A mix-method model for adaptation to climate change in the
agricultural sector: A case study for Italian wine farms
Sandro Sacchelli
*, Sara Fabbrizzi, Marco Bertocci, Enrico Marone, Silvio Menghini,
Iacopo Bernetti
Department of Agricultural, Food and Forest Systems Management, University of Florence, P.le delle Cascine 18, I-50144, Florence, Italy
a r t i c l e i n f o
Article history:Received 4 January 2017 Received in revised form 31 July 2017
Accepted 12 August 2017 Available online 14 August 2017 Handling Editor: Yutao Wang Keywords:
Climate change Adaptation strategy Complex system Metaheuristic model Decision Support System Wine farm accounting
a b s t r a c t
The negative effects of climate change are predicted to impact the agricultural sector in coming decades. Economic losses and modifications of production processes are fundamental issues to consider in coping with the harmful consequences of climate variability. The literature and empirical evidence show that the wine sector is extremely vulnerable to this risk. These studies show that this sector lacks appropriate adaptation strategies due to the complexity of the analysed systems and interrelations between a number of socio-economic and environmental variables. The present study designed a decision support system to identify adaptation strategies for wine farms undergoing climate change. The tool allows for the analysis of a wine farm's economic performance when it adopts measures to cope with climatic variability. Average values for climate change and extreme events were considered to assess different scenarios. A mix-method approach was applied to integrate probability calculations, complex system analyses and operational research (a metaheuristic approach). The model was tested on a case study located in central Italy (Chianti Classico). To maintain and improve futurefinancial performance, organic farming and adjustments to procedural guidelines were recommended as key strategies. Economic variables, such as the average price of wine, seem to have a strong influence on farms' implementation of adaptive measures. An additional result seems to suggest that insurance schemes in areas producing high quality wine are only suitable when low-level deductibles and public funding are available. The present work shows that the decision support system favours analytical sensitivity to different scenarios and variables related to climate change as well as to socio-economic shifts in the viticulture sector.
© 2017 Elsevier Ltd. All rights reserved.
1. Introduction
Climatic change is forecast as a major phenomenon in short-and long-term projections of Intergovernmental Panel on Climate Change (IPCC, 2014). The predicted variability is mainly associated with an increase in average temperature and decrease in average precipitation. These aspects are also combined with a high likeli-hood for the intensification of extreme events such as drought, particularly in Europe (Collins et al., 2013). This framework stresses the need for appropriate actions to cope with the risks related to climate change, taking into account the scale of analysis and local features (Mirzabaev et al., 2015).
Rural areas appear particularly susceptible to climate change due to harmful impacts on important economic and social
dynamics (Dubey et al., 2016) and environmental and land-use trends (Dasgupta et al., 2014). As stated inIPCC (2014),“the term impacts is used primarily to refer to the effects on natural and human systems of extreme weather and climate events and of climate change. Impacts generally refer to effects on lives, liveli-hoods, health, ecosystems, economies, societies, cultures, services and infrastructure […]” (pag. 124). The wine industry, a rural sub-sector, is also at risk for substantial climatic-related threats (Schultz and Jones, 2010). The primary negative effects of climate change on the wine industry are potential losses to product quantity and quality. Additional risks are related to consequences on revenues and production costs throughout the supply chain (Mozell and Thach, 2014). The modification of production processes due to climate variability and extreme events may lead to addi-tional socio-economic impacts for the whole sector and its related activities (Pomarici and Seccia, 2016).
Researchers have made several attempts to identify adaptation
* Corresponding author.
E-mail address:sandro.sacchelli@unifi.it(S. Sacchelli).
Contents lists available atScienceDirect
Journal of Cleaner Production
j o u r n a l h o me p a g e :w w w .e l se v i e r. co m/ lo ca t e / jc le p r o
http://dx.doi.org/10.1016/j.jclepro.2017.08.095
strategies and policies to address the harmful consequences of climate change on the wine industry. Most studies focus on thefirst step of the production process (field phase) showing a predominant interest in water deficit contrast (Sacchelli et al., 2016).Costa et al. (2014)presented a strategy to promote water efficiency and sus-tainable water use and to minimise environmental impact in southern Europe's wine sector.Nicholas and Durham (2012) con-ducted interviews to observe farm-scale adaptive responses to climatic stress and to comprehend the motivations and views of agricultural managers in California. A further study focused on adaptive capacities for aesthetic logic e aligned with environ-mental sustainability and mitigatione and market logic to promote adaptation to localised impacts in Australia (Fleming et al., 2015).
Lereboullet et al. (2013) considered suitable socio-ecological adaptation measures for two case studies (Roussillon e France and McLaren Valee Australia) and evaluated the economic per-formance of adaptive actions from different perspectives.Hadarits (2011)introduced the concepts of adaptive strategies and adaptive capacity and their influence on economic performance of wine farms in Maule region (Chile).De Salvo et al. (2015)paid particular attention to winegrower education, farm technology and wine-growers' awareness of climate change's impacts to cope with these effects in the Moldavia region's (Romania) wine industry.Bernetti et al. (2012) assessed the “adaptation probability” according to professional training, specialisation and capacities to react to un-expected situations among farmers in the area of Brunello di Montalcino (Italy). A recent study measured thefinancial efficiency of vineyard relocation and the adoption of drought-resistant grape varieties in the Chianti region (Italy) (Zhu et al., 2016).
The above literature review suggests that the high territorial peculiarity and differentiation of local wine-growing contexts e
well explained by the terroir concept (Clingeleffer, 2014)e hinder the development of generalisable models. These problems are manifest in researchers' projection of suitable adaptation strategies and depictions of farmers' intentions to adapt to climate change in the medium- to long-term (Arunrat et al., 2017).Snyder et al. (2011)
stated that the assessment of climate change impacts is often “disciplinary-based and not sufficiently integrative across impor-tant disciplinary subcomponents, producing misleading results that have potentially dangerous environmental consequences” (pag. 467). A further issue is the need to introduce the complexity of the analysed systems to facilitate readiness, communication and potential applications for research outputs (Serrao-Neumann et al., 2015). This last aspect is particularly important in the case of me-dium- to long-term scenarios and the involvement of non-expert stakeholders who may perceive the analysed issues as vague and ambiguous (Dulic et al., 2016).
