediction of climate change impact on grape quality (Dai et al.,2009) in a warmer, CO2-enriched environment. As outlined in Daiet al. (2009), the hypothesis of coupling quality models to thosesimply predicting final yield does not seem to be realistic at thepresent time given the mismatch between the complexity of theprocesses simulated in crop growth and the functional models.

2.2.7. Canopy management practicesFor grapevines, different management practices can easily be

introduced in growth models by modifying the number of shootsper plant and the number of plants per square metres, which inturnmodifies intercepted crop radiation. Models accounting for thethree-dimensional shape (i.e. STICS, Lebon et al., 2003, Cola et al.,2014; Celette et al., 2010) may provide a better benchmark to testthe effects of different trellis systems on crop growth as they allowfor selecting different width to height ratios of the crop and therelevant effect on the shadowing among rows. For olive trees, theseeffect were studied e.g. in Connor (2006), but such an approachwasnot yet included in a crop growth model.

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e1510

The simulation of trimmed grapevine shoots was considered inPoni et al. (2006) while the same model was modified in Orlandiniet al. (2008) to account for the effect of different trellis systems ongrapevine biomass accumulation, while the effect of crown shapeof olive tree was given a little attention in literature.

2.2.7.1. Issues related to climate change. The manipulation of cropshape according to different training systems has an influence onmany processes such as carbon assimilation and water balance, andthe study of these responses as well as their modellingmay providemore insight into the adaptation of olive trees and grapevines to awarmer, drier climate.

Intercepted radiation, as influenced by grapevine geometry, is infact a major determinant of grapevine water status (Williams andAyars, 2005), where the systems intercepting more radiation (e.g.a canopy arranged on two parallel wires) generally result in ahigher transpiration rate with respect to those arranged on a singlecurtain (Candolfi-Vasconcelos et al., 1994; Orlandini et al., 2008).The same may apply to the CO2 exchange rate of the whole canopy,which was found to be directly related to the amount of canopylight interception (Poni et al., 2003), where trellis systems withhigher intercepted radiation tend to have a higher assimilation rateand a higher yield per vine (Orlandini et al., 2008). Accordingly, themodelling of these responses would allow for selecting a trainingsystem to balance water loss in relation to carbon gain in a climatechange scenario.

Olive trees have a different canopy structure and plantingstrategy that in any case aim atmaximising the interception of solarradiation with an appropriate distribution of radiation within thetree canopies.

The models presented in this review, while having the potentialto include the effect of different shapes on canopy light interceptionand evapotranspiration, may fail to reproduce the effects that thesechanges have on grapevine growth. The only attempt to comparethe effects of different trellis systems on grapevine growth wasfound in Orlandini et al. (2008), demonstrating that the crop modelwas not able to reproduce higher biomass accumulation asobserved in the lyre trellis system (double curtain) versus the singlecurtain.

According to these observations, an important part of themodelling should be geared towards studying the effects of changesin canopy shape on photosynthetic rate, water relations and canopymicroclimate, with particular attention to the new approachesdeveloped in the last few years (Matese et al., 2014). The use ofstructural-functional models that integrate a virtual representationof the canopy considering its shape and leaf and shoot orientation,may simulate the physiological space-plant processes such astranspiration or photosynthesis. This coupled representation,already available for grapevines (Louarn et al., 2008), makes itpossible to study the effects of a different canopy architecture onlight interception, and thus on canopy energy balance, providingthe expected feedback on carbon assimilation and water demand(Schultz and Stoll, 2010).

2.2.8. Soil organic carbonMore complex models, STICS, Cropsyt and SUCROS, include the

simulation of soil organic matter dynamics by considering thecontribution of crop residue in the soil and the mineralisation ofsoil organic matter. In general the rate of crop residue decompo-sition is negatively dependent on the C/N ratio, while the miner-alisation rate of soil organic matter is positively dependent on soilmoisture and temperature.

2.2.8.1. Issues related to climate change. The implementation ofsub-models to describe the dynamics of soil organic matter in crop

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

models mainly focuses on assessing the kinetics of nitrogen toprovide a feed-back on plant growth while the effects of soilorganic matter on soil structure and properties are not usuallyconsidered. In a climate change context, the effects of soil organicmatter on soil structure and consequently on soil hydraulic prop-erties should be considered fundamental for improving the per-formance of the models.

More importantly, an increase in CO2 concentration is likely tochange not only the total biomass of plants and their relevantresidue, but also the composition of this material. FACE experi-ments on woody plants indicated that N concentration of leavestends to decrease at high CO2 concentrations (Kimball et al., 2002),thus increasing the C/N ratio. This effect should be included in thesimulation of residue decomposition as the C/N ratio affects itsmineralization rate.

3. Discussion

3.1. General overview

Under the present review, empirical and process-based modelswere considered asmutually exclusive because these are developedfor different purposes and therefore there is a mismatch betweentheir underlying approaches.

Empirical models, do not require a heavy load of input param-eters, and the final yield or quality are the only outputs required forthese models. They can be easily calibrated by regression ap-proaches with climate over large areas and are applicable on a largespatial scale using gridded climatic datasets. Conversely, since theunderlying mechanisms that transform climatic input into yield orquality are not explicitly described, the use of these models shouldbe restricted to the range of climate data from which they aredeveloped (Landau et al., 2000).

The description of physiological processes provides good pre-dictive performance especially on a local scale, where all requiredinputs should be available and environmental conditions (soil type,rainfall pattern, agricultural practices, e.g. irrigation, fertilization)are usually well defined.

While it should be noticed that these approaches cannot actu-ally be completely separated since in any case process-basedmodels contains a certain degree of empirism (e.g. related to theapproximation of specific processes or fitting of equation co-efficients), for olive tree and grapevine empirical and process basedmodelling represent two different lines of research with anymutual inter-relationship.

Conversely, some examples in the literature provide the evi-dence that empirical and process based approaches may find acommon field of application, where the results of one may improvethe simulation of the other one. Challinor et al. (2003) developed acombined system where the successful fitting of an empiricalmodel, which links seasonal weather to yield, to a specific spatialscale, provide information on which scale a process based modelshould operate. Landau et al. (2000) used a multi-regressiveapproach to predict wheat yield as function of climate variables,which were aggregated over five phenological stages as deter-mined by a phenological sub-model.

Interestingly, Challinor et al. (2003) highlight that the level ofcomplexity of process-based model (i.e. the number of processesconsidered and their modelling approach) is inherently related tothe working scale where this model is intended to be used. Whenlooking at climate change impact assessment, whose spatial scaleranges from regional to continental areas, the use of complexprocess-based models is particularly unrealistic given the largeinput data required for these models. Conversely, the use ofsimplified process-based models, i.e. considering only the

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 11

processes having an impact on yield, are more appropriate to workon larger scale reducing, at the same time, the gap between ap-proaches with high (i.e. process-based) and low (i.e. empiricapproach) input data requirements.

However, so far large area process-based models for olive treeand grapevine is missing, and this reduces the applicability of thesemodels for climate change impact assessment over wide scales. Anexception is represented by the models of Bindi et al. (1996) andGutierrez et al. (2009) that simulating the basic processes involvedin olive and grapevine yield, matches the indications of Challinoret al. (2003) for a low-input process-based model to be appliedfor impact studies on large scales. These simulations are of aparticular interest specifically for climate change impact assess-ment from regional to continental scale, where the outputs of theprocess-based models (i.e. crop yield) are used in turn input data tooutline possible shift in cultivated area (Ferrise et al., 2015), or tofeed economic models (Moriondo et al., 2011a,b; Zhu et al., 2014;Ponti et al., 2014). This kind of framework allowed to outlinepossible effects of a warming climate on farmer's income and topropose reliable adaptation options. These studies represent a firststep towards an integrated environmental modelling approach forgrapevine and olive tree, where crop modelling is a part of a morecomprehensive system, which in a holistic fashion, include theformulation of an environmental issue, the identification of thesocial-economic-environmental components of the relevant sys-tem, and an integrated modelling methodology to provide infor-mation synthesis (Laniak et al., 2013).

3.2. Specific recommendation

Our analysis showed how the formulation of empirical models isstill lacking in important statistical approaches and predictors,especially when applied in a warmer climate. Major drawbacksinclude the following:

1) Yield quantity and quality are closely related to climatic indicescalculated over specific phenological phases. Since the devel-opment of grapevines and olive trees is a predominantlytemperature-driven process, it is expected that in a changingclimate, warmer temperatures will likely shift the occurrence ofphenological stages, as well as shortening the duration of thegrowing cycles (Webb et al., 2007; Ferrise et al., 2013). This mayaffect or even invalidate the significance of the models cali-brated for this period and the empirical models could fail tosimulate the correct impact of a warmer climate. The coherenceof the approach could be improved by using a phenologicalmodel simulating for each scenario the shift of phenologicalevents (e.g. Landau et al., 2000).

2) The effect of extreme weather events should be carefullyconsidered when empirical models are applied to estimate yieldin future decades. Temperature and rainfall aggregated on amonthly time step may fail to reveal information on extremedaily atmospheric events, i.e. events falling outside the rangegenerally experienced by the crop during its growing cycle.Extremely high or low temperatures as well as intense rainfall orhail, in particular during sensitive growth stages, may limit yieldand yield quality even with general favourable weather condi-tions for the rest of the season. For instance, temperatures above35 �C during flowering and throughout the growth of the grapeberries were reported to have a detrimental impact on finalquality due to premature veraison, high grape mortality, andpartial or total failure of flavour compound formation (Mullinset al., 1992). In addition, Santos et al. (2011) ascribed the poorsimulation of grapevine yield in the Douro region in particular

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

years to missing information about extreme events such as frostor hail occurrence.

3) Prediction of the impact of a warmer climate may be biased bythe linear approach used to model the effect of explanatoryclimatic variables on yield and quality. Jones et al. (2005) andMoriondo et al. (2011b) offer evidence that in many cases anoptimum temperature threshold for ripening exists, abovewhich final quality starts decreasing. The same approach wasapplied by Lobell et al. (2006) to model grapevine yield. Todate, these are the only attempts to introduce a non-linearrelationship between climate and yield or quality in grape-vines, although these dynamics were already demonstrated fora number of crops including corn and soybean (Schlenker andRoberts, 2009), wheat, maize and barley (Lobell and Field,2007).

4) Finally, the projected increase in CO2 concentration is expectedto have a positive effect on grapevine growth and yield (e.g.Bindi et al., 2001), and also olive yield (e.g. Tognetti et al., 2001).Therefore, it should be included as a predictive variable. How-ever, the fertilization effect of CO2 together with the increasedwater use efficiency was never considered in statistical ap-proaches at least for grapevine and olive.

For process based models, we noticed that the perennial natureof olive trees and grapevines is rarely considered and thereforebiomass partitioning to different organs and the relevant contri-bution of old wood or roots in simulating the growth process isnot well represented. Conversely, perennial structures must beconsidered as they provide reserves of carbon and nutrients thatmay play a role in plant recovery or support initial growth. Assuch, these buffering reserves may be particularly important in awarmer climate, where extreme drought or heat events can lead toleaf loss and the recovery process plays a fundamental role inassuring the final yield. More generally, when consideringperennial crops, each year growth and yield are dependent inseveral ways on the previous year's growing conditions. By way ofexample, drought or heat stress during anthesis in olive trees mayresult in a reduced floret fertility or fruit set. While reducing cropyield in the current year, this gives rise to increased production inthe following year, triggering alternate (fruit) bearing for manyyears. In this perspective, more developed models such as STICSand SUCROS, as modified by Nendel and Kersebaum (2004) orPallas et al. (2011), may provide more insight into the issue ofclimate change by including the carry-over effects in theirsimulation.

In any case, the effects of increasing temperatures and CO2concentration as expected in the near future are generally poorlyrepresented in the models under analysis. Major concerns may besummarized as follows:

1. Phenology is usually poorly described and this affects reliablesimulations of crop cycle duration (i.e. the time for biomassaccumulation) in a warmer climate in the processes dependingon phenology (e.g. changes in biomass partitioning). Chill unitrequirements for dormancy release are generally not consid-ered, or modelled with too simple an approach (Cropsyst) forboth olive trees and grapevines, leading to a progressiveadvancement of phenological stages that may actually not occurin response to increasing temperatures. Despite the importanceof this traits for climate change adaptation, observed data ofdormancy break are still lacking thus limiting the developmentor refining of specific modelling approaches (Garcia de Cortazar-Atauri et al., 2009). Moreover, the lengthening of the growingperiod (i.e. the time from bud break to leaf fall or dormancy) asobserved in response to increasing temperatures of plants

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e1512

(Menzel et al., 2006), is not currently considered even though itcould play a role in increasing the capacity of crops to storebiomass in perennial organs.

2. While the effects of increased CO2 and temperature on photo-synthetic efficiency are accounted for in most models or mayeasily be implemented by modifying RUE or the value ofphotosynthetic efficiency, their positive interaction on assimi-lation is neglected especially because there is a lack of experi-mental results on these specific crops. This may lead to asystematic underestimation of biomass accumulation in a futureclimate scenario.