The combination of methods and tools in climate change impact analysis creates a need for procedures that can represent the“wine system” as a structure with non-continuous, reflexive and emer-gent characteristics. These parameters can be addressed with the concept of “complex systems” (Sayama, 2015). As reported in
Chapman (2009), there is no agreement upon definition of complexity, but the presence of non-continuous interactions (i.e. when a change in one variable does not cause a proportionate constant change in a dependent variable) could be considered as the most important characteristic. Chapman states how“the single parameters of a system have a cognitive model of their role and position in the system, but because these cognitive models cannot claim to have complete knowledge of themselves or the system, the partiality of knowledge is the reason that the characteristic behaviour of the whole is seen to be emergent” (Chapman, 2009; Nomenclature
ASi dummy variable referred to i-th adaptation strategy,
that can be 1 if activated, 0 otherwise. ASican be: ASn
anti-hail net, ASoorganic farming, ASgorganic
farmingþ certification, ASffans, AShheater/candles,
ASuunder/over canopy irrigation, ASccultivar
substitution, ASrfixed irrigation plant, ASeemergency
irrigation, ASdprocedural guidelines modification, ASt
increase of phytosanitary treatments, ASsinsurance
Ci cost per each4 category (V/ha y1)
de level of damage for e-th extreme events (%) where e
can be: F frost, N hail, H heat waves, D drought, T phytopathologies
DEN dummy variable: 0 if Controlled and Guaranteed Designation of Origin (DOCG) certification is not present, 1 otherwise
e extreme event category where e can be: F frost, N hail, H heat waves, D drought, T phytopathologies
I intercept of VQfequation
LAN dummy variable: 0 if landscape constraint is not present, 1 otherwise
m number of extreme event categories
n number of cost categories (related to wine production process)
NRc current net revenues (V/ha y1)
NRf future net revenues (V/ha y1)
ORG dummy variable: 0 if certification for organic practices is not present, 1 otherwise
P(e)c current probability of e-th extreme event occurrence
defined as inTable 1(%)
P(e)f future probability of e-th extreme event occurrence
defined as inTable 1(%) pA base price for bulk wine (V/l)
pB base price for bottled wine (V/l)
pa final price of bulk wine (V/l)
pb final price of bottled wine (V/l)
Tsum average temperature in the summer period (C)
VQc current vintage quality rating of wine
VQf future vintage quality rating of wine
Yc current wine production calculated with the growth
simulator model ofBindi et al. (1997)and updated in
Bindi et al. (2005)(l/ha y1)
Yf future wine production calculated with the growth
simulator model ofBindi et al. (1997)and updated in
Bindi et al. (2005)(l/ha y1)
Zmat precipitation in the anthesis-maturity period (mm)
a
percentage of bulk wine (%)g
andd
coefficients of VQfequationεVQ price elasticity in respect to wine quality. Derived from
average value inNeill (2011)andAbraben (2014)
q
indemnity (V/ha y1)l
influence of certification of organic practices on wine price (%). Derived fromAbraben (2014)4 cost category (see section2.2. for more details)
c
deductible (%) for insurance schemesu
influence of DOCG certification on wine price (%). Derived fromContini et al. (2015)pag. 31). Samet illustrated the importance of complexity science from an innovative models viewpoint (Samet, 2011) and the need to apply complexity-based theory in prospective studies (Samet, 2012).Derbyshire (2016)showed a number of benefits and limits derived from complexity-based studies and discussed the potential implications of their use.
A few examples of complexity-based methodologies have been applied in the literature to assess the impact of climate change on the agricultural and food sector. A review by Beermann (2011)
stressed the importance of system thinking to cope with climate change effects on German food firms. Nikolaou et al. (2015)
implemented a model based on system thinking and system dy-namic approaches to investigate the relationships between climate change risks,financial performance and the operational processes offirms for ski resorts and agribusiness in Greece.
According to the authors' knowledge, system thinking and complexity-based approaches have never been used to analyse adaptation measures to climate change in the wine sector.
The aim of the current study is to develop a Decision Support System (DSS) that can guide policymakers, local stakeholders and wine farmers in designing adaptation strategies to cope with the impacts of climate change on viticulture. “Impacts” designate economic damages with and without farm-level or normative in-terventions to cope with climate change. The case study relied on farm accounting databases, territorial information and climatic parameter projections, and it targeted the Controlled and Guaran-teed Designation of Origine DOCG e area of the Chianti Classico district in Tuscany (central Italy). The DOCG label represents a certification of origin and guarantee of high quality in Italian wines. An assumption considered in the study is that farmers will not switch to other farm types. Ricardian studies showed that certain farming activities move north in the case of climate change (Fraga et al., 2013). As a consequence, qualitative wine production in northern regions will have an impact on existing wine regions that implement (expensive) adaptation measures (De Salvo et al., 2015). From a methodological point of view, spatial statistical opera-tions were applied to define the forecast climatic data and extreme events for the test area. A series of spatial overlays were performed to apply the average monthly climatic data to the analysed context. The available growth models and methods to estimate wine quality were used to measure the impact of climate change on a specific Vitis vinifera L. variety (Sangiovese cultivar) (Bindi et al., 1997; updated inBindi et al., 2005 to take into account water-deficit stress). A literature review and cognitive mapping methods facilitated interviews with local experts in the wine sector, which recorded their subjective perceptions of appropriate adap-tation strategies. A metaheuristic model e solved by an evolu-tionary genetic algorithm e was applied to define intervention strategies that would minimise expected economic losses. The study's novel contribution among recent similar studies lies in its application of a mix-method model. The proposed DSS can analyse complex systems through a combination of different techniques, as explained in next section.
2. Material and methods
2.1. General framework: the mix-method approach
The terms“mix-method” refers to the so-called “nested” design, where different strategies are applied in a single study (DePoy and Gitlin, 2016). In a mix-method approach, the research design strategy can be depicted in the concurrent nested approach (Creswell, 2013). Concurrent nested is characterized by techniques that give priority to one of the applied methods, while other are embedded or“nested” in order to seek information from different
levels. In this study, concurrent nested approach was used and the dominant method is operational research (i.e., metaheuristic opti-misation) for the optimisation of the system; it was applied with input data derived from the output of four different models: i) economic model (i.e., literature data from Full Cost Accounting application); ii) agronomic models (i.e., vine growth simulator); iii) calculated likelihood of damage to wine production and iv) system thinking approach (i.e., cognitive mapping). In this case optimisa-tion cannot be performed simultaneously but sequentially (the output of different models are used as input of another model in a subsequent phase) because the system is too complex and involve different techniques such as spatial analysis and goal achievement. A brief description of these techniques is furnished as follow; additional considerations on the models and their integration are described in the following sections.
i) Economic model: the Full Cost Accounting method evaluates the cost of a wine product unit (bottle, litre, etc.), taking into account the remuneration of all factors in its production (i.e., including both explicit and implicit costs). Full Cost Ac-counting data were obtained from the work ofMarone et al. (2014).
ii) A grapevine growth simulator was developed byBindi et al. (1997)and updated inBindi et al. (2005)for Vitis vinifera L. (Sangiovese cultivar). It uses a unique model of climatic data to forecast annual production per hectare. This model was applied to compute current and future production of wine in Chianti Classico district by the use of climatic data from
Hijmans et al. (2005)(mean temperature and precipitation). iii) Further potential damages to wine production are expected losses quantifiable in terms of wine quantity reduction due to extreme events and wine quality decrease. Wine quantity reduction is computed basing on literature data (probability of extreme events). Forecasted climatic data (mean temper-ature and precipitation) were applied to calculate future quality of wine (see section2.3for more details).
iv) Cognitive mapping is a technique to structure and evaluate a specific topic represented in verbal (e.g., by means of in-terviews) or documentary form (Ackermann et al., 2004). Cognitive Maps (CM) are the results of cognitive mapping. CM are graphical, semi-quantitative and dynamic systems composed of concepts (or nodes), and their relations are represented by arrows (Kosko, 1986). The value of nodes and the link between two concepts are weighted to facilitate their interpretation (Papageorgiou et al., 2011). CM was extensively applied in thefield of climate change perception and scenario analysis (Henly-Shepard et al., 2015). The pre-sent study applied CM to interview experts about the po-tential implementation of adaptation strategies to cope with climate change and about costs variation of production process due to adaptation strategies.