3. Evapotranspiration models currently implemented may fail toreproduce the combined effects of increases in temperature andCO2. While a rise in temperature would lead to a higherevapotranspiration rate, this trend may be counterbalanced by adecrease in canopy conductance in response to higher CO2.Furthermore, when using the FAO approach, changes in KCsshould be accounted for because of the effect of a change in theoccurrence of phenological stages and in the new ratio betweenET0 and ETP.

4. Drought stress or nitrogen stress may play a role in changingthe biomass partitioning pattern to increase root growth butin most cases biomass partitioning consists of fixed co-efficients to plant organs. In addition, an increase in CO2 wasdemonstrated to change biomass partitioning in woody plantsbut further experiments in should be performed for grapevineand olive tree, however further experiments should be per-formed to assess new partitioning coefficients to perennialorgans. Change in canopy management practices such aspruning and trellis systems may help to reduce the impact ofincreasing temperature and reduced soil moisture and modelsmay easily be modified to account for different shapes andplant distances. However, current modelling approachesmight be not able to fully reproduce the actual crop physio-logical responses.

5. Vintage quality is a pre-requisite to obtain premium wines tocompete on international markets. The simulation of mostimportant grape quality traits (i.e. sugar and acid concentration)should be therefore pursued to test the impact on climatechange.

6. The simulation of soil organic carbon may provide more insightinto the analysis of the impact of olive tree and grapevine agro-ecosystems on the accumulation/depletion of soil organic car-bon. Agricultural management of perennial crops such asgrapevines and olive trees has the potential to increase theaccumulation of soil organic matter, thereby capturing CO2 fromthe atmosphere (Nieto et al., 2010). The simulation of soilorganic matter dynamics is therefore of maximum importancein evaluating the impact of these crops on the global carboncycle in future scenarios.

7. To improve the reliability of process-based models in awarmer climate, the uncertainties related to the use ofdifferent models as applied for different varieties or cultiva-tion areas, should be firstly evaluated for the present period.Frameworks like those presented in COST 734 for modelintercomparison should be therefore adopted for a firstevaluation of olive tree and grapevine model performancesunder various climatic conditions. However, comprehensiveexperimental datasets to compare observed and simulateddata are scarce. Experimental data as presented in Bindi et al.(1997) (Free Air CO2 Enrichment) and Sadras and Soar (2009)(large scale open-top heating system) for grapevine, whileproviding more insight the effect of a warmer climate on thiscrop, may represent, at the same time, a robust dataset to testthe reliability of crop modelling in a future climate.

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

4. Conclusions

The present critical review of empirical and process-basedmodels for grapevines and olive trees highlighted the main weakpoints in each group, especially in view of possible application tofuture scenarios. The mere assumption on which empirical modelsrely, namely, that the climate is driving the growth process and canbe used as a predictor of yield and quality is true, have only a localvalidity Furthermore, they are generally unreliable for theirpossible application in a changing climate unless to consideradditional predictive variables and improved statistical approaches.Conversely, process-based models may be improved to account forthe effect of higher temperatures and CO2, but the high level ofdetails required to run these models restricts their use to a localscale. Further, the missing or poor simulation of olive tree andgrapevine agro-ecosystems carbon balance does not allow to studythe possible beneficial effects of these cultivations to the globalcarbon cycle, which conversely is an information highly required toevaluate the capacity of these ecosystems to capture C thuscontributing to the mitigation strategies.

As a matter of fact, these limitations reduce the applicability ofthese tools for impact studies that at the moment is limited to a fewcases where crop and economic modelling are merged to provide acomprehensive evaluation of the impact of climate change fromregional to sub-continental scale. A model intercomparison wouldbe therefore highly recommended for testing the applicability ofprocess-based models in different regions and in response to awarmer climate.

Acknowledgements

The authors would like to gratefully acknowledge theconstructive comments provided by the four anonymous refereesthat improved the quality of the original paper.

References

Abdel-Razik, M., 1989. A model of the productivity of olive trees under optionalwater and nutrient supply in desert conditions. Ecol. Model. 45 (3), 179e204.http://dx.doi.org/10.1016/0304-3800(89)90081-1.

Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration -Guidelines for Computing Crop Water Requirements. FAO Irrigation andDrainage Paper 56. FAO, Rome, Italy. http://www.fao.org/docrep/X0490E/x0490e00.htm.

Amir, J., Sinclair, T.R., 1991. A model of water limitation on spring wheat growth andyield. Field Crop Res. 28 (1e2), 59e69. http://dx.doi.org/10.1016/0378-4290(91)90074-6.

Asseng, S., Ewert, F., Rosenzweig, C., Jones, J.W., Hatfield, J.L., Ruane, A.C., Boote, K.J.,Thorburn, P.J., Rotter, R.P., Cammarano, D., Brisson, N., Basso, B., Martre, P.,Aggarwal, P.K., Angulo, C., Bertuzzi, P., Biernath, C., Challinor, A.J., Doltra, J.,Gayler, S., Goldberg, R., Grant, R., Heng, L., Hooker, J., Hunt, L.A., Ingwersen, J.,Izaurralde, R.C., Kersebaum, K.C., Muller, C., Naresh Kumar, S., Nendel, C.,O'Leary, G., Olesen, J.E., Osborne, T.M., Palosuo, T., Priesack, E., Ripoche, D.,Semenov, M.A., Shcherbak, I., Steduto, P., Stockle, C., Stratonovitch, P., Streck, T.,Supit, I., Tao, F., Travasso, M., Waha, K., Wallach, D., White, J.W., Williams, J.R.,Wolf, J., 2013. Uncertainty in simulating wheat yields under climate change.Nat. Clim. Change 3, 827e832. http://dx.doi.org/10.1038/Nclimate1916.

Avolio, E., Orlandi, F., Bellecci, C., Fornaciari, M., Federico, S., 2012. Assessment of theimpact of climate change on the olive flowering in Calabria (southern Italy).Theor. Appl. Climatol. 107 (3e4), 531e540. http://dx.doi.org/10.1007/S00704-011-0500-2.

Barnuud, N.N., Zerihun, A., Gibberd, M., Bates, B., 2013. Berry composition andclimate: responses and empirical models. Int. J. Biometeorol. 58, 1207e1223.http://dx.doi.org/10.1007/S00484-013-0715-2.

Ben-Asher, J., Van Dam, J., Feddes, R.A., Jhorar, R.K., 2006. Irrigation of grapevineswith saline water: II. Mathematical simulation of vine growth and yield. Agric.Water Manag. 83 (1), 22e29. http://dx.doi.org/10.1016/j.agwat.2005.11.006.

Bindi, M., Fibbi, L., Gozzini, B., Orlandini, S., Miglietta, F., 1996. Modelling the impactof future climate scenarios on yield and yield variability on grapevine. Clim. Res.7, 213e224.