Once the general framework of the problem was assessed the outputs of the above techniques were integrated in a model (the DSS) to perform the optimisation of the system. Since the imple-mented model is characterized by non-continuous functionse e.g., for the presence of IF-THEN operatore a metaheuristic optimisa-tion designed to reach specific objective functions was carried out. The simplified structure of the problem is shown inFig. 1. Solid lines describe current scenario (section2.2); dotted lines highlight climate change impact (section2.3). Potential adaptation to climate change (section2.4) are expressed by dashed lines. The results will refer to a farm representing the average characteristics of the Chianti Classico district. The expression“average characteristics” denotes that all input data of the model (wine prices, unit costs,
number/typology/time of workers) derive from the average values of organisation and production costs related to a sample of wine farms as reported in section2.2(Marone et al., 2014).
Climate change has a direct influence on two portions of the chain: wine quantity and wine quality. Quality refers to product characteristics based on procedural guidelines and consumer ex-pectations. The procedural guideline of Chianti Classico is a docu-ment that gives guidelines for vine cultivation and wine production to obtain DOCG certification. The certification is verified not only by monitoring production activities but also by analysing and tasting thefinal product. For example, procedural guidelines indicate the area of production, maximum yield per grapevine, vine varieties allowed, and additional constraints (e.g., the exclusion of irrigation e apart from emergency irrigation e to maintain DOCG certification).
Wine quantity and quality, coupled with the organisation of wine production processes, have an effect on both the revenues and costs of wine selling and production. The difference between rev-enues and cost yields net revrev-enues. To cope with current extreme events and with future negative climate change effects, a series of adaptation strategies can be adopted. These measures target pro-duction processes. The present work recommends adaptation schemes among approaches to maximise farmers' economic ben-efits (i.e., net revenues).
2.2. Current scenario
Fig. 1shows the current organisation of the wine industry for the case study (solid lines). The organisation and costs related to production derive from literature data. Data from the work of
Marone et al. (2014) were applied. This study collected data through 40 face-to-face interviews in a stratified sample e based on dimension (number of workers) and surface of vineyards e of wineries representative of the Chianti Classico area. The data were processed using a Full Cost Accounting model (Marone et al., 2014); costs were computed and classified according to production phases and production factors (Marone et al., 2014). The following cost categories were defined (expressed in V/ha y1): management
(family members); administration (family members); labour (family members); employees (management and administration); employees (labour); administrative expenses; energy, water and bioproducts; insurance; consultations; financial; maintenance; phytosanitary products; fuels; winemaking products (e.g., yeast);
third party activities; sellers; amortizations; rents; packaging and certification. The partition of bulk and bottled DOCG was also derived from farm data (face-to-face interviews) taking into ac-count the average value of the last three years (Marone et al., 2014). Based on the same period, the current vintage quality rating of Chianti Classico wine (VQc) was derived from the 2016 Vintage
Chart (Wine Enthusiast Magazine, 2016). VQ is a score in the range 0e100 that worldwide experts attribute to wines. Revenues from sales were derived for the product based on the quantity and weighted average prices for bulk and bottled wine. The data for bulk and bottled wine prices for the Chianti region are depicted in the weekly bulletins of theChamber of Commerce, Industry, Crafts and Agriculture of Florence (2016)and theChamber of Commerce, Industry, Crafts and Agriculture of Siena (2016).
Wine production was calculated with the growth simulator model of Bindi et al. (1997)and updated inBindi et al. (2005). Current production was computed running the above model with WorldClim 1.4 (current conditions) spatial-based climatic data of mean temperature and precipitation (Hijmans et al., 2005). Cli-matic data, needed as inputs for the model, derived from zonal statistics operations based on the average value of WorldClim 1.4 spatial data in the study area.
The expected current yield reduction (amount of wine, in vol-ume)e E(YRc)e due to extreme events was analysed with a
sta-tistical computation that evaluated the current probability of event occurrence and potential level of damage. The analysis was based on the detailed literature review reported inTable 1. Yield reduc-tion was examined for different extreme events: hail, drought, heat waves, frost and phytopathologies.
The above consideration, i.e. current net revenues computation as well as probability calculation, can be modelled as in Eqs.(1) and (2). NRc¼ 8 < : h Yc,
a
,paþ Yc,ð1a
Þ,pb i Xn 4¼1 Ci 9 = ;,ð1 EðYRcÞÞ (1) where: EðYRcÞ ¼ Xm e¼1 PðeÞc$de (2)A default value of dewas set (15%) due to a high variability in
potential damage and a sensitivity analysis of d was carried out; it is discussed in the Results section. The level of damage was con-strained in the range 0e100%. The likelihood of extreme events was considered an ‘independent variable’ probability. Therefore, the probabilities of yield reduction due to extreme events were added to each other.
Thefinal prices of bulk and bottled wine depend on the pres-ence of DOCG denomination (Contini et al., 2015), organic certi fi-cation (Abraben, 2014) and the price elasticity in respect of wine quality (average value from Neill, 2011; Abraben, 2014). When DOCG was not present, the hypothesis was a shift to a lower cate-gory of denomination (Typical Geographical Indication - IGT) because it is a current practice in the study area. The price can be calculated as: pa¼ pАþ ðpА,DEN,
u
Þ þ ðpА,ORG,l
Þ þpА,D
VQ,εVQ (3) pb¼ pВþ ðpВ,DEN,u
Þ þ ðpВ,ORG,l
Þ þpB,D
VQ,εVQ (4)with
D
VQ variation of wine quality from current (VQc) to future(VQf) scenario.
The input data for Eqs.(1)e(4) are reported in the Supplemen-tary material (Table S1).
2.3. Climate change impact
The projected climatic data refer to the Representative Con-centration Pathway RCP4.5 for the year 2050 (medium-term sce-nario), which was developed by the IPCC (Moss et al., 2010). RCP4.5 outlines a stabilisation scenario in which total radiative forcing is stabilised before 2100 through the use of technologies and strate-gies to reduce greenhouse gas emissions (Clarke et al., 2007). Spatial RCP4.5 data were selected because they represent an
intermediate emission framework for the European context with respect to the other available scenario, at 2050. That timeline was chosen because it offers a suitable time range for the imple-mentation of structural adaptation strategies in the examined sector (e.g., change of vine cultivar, facilities installation). The impact on wine productivity and quality related to climatic change is due to two effects: i) average climatic parameters variation and ii) extreme events.
The amount of produced wine in the medium-term was also computed in this case by means of Bindi et al.'s (1997)model, updated in Bindi et al. (2005)through the method described in section2.2with the application of RCP4.5 climatic spatial data. For the analysis of future extreme events probability E(YRf) seeTable 1.
The expected net revenues for a future scenario were evaluated as: NRf ¼ 8 < : h Yf,
a
,paþ Yf,ð1a
Þ,pb i X n 4¼1 Ci 9 = ;, 1 EYRf (5) where: EYRf¼Xm e¼1 PðeÞf$de (6)Climate change can influence a wine's future vintage quality rating (VQf) (Jones et al., 2005). Some models that compute VQ for
the Tuscany region exist (Moriondo et al., 2011). To design a sta-tistical model for the Chianti Classico area, the present study modified a proposed model byMoriondo et al. (2011)as follows:
VQf ¼ I þ
g
,Zmatþd
,Tsum (7)Eq. (7) was calculated with 2016 Vintage Chart data (Wine Table 1
Definition of extreme events probability.
Extreme eventsa Current probability Future probability
Reference(s) Note Reference(s) Note
Hail (Baldi et al., 2014) Based on data of Florence and Siena province (Zollo et al., 2016) Based on variation of Simple Daily Intensity Index (SDII) for Tuscany
Drought (JRC, 2014) Number of consecutive days without rainfall (P) (precipitation<1 mm) was computed. Then, a threshold of“drought” (D) was defined by experts for examined area (number of days without rainfall). The probability of drought per year (pDy) was calculated as pDy¼ P/D. The
average of pDyfor the period 1981e2010
represents current likelihood (pD).
(Zollo et al., 2016) Based on variation of maximum number of consecutive dry days (CDD) for Tuscany
Heat waves (Fischer and Sch€ar, 2010) Full methodology described in (Fischer and Sch€ar, 2010)
(Fischer and Sch€ar, 2010) Full methodology described in (Fischer and Sch€ar, 2010) Frost (JRC, 2014) Presence of days with minimum temperature
(DTmin) less than threshold defined by experts
was computed for the period MarcheApril for every year (1: presence at less of one day; 0: no presence). The probability of frost (pF) was
calculated as pF¼
P
i
DTmin
i where i is the number
of years in the period 1981e2010.
(IPCC, 2012) A decrease in cold extreme is forecasted with very likely - 90 e100% - probability
Phytopathology (Tuscany Plant protection service, 2016)
Value computed on frequency and intensity for Chianti area of Florence and Siena provinces. Considered pathogens: Lobesia botrana, Botrytis cinerea, Plasmopara viticola, Uncinula necator, Daktulosphaira vitifoliae, Metcalfa pruinosa, vine acarus, black-rot
(Jones et al., 2005; Stock et al., 2005; Caffarra et al., 2012)
Results from literature review show how potential spread of pests and plant diseases will be rather variable and region-specific. A few augmented level of damage will be hypothesized (5%)
Enthusiast Magazine, 2016) and Joint Research Centre (JRC, 2014) historical climatic data. Inputs for the model (Zmat and Tsum)
derived from zonal statistics operation based on the spatial data of the RCP4.5 climatic scenario at 2050 in the study area (Hijmans et al., 2005). For more details, see Supplementary material (Table S1).
2.4. Adaptation strategies for climate change and their influence on the production process
A literature review and cognitive mapping techniques involving local experts and researchers were used to define appropriate adaptation strategies and their effects on damage, production process, denomination, certification and e also indirectly e on revenues and costs. CM was elaborated in a focus group of nine people including farmers, delegates of local consortia and re-searchers. This number is considered adequate for focus group as specified inMorgan (1997). CM was implemented with the pro-cedure suggested inAckermann et al. (2004). The result of the focus group was CM that transform the implementation of adaptation strategies in variation of costs (percentage variation, with respect to current significance). These values were confirmed and/or inte-grated through a literature analysis and consultation of specialised magazines. A tabular version of the obtained CM related to cost variation is reported inTable S2.
The adaptation strategies and their effects on average yield or expected damages, the restoration of economic losses and the loss/ maintenance of certification as well as denomination are sum-marised inTable 2and modelled in the algorithm 8.
IF ASf ¼ 1∨ASh¼ 1∨ASu¼ 1 THEN dF¼ 0 IF ASn¼ 1 THEN dN¼ 0 IF ASo¼ 1 THEN ðdH¼ 0Þ ∧ðdD*¼ dD,0:4Þ IF ASg¼ 1 THEN ðdH¼ 0Þ ∧ðdD*¼ dD,0:4Þ∧ðORG ¼ 1Þ IF ASe¼ 1 THEN ðdD¼ 0Þ IF ASt ¼ 1 THEN ðdT¼ 0Þ IF ASc¼ 1 THEN Yf ¼ Yc ∧ðDEN ¼ 0Þ IF ASr¼ 1 THEN Yf ¼ Yc ∧ðDEN ¼ 0Þ ∧ðdD¼ 0Þ IF ASd¼ 1 THEN ðDEN ¼ 1Þ (8)
Insurance schemes are an indirect form of adaptation strategies.
Current insurance characteristics derive from national legislation, which represents an expansion of combined and comprehensive insurance schemes (Ministero delle Politiche Agricole Alimentari e Forestali - MIPAAF, 2016). For more details about national agricul-tural insurance systems, please seePerez-Blanco et al. (2014). To ensure aflexible evaluation of insurance performance, the model includes the following hypotheses and parameters: i) a premium is defined as the annual fee due from a farmer to a multi-risk insur-ance to cover all damages from extreme events; ii) a deductible represents the threshold of damage (computed on yield) that has to be crossed to pay indemnity (i.e., a deductible is the initial share of the damage that is not protected) (Perez-Blanco et al., 2014) and iii) an incentive is the percentage of a premium covered by public funding. Indemnity (payment to farmer in case of damage) is calculatede for both current and predicted extreme events e based on lost net revenues with respect to current production. Written as a formula,
IF ðASs¼ 1Þ∧ðEðYRÞ >
c
Þ THEN ½q
¼ NR,ðEðYRÞc
Þ(9)
2.5. Optimisation of the system
In the present work,“optimisation” of the system denotes a farm's optimal economic performance. The maximisation of net revenues (NR) was set as objective function. Formally, each scenario reported in Results section can be represented as follows:
max NR subject to: ASi¼ binary
(10)
where ASiis the i-th adaptation strategy that can be 1 if activated
and 0 if non-activated. Additional constraints provided for i) the presence of landscape constraints1that prevent the installation of fans and anti-hail nets; ii) conflict between organic farming certi-fication and the use of phytosanitary treatments; iii) the exclusion of the simultaneous presence of irrigation practices and emergency irrigation (i.e., if afixed irrigation plant is installed, no emergency irrigation is needed) and iv) the priority given to Chianti Classico procedural guidelines over the consequences of other adaptation
Table 2
Schematisation of adaptation strategies.