Bindi, M., Fibbi, L., Miglietta, F., 2001. Free Air CO2 Enrichment (FACE) of grapevine(Vitis vinifera L.): II. Growth and quality of grape and wine in response to

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 13

elevated CO2 concentrations. Eur. J. Agron. 14 (2), 145e155. http://dx.doi.org/10.1016/S1161-0301(00)00093-9.

Bindi, M., Miglietta, F., Gozzini, B., Orlandini, S., Seghi, L., 1997. A simple model forsimulation of growth and development in grapevine (Vitis vinifera L.). II. Modelvalidation. Vitis-Geilweilerhof 36, 73e76.

Bongi, P.A., 1994. Olive. In: Handbook of Environmental Physiology of Fruit Crops:Temperate Crops, pp. 165e187.

Bota, J., Stasyk, O., Flexas, J., Medrano, H., 2004. Effect of water stress on partitioningof 14C-labelled photosynthates in Vitis vinifera. Funct. Plant Biol. 31 (7),697e708. http://dx.doi.org/10.1071/FP03262.

Boussadia, O., Steppe, K., Zgallai, H., Ben El Hadj, S., Braham, M., Lemeur, R., VanLabeke, M.C., 2010. Effects of nitrogen deficiency on leaf photosynthesis, car-bohydrate status and biomass production in two olive cultivars ‘Meski’and‘Koroneiki’. Sci. Hortic-Amsterdam 123 (3), 336e342. http://dx.doi.org/10.1016/j.scienta.2009.09.023.

Brisson, N., Launay, M., Mary, B., Beaudoin, N., 2009. Conceptual Basis, Formaliza-tions and Parameterization of the STICS Crop Model. Quae, Versailles.

Brisson, N., Gary, C., Justes, E., Roche, R., Mary, B., Ripoche, D., Zimmer, D., Sierra, J.,Bertuzzi, P., Burger, P., Bussiere, F., Cabidoche, Y.M., Cellier, P., Debaeke, P.,Gaudillere, J.P., Henault, C., Maraux, F., Seguin, B., Sinoquet, H., 2003. An over-view of the crop model STICS. Eur. J. Agron. 18, 309e332. http://dx.doi.org/10.1016/S1161-0301(02)00110-7.

Buttrose, M.S., 1969. Vegetative growth of grapevine varieties under controlledtemperature and light-intensity. Vitis 8, 280e285.

Caffarra, A., Eccel, E., 2010. Increasing the robustness of phenological models forVitis vinifera cv. Chardonnay. Int. J. Biometeorol. 54 (3), 255e267. http://dx.doi.org/10.1007/s00484-009-0277-5.

Candolfi-Vasconcelos, M.C., Koblet, W., Howell, G.S., Zweifel, W., 1994. Influence ofdefoliation, rootstock, training system, and leaf position on gas exchange ofPinot noir grapevines. Am. J. Enol. Vitic. 45 (2), 173e180.

Celano, G., Dichio, B., Montanaro, G., Nuzzo, V., Palese, A.M., Xiloyannis, C., 1997.Distribution of dry matter and amount of mineral elements in irrigated andnon-irrigated olive trees. Acta Hortic. 474, 381e384.

Celette, F., Ripoche, A., Gary, C., 2010. WaLISdA simple model to simulate water parti-tioning in a crop association: the example of an intercropped vineyard. Agric.WaterManag. 97 (11), 1749e1759. http://dx.doi.org/10.1016/j.agwat.2010.06.008.

Cesaraccio, C., Spano, D., Snyder, R.L., Duce, P., 2004. Chilling and forcing model topredict bud-burst of crop and forest species. Agric. For. Meteorol. 126 (1), 1e13.http://dx.doi.org/10.1016/j.agrformet.2004.03.002.

Chanishvili, S.S., Badridze, G.S., Barblishvili, T.F., Dolidze, M.D., 2005. Defoliation,photosynthetic rates, and assimilate transport in grapevine plants. Russ. J. PlantPhysiol. 52 (4), 448e453 doi: 1021-4437/05/5204.

Challinor, A.J., Slingo, J.M., Wheeler, T.R., Craufurd, P.Q., Grimes, D.I.F., 2003. Towardsa combined seasonal weather and crop productivity forecasting system:determination of the spatial correlation scale. J. Appl. Meteorol. 42, 175e192.

Chuine, I., 2000. A unified model for budburst of trees. J. Theor. Biol. 207 (3),337e347. http://dx.doi.org/10.1006/jtbi.2000.2178.

Cola, G., Mariani, L., Salinari, F., Civardi, S., Bernizzoni, F., Gatti, M., Poni, S., 2014.Description and testing of a weather-based model for predictingphenology, can-opydevelopment and sourceesink balance inVitis vinifera L. cv. Barbera. Agric. For.Meteorol. 184, 117e136. http://dx.doi.org/10.1016/j.agrformet.2013.09.008.

Connor, D.J., 2006. Towards optimal designs for hedgerow olive orchards. Aust. J.Agric. Res. 57, 1067e1072. http://dx.doi.org/10.1071/AR05448.

Dai, Z.W., Vivin, P., Barrieu, F., Ollat, N., Delrot, S., 2010. Physiological and modelingapproaches to understand water and carbon fluxes during grape berry growthand quality development: a review. Aust. J. Grape Wine Res. 16, 70e85. http://dx.doi.org/10.1111/j.1755-0238.2009.00071.x.

Dai, Z.W., Vivin, P., Robert, T., Milin, S., Li, S.H., G�enard, M., 2009. Model-basedanalysis of sugar accumulation in response to sourceesink ratio and watersupply in grape (Vitis vinifera) berries. Funct. Plant Biol. 36 (6), 527e540. http://dx.doi.org/10.1071/FP08284.

De Melo-Abreu, J.P., Barranco, D., Cordeiro, A.M., Tous, J., Rogado, B.M.,Villalobos, F.J., 2004. Modeling olive flowering date using chilling for dormancyrelease and thermal time. Agric. For. Meteorol. 125 (1e2), 117e127. http://dx.doi.org/10.1016/j.agrformet.2004.02.009.

Duchene, E., Huard, F., Dumas, V., Schneider, C., Merdinoglu, D., 2010. The challengeof adapting grapevine varieties to climate change. Clim. Res. 41 (3), 193.

Eitzinger, J., Thaler, S., Schmid, E., Strauss, F., Ferrise, R., Moriondo, M., Bindi, M.,Palosuo, T., R€otter, R., Kersebaum, K.C., Olesen, J.E., Patil, R.H., Saylan, L.,Çaldag, B., Çaylak, O., 2013. Sensitivities of crop models to extreme weatherconditions during flowering period demonstrated for maize and winter wheatin Austria. J. Agric. Sci. 1 (1), 1e23. http://journals.cambridge.org/abstract_S0021859612000779.