Adaptation strategies Notes and effect/s
Anti-hail net Effect on hail
Organic farming E.g. grassing or modification of pruning. Effect on drought and heat waves
Organic farmingþ certification E.g. grassing or modification of pruning. Effect on drought and heat waves. Effect on ORG
Fans Effect on frost
Heater/candles Effect on frost
Under/over canopy irrigation Effect on frost
Cultivar substitution Effect on yield. Effect on DEN Fixed irrigation plant Effect on drought. Effect on DEN
Emergency irrigation Effect on drought
Disciplinary modification Effect on DEN
Increase of phytosanitary treatments Effect on phytopathologies
Insurance Restore net revenues in case of extreme events damage (multi-risk insurance)
1 Italian legislation protects the landscape and promotes its enhancement. On the
basis of these assumptions, the law has adopted the“Code of the cultural and landscape heritage” approved by Legislative Decree 42/2004 and subsequent leg-islative, corrective or supplemental decrees. In this sense, if landscape constraints are currently or potentially active in a national area, some facilities and in-terventions could not be allowed in order to protect scenic beauty.
strategies. Procedural guidelines modification supposes that fixed irrigation plant and cultivar substitution are allowed without loss of DOCG certification.
A modelling of the above constraints is shown in Algorithm 11.
Due to the presence of non-continuous function, the model can be solved with a metaheuristic approach. In case of high complexity and difficulty of optimisation, classical problem solving approaches may require significant computational resources (time) and guar-antee optimal solutions only for small-scale problems (Bianchi et al., 2009). In contrast, techniques based on metaheuristics usu-ally find good e and sometimes optimal e solutions in less computation time. In this study, the model was solved using a ge-netic algorithm (Differential Evolution and Particle Swarm - DEPS), which applies Particle Swarm optimisation and a Differential Evo-lution algorithm to improve the system (Parsopoulos et al., 2004). DEPS was used because of its stability, good performance and multiple applications in complex systems (Petalas et al., 2007). A DEPS algorithm was applied because it is easily used in an open-source Calc spreadsheet of Apache OpenOffice software. Calc can be equipped with the DEPS Evolutionary Algorithm Solver (Xie, 2009), an extension with a user-friendly graphical interface and algorithms to solve non-continuous and complex problems. The DSS was implemented in Calc software.
3. Results
Table 3reports the expected yield reduction (Eqs.(2) and (6)), variation of quality (Eq.(7)) and economic parameters (Eqs.(1) and (5)) for current and future scenarios without adaptation strategies. The output is displayed per unit of surface (1 ha).
Compared to today, the RCP4.5 projection at 2050 predicts a strong impact on wine yields (18%) due to the variation of average climatic data. In terms of extreme events, the greater risk of yield loss is related to drought, frost, hail and heat waves. In the future,
the negative effects of these extreme events (with the exception of frost) seem to increase. A strong progression is linked to heat waves. Quality does not seem to be significantly influenced (only a small improvement was shown:þ1%).
A reduction in revenue and cost is projected due to yield reduction. The combination of these variables leads tofinancial damage (losses of NR) from 10,853 to 26,989V/ha y1, depending on the level of damage for distinct extreme events.
Due to the importance of landscape in the studied area from a socio-economic perspective, the analysis of adaptation strategies took into account the following scenarios (SC) in the optimisation model:
- SC1: current climatic condition; no landscape constraint; - SC2: current climatic condition; landscape constraint; - SC3: future climatic condition; no landscape constraint; - SC4: future climatic condition; landscape constraint.
Output refers to the average level of damage to grapevines due to each extreme event, equal to 15% (Fig. 2).
The number of relevant adaptation measures increases with the value of production. As expressed inFig. 1(dashed lines) and in
Table S2, the implementation of adaptation schemes indirectly leads to additional costs. These costs can be covered through a combination of increased wine prices and decreased damage. The optimal combination is identified via metaheuristic algorithm.
Organic farming is highlighted as a fundamental strategy in all scenarios, but currently no in-depth study of organic farming cer-tification influences on buyers exists (Abraben, 2014). Frost ex-tremes can be overcome with fans only when no landscape constraints are in place. If climate change occurs or if a landscape has to be protected, under/over canopy irrigation seems to be the strategy to adopt to cope with frost effect. Emergency irrigation is another adoptable measure in current scenarios. The strong impact
Table 3
Expected impact on wine quantity, quality and economic indexes based on current/future scenario and level of damage defor each extreme event.
Level of damage (15%) Level of damage (25%)
Current scenario RCP4.5 scenario at 2050 Current scenario RCP4.5 scenario at 2050
E(YRave)a e 18% e 18% E(YRfrost)b 9% 6% 15% 10% E(YRdrought)b 13% 15% 22% 25% E(YRheat_waves)b 5% 15% 8% 25% E(YRhail)b 7% 9% 11% 14% E(YRphytopathology)b 1% 2% 2% 3% DVQc e 1% e 1% Revenues (V/ha y1) 29,614 17,588 19,076 229 Costs (V/ha y1) 20,458 19,286 19,543 18,063 NR (V/ha y1) 9155 1698 467 17,834
aPercentage variation in respect to current situation. E(YR
ave)¼(1-Yf/Yc),100. It is independent from the level of damage due to extreme events. b Percentage variation in respect to hypothetical situation without extreme events, i.e. P(e)¼ 0.
c Percentage variation in respect to current situation. It is independent from the level of damage due to extreme events.
IF ðLAN ¼ 1Þ THEN ðASn¼ 0Þ∧
ASf ¼ 0 IF ASo¼ 1∨ASg¼ 1 THEN ðASt¼ 0Þ
IF ðASr¼ 1Þ THEN ðASe¼ 0Þ
IF ðASd¼ 0Þ∧ðASr¼ 1∨ASc¼ 1Þ THEN ðDEN ¼ 0Þ ELSEðDEN ¼ 1Þ
of climate change makes emergency irrigation insufficient for economic efficiency over the middle-term. Future scenarios stress that fixed irrigation plants should be installed in vineyards, but procedural guidelines must be modified for winemakers to retain high quality certification (DOCG). Cultivar substitution is a very expensive activity from an economic and organisational viewpoint (see the increase of 125% with respect to current values for amor-tisation costs inTable S2). The results confirm that this strategy should be activated only when climate change impacts occur and wine prices are at least 20V/l. Established insurance variables are the following: i) annual multi-risk insurance premium: 300V/ha; ii) deductible: 30% and iii) public funding: 50%. With these setting of parameters insurance seems to be an invalid measure. The high quality of Chianti Classico production suggests that yield mainte-nance is probably preferable to damage coverage. Such as example, a preliminary test developed by DSS demonstrated how in climate change scenarios insurance only becomes convenient when de-ductibles decrease to 10% (with a 50% of public funding).