Farquhar, G.D., von Caemmerer, S., Berry, J.A., 1980. A biochemical model ofphotosynthetic CO2 assimilation in leaves of C3 species. Planta 149 (1), 78e90.http://dx.doi.org/10.1007/BF00386231.

Ferrise, R., Moriondo, M., Trombi, G., Miglietta, F., Bindi, M., 2013. Climate changeimpacts on typical Mediterranean crops and evaluation of adaptation strategiesto cope with. In: Navarra, A., Tubiana, L. (Eds.), Regional Assessment of ClimateChange in the Mediterranean. Springer, Dordrecht, The Netherlands.

Ferrise, R., Trombi, G., Moriondo, M., Bindi, M., 2015. Climate change and grape-vines: a simulation study for the Mediterranean Basin. J. Wine Econ. 1e17.http://dx.doi.org/10.1017/jwe.2014.30 (in press).

Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., Santos, J.A., 2012. An overview ofclimate change impacts onEuropeanviticulture. FoodEnergy Secur.1 (2), 94e110.

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

Gal�an, C., García-Mozo, H., V�azquez, L., Ruiz, L., Díaz De La Guardia, C., Domínguez-Vilches, E., 2008. Modeling olive crop yield in Andalusia, Spain. Agron. J. 100 (1),98e104. http://dx.doi.org/10.2134/agrojnl2006.0345.

Gal�an, C., V�azquez, L., Garcıa-Mozo, H., Domınguez, E., 2004. Forecasting olive (OleaEuropaea) crop yield based on pollen emission. Field Crop Res. 86 (1), 43e51.http://dx.doi.org/10.1016/S0378-4290(03)00170-9.

García de Cort�azar-Atauri, I., 2006. Adaptation du mod�ele STICS �a la vigne (Vitisvinifera L.). Utilisation dans le cadre d’une �etude du changement climatique �al’�echelle de la France. PhD thesis of Ecole Sup�erieur Nationale d’Agronomie deMontpellier. Available at: http://www.inra.fr/ea/sources/index.php?page¼detail_these&id¼200.

Garcia de Cortazar-Atauri, I., Chuine, I., Donatelli, M., Parker, A.K., van Leeuwen, C.,2010. A curvilinear process-based phenological model to study impacts ofclimate change on grapevine (Vitis vinifera L.). In: Proceedings of Agro 2010: the11th ESA Congress, Montpellier, France (Agropolis International Editions:Montpellier), pp. 907e908.

Garcia de Cortazar-Atauri, I.G., Brisson, N., Ollat, N., Jacquet, O., Payan, J.C., 2009.Asynchronous dynamics of grapevine (Vitis vinifera) maturation: experi-mental study for a modeling approach. J. Int. Sci. de la Vigne du Vin 43 (2),83e97.

Godwin, D., White, B., Sommer, K., Walker, R., Goodwin, I., Clingeleffer, P., 2002.VineLOGIC e a model of grapevine growth, development and water use. In:Dundon, C., Hamilton, R., Johnstone, R., Partridge, S. (Eds.), Managing Water.Australian Society of Viticulture and Oenology Inc., Adelaide, pp. 46e50.

Grechi, I., Vivin, P.H., Hilbert, G., Milin, S., Robert, T., Gaudill�ere, J.P., 2007. Effect oflight and nitrogen supply on internal C: N balance and control of root-to-shootbiomass allocation in grapevine. Environ. Exp. Bot. 59 (2), 139e149. http://dx.doi.org/10.1016/j.envexpbot.2005.11.002.

Gutierrez, A.P., Ponti, L., Cossu, Q.A., 2009. Effects of climate warming on Olive andolive fly (Bactrocera oleae (Gmelin)) in California and Italy. Clim. Change 95(1e2), 195e217.

Gutierrez, A.P., Williams, D.W., Kido, H., 1985. A model of grape growth anddevelopment: the mathematical structure and biological considerations. CropSci. 25, 721e728.

Jones, G.V., Davis, R.E., 2000. Climate influences on Grapevine phenology, grapecomposition, and wine production and quality for bordeaux, France. Am. J. Enol.Vitic. 51 (3), 249e261.

Jones, G.V., White, M.A., Cooper, O.R., Storchmann, K., 2005. Climate change andglobal wine quality. Clim. Change 73 (3), 319e343. http://dx.doi.org/10.1007/s10584-005-4704-2.

Keller, M., Koblet, W., 1995. Dry matter and leaf area partitioning, bud fertility andsecond season growth of Vitis vinifera L.: responses to nitrogen supply andlimiting irradiance. Vitis 34 (2), 77e83.

Kimball, B.A., Kobayashi, K., Bindi, M., 2002. Responses of agricultural crops to free-air CO2 enrichment. Adv. Agron. 77, 293e368. http://handle.nal.usda.gov/10113/50534.

Landau, S., Mitchell, R.A.C., Barnett, V., Colls, J.J., Craigon, J., Payne, R.W., 2000.A parsimonious, multiple-regression model of wheat yield response to envi-ronment. Agric. For. Meteorol. 101 (2e3), 151e166. http://dx.doi.org/10.1016/S0168-1923(99)00166-5.

Laniak, G.F., Olchin, G., Goodall, J., Voinov, A., Hill, M., Glynn, P., Whelan, G.,Geller, G., Quinn, N., Blind, M., Peckham, S., Reaney, S., Gaber, N., Kennedy, R.,Hughes, A., 2013. Integrated environmental modeling: a vision and roadmap forthe future. Environ. Model. Softw. 39, 3e23.

Lebon, E., Dumas, V., Pieri, P., Schultz, H.R., 2003. Modelling the seasonal dynamicsof the soil water balance of vineyards. Funct. Plant Biol. 30, 699e710. http://dx.doi.org/10.1071/FP02222.

Lebon, E., Pellegrino, A., Louarn, G., Lecoeur, J., 2006. Branch development controlsleaf area dynamics in Grapevine (Vitis vinifera) growing in drying soil. Ann. Bot.98 (1), 175e185. http://dx.doi.org/10.1093/aob/mcl085.

Lebon, E., Pellegrino, A., Tardieu, F., Lecoeur, J., 2004. Shoot development ingrapevine (Vitis vinifera) is affected by the modular branching pattern of thestem and intra- and inter-shoot trophic competition. Ann. Bot. 93, 263e274.http://dx.doi.org/10.1093/aob/mch038.

Liu, H.F., Wu, B.H., Fan, P.G., Xu, H.Y., Li, S.H., 2007. Inheritance of sugars and acids inberries of grape (Vitis vinifera L.). Euphytica 153, 99e107. http://dx.doi.org/10.1007/s10681-006-9246-9.

Lobell, D.B., Field, C.B., Cahill, K.N., Bonfils, C., 2006. Impacts of future climatechange on California perennial crop yields: model projections with climate andcrop uncertainties. Agric. For. Meteorol. 141 (2e4), 208e218. http://dx.doi.org/10.1016/j.agrformet.2006.10.006.

Lobell, D.B., Field, C.B., 2007. Global scale climateecrop yield relationships and theimpacts of recent warming. Environ. Res. Lett. 2 (1), 14e21. http://dx.doi.org/10.1088/1748-9326/2/1/014002.

Louarn, G., Lecoeur, J., Lebon, E., 2008. A three-dimensional statistical reconstruc-tion model of Grapevine (Vitis vinifera) simulating canopy structure variabilitywithin and between Cultivar/Training system pairs. Ann. Bot. 101 (8),1167e1184. http://dx.doi.org/10.1093/aob/mcm170.

Loumou, A., Giourga, C., 2003. Olive groves: “The life and identity of the Mediter-ranean”. Agric. Hum. Values 20 (1), 87e95. http://dx.doi.org/10.1023/A%3A1022444005336.

Lovelli, S., Perniola, M., Di Tommaso, T., Ventrella, D., Moriondo, M., Amato, M.,2010. Effects of rising atmospheric CO2 on crop evapotranspiration in a Medi-terranean area. Agric. Water Manag. 97 (9), 1287e1292. http://dx.doi.org/10.1016/j.agwat.2010.03.005.

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e1514

Luo, Y., Jackson, R.B., Field, C.B., Mooney, H.A., 1996. Elevated CO2 increasesbelowground respiration in California grasslands. Oecologia 108 (1), 130e137.http://dx.doi.org/10.1007/BF00333224.

Maselli, F., Chiesi, M., Brilli, L., Moriondo, M., 2012. Simulation of olive fruit yield inTuscany through the integration of remote sensing and ground data. Ecol.Model. 244, 1e12. http://dx.doi.org/10.1016/j.ecolmodel.2012.06.028.

Matese, A., Crisci, A., Di Gennaro, S.F., Primicerio, J., Tomasi, D., Marcuzzo, P.,Guidoni, S., 2014. Spatial variability of meteorological conditions at differentscales in viticulture. Agric. For. Meteorol. 189e190, 159e167. http://dx.doi.org/10.1016/j.agrformet.2014.01.020.

Melgar, J.C., Syvertsen, J.P., García-S�anchez, F., 2008. Can elevated CO2 improve salttolerance in olive trees? J. Plant Physiol. 165 (6), 631e640. http://dx.doi.org/10.1016/j.jplph.2007.01.015.

Menzel, A., Sparks, T.H., Estrella, N., Koch, E., Aasa, A., Ahas, R., Alm-Kübler, K.,Bissolli, P., Braslavsk�a, O., Briede, A., Chmielewski, F.M., Crepinsek, Z., Curnel, Y.,Dahl, A., Defila, C., Donnelly, A., Filella, Y., Jatczak, K., Måge, F., Mestre, A.,Nordli, O., Pe~nuelas, J., Pirinen, P., Remi�sov�a, V., Scheifinger, H., Striz, M.,Susnik, A., Van Vliet, A.J.H., Wielgolaski, F.E., Zach, S., Zust, A., 2006. Europeanphenological response to climate change matches the warming pattern. Glob.Change Biol. 12, 1969e1976. http://dx.doi.org/10.1111/j.1365-2486.2006.01193.x.

Moriondo, M., Orlandini, S., De Nuntiss, P., Mandrioli, P., 2001. Effect of agro-meteorological parameters on the phenology of pollen emission and produc-tion of olive trees (Olea europea L.). Aerobiologia 17 (3), 225e232. http://dx.doi.org/10.1023/A:1011893411266.

Moriondo, M., Giannakopoulos, C., Bindi, M., 2011a. Climate change impactassessment: the role of climate extremes in crop yield simulation. Clim. Change104 (3e4), 679e701. http://dx.doi.org/10.1007/s10584-010-9871-0.

Moriondo, M., Bindi, M., Fagarazzi, C., Ferrise, R., Trombi, G., 2011b. Framework forhigh-resolution climate change impact assessment on grapevines at a regionalscale. Reg. Environ. Change 11 (3), 553e567. http://dx.doi.org/10.1007/s10113-010-0171-z.

Moriondo, M., Jones, G.V., Bois, B., Dibari, C., Ferrise, R., Trombi, G., Bindi, M., 2013a.Projected shifts of wine regions in response to climate change. Clim. Change 119(3e4), 825e839. http://dx.doi.org/10.1007/s10584-013-0739-y.

Moriondo, M., Trombi, G., Ferrise, R., Brandani, G., Dibari, C., Amman, C.M., Mariotti-Lippi, M., Bindi, M., 2013b. Olive trees as bio-indicators of climate evolution inthe Mediterranean Basin. Glob. Ecol. Biogeogr. 22, 818e833. http://dx.doi.org/10.1111/geb.12061.

Moutinho-Pereira, J., Correia, C., Falco, V., 2006. Effects of elevated CO2 on grape-vines grown under Mediterranean field conditionseimpact on grape and winecomposition. Aust. J. Grape Wine Res. 6, 2e12. http://www.oiv2007.hu/documents/viticulture/156_oiv2007_falco.pdf.

Mullins, M.G., Bouquet, A., Williams, L.E., 1992. Biology of the Grapevine. CambridgeUniversity Press. Available at: http://books.google.it/books?hl¼enandlr¼andid¼wnNvmRjfxgQCandoi¼fndandpg¼PR9anddq¼Biologyþofþtheþgrapevineþmullinsþ1992andots¼L_l02Fdtctandsig¼iXymrntBhkMM9xiT0sA88D1QfOU (accessed 29.10.13).

Nemani, R.R., White, M.A., Cayan, R.D., Jones, G.V., Running, S.W., Coughlan, J.C.,Peterson, D.L., 2001. Asymmetric warming over coastal California and its impacton the premium wine industry. Clim. Res. 19 (1), 25e34.

Nendel, C., Kersebaum, K., 2004. A simple model approach to simulate nitrogendynamics in vineyard soils. Ecol. Model. 177 (1e2), 1e15. http://dx.doi.org/10.1016/j.ecolmodel.2004.01.014.

Nieto, O.M., Castro, J., Fern�andez, E., Smith, P., 2010. Simulation of soil organiccarbon stocks in a Mediterranean olive grove under different soil-managementsystems using the RothC model. Soil Use Manage. 26 (2), 118e125. http://dx.doi.org/10.1111/j.1475-2743.2010.00265.x.