4. Discussion
The results highlight that in the studied area climate change can have a negative impact on wine quantity due to the variation of average climatic data and extreme events. A small positive impact is forecast for wine quality. Quality is also a key indicator in the selection of adaptation measures. Highly valued wines (strictly related to quality) show a higher potential for the adoption of adaptation strategies. The good rating of Chianti Classico also suggests that maintaining a specific level of production seems to be preferable to insurance unless favourable conditions for farmers are met, e.g. a low percentage of deductibles proposed by insurance companies. Future studies could analyse and investigate in-depth
this aspect as well as proper insurance schemes.
Among adaptation strategies, the results suggest a combination of fans, anti-hail nets and emergency irrigation for current condi-tions. Forecasted climatic variations suggest winemakers may integrate or substitute these measures withfixed irrigation plants. Normative interventions (i.e., modification of Chianti Classico procedural guidelines) will enable the introduction of these facil-ities. When landscape constraints are in place, the risk of frost damage may be overcome with under/over canopy irrigation, which has less impact on the landscape even if it is more expensive than fans. According to results, organic farming seems to be a suitable intervention in all scenarios. Additional analyses should estimate the real influence of organic certification on wine prices. This study'sfindings suggest the need for procedural guidelines modifications to counteract winemakers' potential loss of denom-ination due to new vine cultivars orfixed irrigation practices.
A major assumption in this study is that farmers are rational profit maximisers. However, the actual situation and empirical evidence show that diverse levels of risk aversion/propensity exist among farmers. Future studies might integrate risk aversion into economic models. Different objective functions (not only economic ones) could also be introduced (e.g., maintenance of familiar ac-tivity, improvement of ecosystem services, etc.).
The limited number of interviewed in the focus group to implement CM (nine people) is suitable based onMorgan (1997)
that affirmed how a number of 6e10 peoples is adequate for focus group.€Ozesmi and €Ozesmi (2004)also stated that the num-ber of new variables in a CM added per interview levels off within a small number of interviews. Future studies should validate per-ceptions of measures and related impacts. This lack is acceptable due to the purpose and approach of the current study (i.e., intro-duction and testing of a new methodology and exploratory
analysis). New studies may consider technological and agronomical advances to update the types and influences of adaptation strate-gies. Results on the need of deductibles and public funding to have insurance schemes is mainly based on the assumptions made. More policy and subsidies-oriented scenario and approaches should also be investigated and used to support policy making.
The DSS (optimisation model) could be applied at the farm level because it represents a farm accounting model. It could also be extended to the Chianti Classico district through the application of data on climatic and economic averages, promoting local devel-opment and farm improvement. The application of the model to other parts of the wine production and supply chain and other agricultural sectors needs further analysis. The DSS could be extended to other regions, but a verification of similar input models is recommended (e.g., due to potential differences in quality sys-tems, procedural guidelines and norms).
The DSS is available on request by emailing the corresponding author.
5. Conclusions
The scientific contribution of the paper lies in the imple-mentation of optimised adaptation strategies to cope with climate change in the wine sector. The implemented Decision Support System (DSS) allows for the study of economic performance and adaptation strategies for wine farms in current climatic conditions and in the event of climate change. Impacts on thefield phase of the production process can be evaluated through a mix-method approach. The present study employed different techniques to assess the probability of future events and stakeholder perceptions on the topic. From a practical viewpoint, cognitive mapping facili-tates the analysis of climate change impacts on the wine industry using a complex system approach. The introduction of operational research techniques (i.e., metaheuristic optimisation) in a complex system framework allows DSS' users to address hypothetical (what-if) scenarios. The proposed tool can be primarily applied to indi-vidual farms to highlight interventions and their effect on a farm's economic output. A district-level application can promote the local development of the wine sector and normative modifications.
To optimise the model in future analyses, researchers may consider different objective functions and the updating of adapta-tion strategies. The present study analysed“active” and “on site” adaptation strategies. Future “active” forms of adaptation could focus on the technological aspects of climate-smart agriculture (Long et al., 2016) as well as winemaking operations. Site selection is another potential measure to cope with climate change. The current study did not consider a shift of location in evaluating the risk for current vineyard sites. In future research, an evaluation of consumer satisfaction related to product quality could also be introduced.
Acknowledgements
This work is part of the project“Climate change and wine sector in Tuscany: scenario assessment and short- to medium term adaptation strategies [Cambiamenti climatici e sistema vitivinicolo toscano: scenari evolutivi e prospettive di adattamento di breve e lungo periodo]” funded by the Ente Cassa di Risparmio di Firenze Foundation (grant nº 2013/0684). The authors wish to acknowledge the Foundation for its contribution to the research.
Appendix A. Supplementary data
Supplementary data related to this article can be found athttp:// dx.doi.org/10.1016/j.jclepro.2017.08.095.
References
Abraben, L.A., 2014. The Value of Organic Wines from Tuscany: a Hedonic Price Analysis. University of Central Florida Orlando, Florida. http://ufdc.ufl.edu/ UFE0046791/00001. (Accessed 31 July 2017).
Ackermann, F., Eden, C., Cropper, S., 2004. Getting started with cognitive mapping. In: Proceedings of 7th Young OR Conference. University of Warwick, pp. 65e82.
www.banxia.com/pdf/de/GettingStartedWithCogMapping.pdf. (Accessed 28 February 2017).
Arunrat, N., Wang, C., Pumijumnong, N., Sereenonchai, S., Cai, W., 2017. Farmers' intention and decision to adapt to climate change: a case study in the Yom and Nan basins, Phichit province of Thailand. J. Clean. Prod. 143, 672e685.
Baldi, M., Ciardini, V., Dalu, J.D., De Filippis, T., Maracchi, G., Dalu, G., 2014. Hail occurrence in Italy: towards a national database and climatology. Atmos. Res. 138, 268e277.
Beermann, M., 2011. Linking corporate climate adaptation strategies with resilience thinking. J. Clean. Prod. 19, 836e842.
Bernetti, I., Menghini, S., Marinelli, N., Sacchelli, S., Alampi Sottini, V., 2012. Assessment of climate change impact on viticulture: economic evaluations and adaptation strategies analysis for the Tuscan wine sector. Wine Econ. Policy 1, 73e86.