Norby, R.J., Hanson, P.J., ONeill, E.G., Tschaplinski, T.J., Weltzin, J.F., Hansen, R.A.,Cheng, W.X., Wullschleger, S.D., Gunderson, C.A., Edwards, N.T., Johnson, D.W.,2002. Net primary productivity of a CO2-enriched deciduous forest and theimplications for carbon storage. Ecol. Appl. 12 (5), 1261e1266. http://dx.doi.org/10.2307/3099969.

Orlandi, F., Avolio, E., Bonofiglio, T., Federico, S., Romano, B., Fornaciari, M., 2013.Potential shifts in olive flowering according to climate variations in southernItaly. Meteorol. Appl. 20, 497e503. http://dx.doi.org/10.1002/met.1318.

Orlandini, S., Dalla Marta, A., Mattii, G.B., 2008. Analysis and agrometeorologicalmodeling of grapevine responses to different trellising system. Vitis 47 (2),89e96.

Oyarzun, R.A., St€ockle, C.O., Whiting, M.D., 2007. A simple approach to modelingradiation interception by fruit-tree orchards. Agric. For. Meteorol. 142 (1),12e24. http://dx.doi.org/10.1016/j.agrformet.2006.10.004.

Pallas, B., Loi, C., Christophe, A., Courn�ede, P.H., Lecoeur, J., 2011. Comparison ofthree approaches to model grapevine organogenesis in conditions of fluctuatingtemperature, solar radiation and soil water content. Ann. Bot. 107 (5), 729e745.http://dx.doi.org/10.1093/aob/mcq173.

Palosuo, T., Kersebaum, K.C., Angulo, C., Hlavinka, P., Moriondo, M., Olesen, J.E.,Patil, R.H., Ruget, F., Rumbaur, C., Tak�a�c, J., Trnka, M., Bindi, M., Caldag, B.,Ewert, F., Ferrise, R., Mirshel, W., Saylan, L., Siska, B., R€otter, R., et al., 2011.Simulation of winter wheat yield and its variability in different climates ofEurope: a comparison of eight crop growth models. Eur. J. Agron. 35 (3),103e114. http://dx.doi.org/10.1016/j.eja.2011.05.001.

Poni, S., Magnanini, E., Bernizzoni, F., 2003. Degree of correlation between totallight interception and whole-canopy net CO2 exchange rate in two grapevinegrowth systems. Aust. J. Grape Wine Res. 9 (1), 2e11. http://dx.doi.org/10.1111/j.1755-0238.2003.tb00226.x.

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

Poni, S., Palliotti, A., Bernizzoni, F., 2006. Calibration and evaluation of a STELLASoftware-based daily CO2 balance model in Vitis vinifera L. J. Am. Soc. Hortic.Sci. 131 (2), 273e283.

Ponti, L., Gutierrez, A.P., Ruti, P.M., Dell’Aquila, A., 2014. Fine-scale ecologicaland economic assessment of climate change on olive in the MediterraneanBasin reveals winners and losers. Proc. Natl. Acad. Sci. U.S.A. 111 (15),5598e5603.

Quiroga, S., Iglesias, A., 2009. A comparison of the climate risks of cereal, citrus,grapevine and olive production in Spain. Agric. Syst. 101 (1), 91e100. http://dx.doi.org/10.1016/j.agsy.2009.03.006.

Rodríguez-Lovelle, B., Gaudill�ere, J.P., 2002. Carbon and nitrogen partitioning ineither fruiting or non-fruiting grapevines: effects of nitrogen limitation beforeand after veraison. Aust. J. Grape Wine Res. 8 (2), 86e94. http://dx.doi.org/10.1111/j.1755-0238.2002.tb00216.x.

R€otter, R.P., Palosuo, T., Kersebaum, K.C., Angulo, C., Bindi, M., Ewert, F., Ferrise, R.,Hlavinka, P., Moriondo, M., Nendel, C., Olesen, J.E., Patil, R.H., Ruget, F., Tak�a�c, J.,Trnka, M., 2012. Simulation of spring barley yield in different climatic zones ofNorthern and Central Europe: a comparison of nine crop models. Field Crop Res.133, 23e36. http://dx.doi.org/10.1016/j.fcr.2012.03.016.

Sadras, V.O., Soar, C.J., 2009. Shiraz vines maintain yields in response to a 2e4�Cincrease in maximum temperature using an open-top heating system at keyphenostages. Eur. J. Agron. 31, 250e258.

Sage, R.F., Kubien, D.S., 2007. The temperature response of C3 and C4 photosyn-thesis. Plant Cell Environ. 30, 1086e1106. http://dx.doi.org/10.1111/j.1365-3040.2007.01682.x.

Salazar-Parra, C., Aguirreolea, J., S�anchez-Díaz, M., Irigoyen, J.J., Morales, F., 2012a.Climate change (elevated CO2, elevated temperature and moderate drought)triggers the antioxidant enzymes' response of grapevine cv. Tempranillo,avoiding oxidative damage. Physiol. Plant. 144 (2), 99e110. http://dx.doi.org/10.1111/j.1399-3054.2011.01524.x (Epub 10.11.11.).

Salazar-Parra, C., Aguirreolea, J., S�anchez-Díaz, M., Irigoyen, J.J., Morales, F., 2012b.Photosynthetic response of Tempranillo grapevine to climate change scenarios.Ann. Appl. Biol. 161 (3), 277e292. http://dx.doi.org/10.1111/j.1744-7348.2012.00572.x.

Santos, J.A., Malheiro, A.C., Karremann, M.K., Pinto, J.G., 2011. Statistical modeling ofgrapevine yield in the Port Wine region under present and future climateconditions. Int. J. Biometeorol. 55 (2), 119e131. http://dx.doi.org/10.1007/s00484-010-0318-0. Epub 2010 May 12.

Schlenker, W., Roberts, M.J., 2009. Nonlinear temperature effects indicate severedamages to US crop yields under climate change. Proc. Natl. Acad. Sci. U.S.A. 106(37), 15594e15598. http://dx.doi.org/10.1073/pnas.0906865106.

Schultz, H.R., Stoll, M., 2010. Some critical issues in environmental physiology ofgrapevines: future challenges and current limitations. Aust. J. Grape Wine Res.16 (s1), 4e24. http://dx.doi.org/10.1111/j.1755-0238.2009.00074.x.

Schultz, H.R., 2000. Climate change and viticulture: a European perspective onclimatology, carbon dioxide and UV-B effects. Aust. J. Grape Wine Res. 6 (1),2e12.

Shuttleworth, J.W., Wallace, J.S., 1985. Evaporation of sparse crops-an energycombination theory. Quart. J. R. Meteor. Soc. 3, 839e885.