Bianchi, L., Dorigo, M., Gambardella, L.M., Gutjahr, W.J., 2009. A survey on meta-heuristics for stochastic combinatorial optimization. Nat. Comput. 8 (2), 239e287.
Bindi, M., Miglietta, F., Gozzini, B., Orlandini, S., Seghi, L., 1997. A simple model for simulation of growth and development in grapevine (Vitis vinifera L.). I. model description. Vitis 36 (2), 67e71.
Bindi, M., Bellesi, S., Orlandini, S., Fibbi, L., Moriondo, M., Sinclair, T., 2005. Influence of water deficit stress on leaf area development and transpiration of Sangiovese grapevines grown in pots. Am. J. Enol. Viticult. 56 (1), 68e72.
Caffarra, A., Rinaldi, M., Eccel, E., Rossi, V., Pertot, I., 2012. Modelling the impact of climate change on the interaction between grapevine and its pests and path-ogens: European grapevine moth and powdery mildew. Agric.Agric. Ecosyst. Environ. 148, 89e101.
Chamber of Commerce, Industry, Crafts and Agriculture of Florence, 2016a. Weekly Prices. www.fi.camcom.gov.it/listini_prezzi_settimanali.asp?ln¼&idtema¼1& page¼informazioni&index¼1&idtemacat¼1&idcategoria¼6126. (Accessed 27 February 2017).
Chamber of Commerce, Industry, Crafts and Agriculture of Siena, 2016. Weekly Prices.www.si.camcom.gov.it/#div_contenuti. (Accessed 27 February 2017).
Chapman, G., 2009. Chaos and complexity. Int. Encycl. Hum. Geogr. 31e39.
Clarke, L., Edmonds, J., Jacoby, H., Pitcher, H., Reilly, J., Richels, R., 2007. Scenarios of Greenhouse Gas Emissions and Atmospheric Concentrations. Sub-report 2.1A of Synthesis and Assessment Product 2.1 by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research. Department of Energy, Office of Biological & Environmental Research, Washington, 7 DC., USA.
Clingeleffer, P., 2014. Terroir: the Application of an Old Concept in Modern Viti-culture. Reference Module in Food Science, Encyclopedia of Agriculture and Food Systems, pp. 277e288.
Collins, M., Knutti, R., Arblaster, J., Dufresne, J.L., Fichefet, T., Friedlingstein, P., Gao, X., Gutowski, W.J., Johns, T., Krinner, G., Shongwe, M., Tebaldi, C., Weaver, A.J., Wehner, M., 2013. Long-term climate change: projections, com-mitments and irreversibility. In: Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M. (Eds.), Climate Change 2013: the Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1029e1136.
Contini, C., Romano, C., Scozzafava, G., Boncinelli, F., Casini, L., 2015. Wine con-sumption and sales strategies: the evolution of Mass Retail Trading in Italy. Wine Econ. Policy 4, 116e127.
Costa, J.M., Vaz, M., Escalona, J., Egipto, R., Lopes, C., Medrano, H., Chaves, M.M., 2014. Modern viticulture in southern Europe: vulnerabilities and strategies for adaptation to water scarcity. Agric.Agric. Water Manag. 164 (1), 15e18.
Creswell, J.W., 2013. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. SAGE Publications, p. 246. ISBN 0761924426, 9780761924425.
Dasgupta, P., Morton, J.F., Dodman, D., Karapinar, B., Meza, F., Rivera-Ferre, M.G., Toure Sarr, A., Vincent, K.E., 2014. Rural areas. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken, S., Mastrandrea, P.R., White, L.L. (Eds.), Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 613e657.
DePoy, E., Gitlin, L.N., 2016. Mixed method designs. In: DePoy, E., Gitlin, L.N. (Eds.), Introduction to Research - Understanding and Applying Multiple Strategies, fifth ed. Elsevier, pp. 173e179.
Derbyshire, J., 2016. The implications, challenges and benefits of a complexity-orientated futures studies. Futures 77, 45e55.
De Salvo, M., Begalli, D., Capitello, R., Signorello, G., 2015. A spatial micro-econometric approach to estimating climate change impacts on wine firm performance: a case study from Moldavia region, Romania. Agric.Agric. Syst. 141, 48e57.
Dubey, P.K., Singh, G.S., Abhilash, P.C., 2016. Agriculture in a changing climate. Book review. J. Clean. Prod. 113 (1), 1046e1047.
Dulic, A., Angel, J., Sheppard, S., 2016. Designing futures: inquiry in climate change communication. Futures 81, 54e67.
Fischer, E.M., Sch€ar, C., 2010. Consistent geographical patterns of changes in high-impact European heatwaves. Nat. Geosci. 3, 398e403.
Fleming, A., Rickards, L., Dowd, A.M., 2015. Understanding convergence and divergence in the framing of climate change responses: an analysis of two wine companies. Environ. Sci. Policy 51, 202e214.
Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., Santos, J.A., 2013. Future scenarios for viticultural zoning in Europe: ensemble projections and uncertainties. Int. J. Biometeorol. 57 (6), 909e925.
Hadarits, M., 2011. Climate change and economic uncertainty in the wine sector: a case study of the Maule region, Chile. In: Leal Filho, W. (Ed.), The Economic, Social and Political Elements of Climate Change. Climate Change Management. Springer, Berlin, Heidelberg, pp. 245e265.
Henly-Shepard, S., Gray, S.A., Cox, L.J., 2015. The use of participatory modeling to promote social learning and facilitate community disaster planning. Environ. Sci. Policy 45, 109e122.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A., 2005. Very high reso-lution interpolated climate surfaces for global land areas. Int. J. Climatol. 25, 1965e1978.
IPCC, 2012. In: Field, C.B., Barros, V., Stocker, T.F., Qin, D., Dokken, D.J., Ebi, K.L., Mastrandrea, M.D., Mach, K.J., Plattner, G.-K., Allen, S.K., Tignor, M., Midgley, P.M. (Eds.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation, Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, and New York, NY, USA.
IPCC, 2014. In: Writing Team, Core, Pachauri, R.K., Meyer, L.A. (Eds.), Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland.
Jones, G.V., White, M.A., Cooper, O.R., Storchmann, K., 2005. Climate change and global wine quality. Clim. Change 73, 319e343.
JRC, 2014. Gridded Agro-Meteorological Data in Europe. Version 2014-1.0, published 14/01/2014. agri4cast.jrc.ec.europa.eu/DataPortal/Index.aspx?o¼d. (Accessed 28 February 2017).
Kosko, B., 1986. Fuzzy cognitive maps. Int. J. Man Mac. Stud. 24, 65e75.
Lereboullet, A.L., Beltrando, G., Bardsley, D.K., 2013. Socio-ecological adaptation to climate change: a comparative case study from the Mediterranean wine in-dustry in France and Australia. Agric.Agric. Ecosyst. Environ. 164, 273e285.
Long, T.B., Blok, V., Coninx, I., 2016. Barriers to the adoption and diffusion of tech-nological innovations for climate-smart agriculture in Europe: evidence from The Netherlands, France, Switzerland and Italy. J. Clean. Prod. 112, 9e21.
Marone, E., Bertocci, M., Marinelli, N., Boncinelli, F., 2014. Full cost analysis for the creation of managerial benchmarks in the wine sector: a case study in Tuscany. In: Proceedings of Academy of Wine Business Research 8th International Con-ference, pp. 1e7. Geisenheim, Germany.
Mastrandrea, M.D., Field, C.B., Stocker, T.F., Edenhofer, O., Ebi, K.L., Frame, D.J., Held, H., Kriegler, E., Mach, K.J., Matschoss, P.R., Plattner, J.K., Yohe, G.W., Zwiers, F.W., 2010. Guidance Note for Lead Authors of the IPCC Fifth Assessment Report on Consistent Treatment of Uncertainties. Intergovernmental Panel on Climate Change (IPCC).www.ipcc.ch. (Accessed 2 March 2017).
Ministero delle Politiche Agricole Alimentari e Forestali - MIPAAF, 2016. National Agricultural Insurance Plan 2016. www.politicheagricole.it/flex/cm/pages/ ServeBLOB.php/L/IT/IDPagina/4617. (Accessed 1 March 2017).
Mirzabaev, A., Nkonya, E., von Braun, J., 2015. Economics of sustainable land management. Curr. Opin. Environ. Sust. 15, 9e19.
Morgan, D.L., 1997. In: Focus Groups as Qualitative Research, vol. 16. Sage University Press, Thousand Oaks, CA.
Moriondo, M., Bindi, M., Fagarazzi, C., Ferrise, R., Trombi, G., 2011. Framework for high-resolution climate change impact assessment on grapevines at a regional scale. Reg. Environ. Change 11, 553e567.
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P., Carter, T.R., Emori, S., Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B., Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J., Thomson, A.M., Weyant, J.P.,
Wilbanks, T.J., 2010. The next generation of scenarios for climate change research and assessment. Nature 463, 747e756.
Mozell, R.N., Thach, L., 2014. The impact of climate change on the global wine in-dustry: challenges& solutions. Wine Econ. Policy 3 (2), 81e89.
Neill, K.D., 2011. A Hedonic Analysis of the Effect of Expert Wine Ratings on Price and Retailer Profits. University of Central Florida Orlando, Florida.
Nicholas, K.A., Durham, W.H., 2012. Farm-scale adaptation and vulnerability to environmental stresses: insights from winegrowing in Northern California. Glob. Environ. Chang. 22, 483e494.
Nikolaou, I., Evangelinos, K., Leal Filho, W., 2015. A system dynamic approach for exploring the effects of climate change risks onfirms' economic performance. J. Clean. Prod. 103, 499e506.
€Ozesmi, U., €Ozesmi, S.L., 2004. Ecological models based on people's knowledge: a multi-step fuzzy cognitive mapping approach. Ecol. Model. 176, 43e64.
Papageorgiou, E.I., Markinos, A.T., Gemtos, T.A., 2011. Fuzzy cognitive map based approach for predicting yield in cotton crop production as a basis for decision support system in precision agriculture application. Appl. Soft Comput. 11, 3643e3657.
Parsopoulos, K.E., Papageorgiou, E.I., Groumpos, P.P., Vrahatis, M.N., 2004. Evolu-tionary computation techniques for optimizing Fuzzy Cognitive maps in radi-ation therapy systems. Lect. Notes Comput. Sci. 3102, 402e413.
Perez-Blanco, C.D., Mysiak, J., Gutierrez-Martìn, C., De Salvo, M., 2014. What Role for Income Stabilisation Insurance in EU Agriculture? the Case of the Regione Emilia Romagna in Italy. Centro Euro-Mediterraneo sui Cambiamenti Climatici. Research Papers RP0242.
Petalas, Y.G., Parsopoulos, K.E., Papageorgiou, E.I., Groumpos, P.P., Vrahatis, M.N., 2007. Enhanced learning in fuzzy simulation models using memetic particle swarm optimization. In: Proceedings of the 2007 IEEE Swarm Intelligence Symposium (SIS 2007), pp. 16e22.
Pomarici, E., Seccia, A., 2016. Economic and social impacts of climate change on wine production. Ref. Module Food Sci. http://dx.doi.org/10.1016/B978-0-08-100596-5.03062-6.
Sacchelli, S., Fabbrizzi, S., Menghini, S., 2016. Climate change effects and adaptation strategies in the wine sector: a quantitative literature review. Wine Econ. Policy 5, 114e126.
Samet, R.H., 2011. Exploring the future with complexity science: the emerging models. Futures 43, 831e839.
Samet, R.H., 2012. Complexity science and theory development in the futuresfield. Futures 44, 504e513.
Sayama, H., 2015. Introduction to the Modeling and Analysis of Complex Systems. Open SUNY Textbooks, Milne Library State University of New York.
Schultz, H.R., Jones, G.V., 2010. Climate induced historic and future changes in viticulture. J. Wine Res. 21, 137e145.
Serrao-Neumann, S., Schuch, G., Harman, B., Crick, F., Sano, M., Sahin, O., van Staden, R., Baum, S., Low Choy, D., 2015. One human settlement: a trans-disciplinary approach to climate change adaptation research. Futures 65, 97e109.
Snyder, C.W., Mastrandrea, M.D., Schneider, S.H., 2011. The complex dynamics of the climate system: constraints on our knowledge, policy implications and the necessity of systems thinking. Philos. Complex Syst. 10, 467e505.
Stock, M., Gerstengarbe, F.W., Kartschall, T., Werner, P.C., 2005. Reliability of climate change impact assessments for viticulture. Acta Hortic. 689, 29e39. Tuscany Plant protection service, 2016. Crop Monitoring.agroambiente.info.arsia.
toscana.it/arsia/arsia. (Accessed 27 February 2017).
Wine Enthusiast Magazine, 2016. Vintage Chart.www.winemag.com/wp-content/ uploads/2016/01/Vintage_Chart_2016.pdf. (Accessed 28 February 2017). Xie, X.F., 2009. Solver for Nonlinear Programming. http://extensions.services.
openoffice.org/project/NLPSolver. (Accessed 28 February 2017).
Zhu, X., Moriondo, M., Ierland, E.C., Trombi, G., Bindi, M., 2016. A model-based assessment of adaptation options for Chianti wine production in Tuscany (Italy) under climate change. Reg. Environ. Change 16 (1), 85e96.
Zollo, A.L., Rillo, V., Bucchignani, E., Montesarchio, M., Mercogliano, P., 2016. Extreme temperature and precipitation events over Italy: assessment of high-resolution simulations with COSMO-CLM and future scenarios. Int. J. Climatol. 36, 987e1004.