Sinclair, T.R., Muchow, R.C., 1999. Radiation use efficiency. Adv. Agron. 65, 215e265.Sinclair, T.R., Seligman, N.A.G., 1996. Crop modeling: from infancy to maturity.

Agron. J. 88 (5), 698e704.Sofo, A., Manfreda, S., Dichio, B., Fiorentino, M., Xiloyannis, C., 2008. The olive tree: a

paradigm for drought tolerance in Mediterranean climates. Hydrol. Earth Syst.Sci. 12 (1), 293e301. http://dx.doi.org/10.5194/hessd-4-2811-2007.

St€ockle, C.O., Donatelli, M., Nelson, R., 2003. Cropsyst, a cropping systems simula-tion model. Eur. J. Agron. 18 (3), 289e307. http://dx.doi.org/10.1016/S1161-0301(02)00109-0.

Tanasijevic, L., Todorovic, M., Pereira, L.S., Pizzigalli, C., Lionello, P., 2014. Impacts ofclimate change on olive crop evapotranspiration and irrigation requirements inthe Mediterranean region. Agric. Water Manag. 144, 54e68.

Tognetti, R., Sebastiani, L., Vitagliano, C., Raschi, A., Minnocci, A., 2001. Responses oftwo olive tree (Olea europaea l.) cultivars to elevated CO2 concentration in thefield. Photosynthetica 39 (3), 403e410.

Tubiello, F.N., Ewert, F., 2002. Simulating the effects of elevated CO2 on crops:approaches and applications for climate change. Eur. J. Agron. 18 (1e2), 57e74.http://dx.doi.org/10.1016/S1161-0301(02)00097-7.

Tubiello, F.N., Donatelli, M., Rosenzweig, C., Stockle, C.O., 2000. Effects of climatechange and elevated CO2 on cropping systems: model predictions at two Italianlocations. Eur. J. Agron. 13 (2), 179e189. http://dx.doi.org/10.1016/S1161-0301(00)00073-3.

Vald�es-G�omez, H., Celette, F., García De Cort�a-Zar-Atauri, I., Jara-Rojas, F., Gary, C.,2009. Modelling soil water content and grapevine growth and developmentwith the STICS crop soil model under two different water management stra-tegies. J. Int. Sci. de la Vigne du Vin 43 (1), 13e28.

Veroustraete, F., Sabbe, H., Rasse, D.P., Bertels, L., 2004. Carbon mass fluxes of forestsin Belgium determined with low resolution optical sensors. Int. J. Remote Sens.25 (4), 769e792. http://dx.doi.org/10.1080/0143116031000115238.

Villalobos, F.J., Testi, L., hidalgo, J., Pastor, M., Orgaz, F., 2006. Modeling potentialgrowth and yield of olive (Olea europaea L.) canopies. Eur. J. Agron. 24 (4),296e303. http://dx.doi.org/10.1016/j.eja.2005.10.008.

Villalobos, F.J., Testi, L., Orgaz, F., García-Tejera, O., Lopez-Bernal, A., Gonz�alez-Dugo, M.V., Ballester-Lurbe, C., Castel, J.R., Alarc�on-Caba~nero, J.J., Nicol�as-Nicol�as, E., Girona, J., Marsal, J., Fereres, E., 2013. Modeling canopy conductanceand transpiration of fruit trees in Mediterranean areas: a simplified approach.

es and grapevines in a changing climate, Environmental Modelling &

M. Moriondo et al. / Environmental Modelling & Software xxx (2015) 1e15 15

Agric. For. Meteorol. 171e172, 93e103. http://dx.doi.org/10.1016/j.agrformet.2012.11.010.

Viola, F., Noto, L.V., Cannarozzo, M., Lal Loggia, G., Porporato, A., 2012. Olive yield asa function of soil moisture dynamics. Ecohydrology 5 (1), 99e107. http://dx.doi.org/10.1002/eco.208.

Vossen, P., 2007. Olive oil: history, production, and characteristics of the world'sclassic oils. HortScience 42 (5), 1093e1100.

Webb, L.B., Whetton, P.H., Barlow, E.W.R., 2007. Modelled impact of future climatechange on the phenology of winegrapes in Australia. Aust. J. Grape Wine Res. 13(3), 165e175. http://dx.doi.org/10.1111/j.1755-0238.2007.tb00247.x.

Wermelinger, B., Baumgartner, J., 1990. Application of a demographic crop growthmodel: an explorative study on the influence of nitrogen on grapevine per-formance. Acta Hort. 276, 113e121.

Wermelinger, B., Baumgartner, J., Gutierrez, A., 1991. A demographic model ofassimilation and allocation of carbon and nitrogen in grapevines. Ecol. Model.53, 1e26. http://dx.doi.org/10.1016/0304-3800(91)90138-q.

White, J.W., Hoogenboom, G., Kimball, B.A., Wall, G.W., 2011. Methodologies forsimulating impacts of climate change on crop production. Field Crops Res. 124,357e368.

Please cite this article in press as: Moriondo, M., et al., Modelling olive treSoftware (2015), http://dx.doi.org/10.1016/j.envsoft.2014.12.016

Williams, L.E., Ayars, J.E., 2005. Grapevine water use and the crop coefficient arelinear functions of the shaded area measured beneath the canopy. Agric. For.Meteorol. 132 (3e4), 201e211. http://dx.doi.org/10.1016/j.agrformet.2005.07.010.

Williams, D.W., Williams, L.E., Barnett, W.W., Kelley, K.M., McKenry, M.V., 1985.Validation of a model for the growth and development of the ThompsonSeedless grapevine. I. Vegetative growth and fruit yield. Am. J. Enol. Vitic. 36,275e282.

Zerihun, A., Treeby, M.T., 2002. Biomass distribution and nitrate assimilation inresponse to N supply for Vitis vinifera L. cv. Cabernet Sauvignon on five Vitisrootstock genotypes. Aust. J. Grape Wine Res. 8 (3), 157e162. http://dx.doi.org/10.1111/j.1755-0238.2002.tb00251.x.

Zhang, X., Walker, R.R., Godwin, D., White, B., 2002. Using VineLOGIC to predictgrapevine phenology, yield and salinity impacts in irrigated vineyards: a casestudy. In: Walker, R. (Ed.), VineLOGIC Education Package. CRCV and CSIRO PlantIndustry, Merbein, Australia.

Zhu, X., Moriondo, M., van Ierland, E.C., Trombi, G., Bindi, M., 2014. A model-basedassessment of adaptation options for Chianti wine production in Tuscany (Italy)under climate change. Reg. Environ. Change 1e12. http://dx.doi.org/10.1007/s10113-014-0622-z (in press).

es and grapevines in a changing climate, Environmental Modelling &