The full spectrum of climate change adaptation: testing an analytical framework in Tyrolean mountain agriculture (Austria)

Because of the observability of impacts of climate change (CC) in ecosystems, societies and economies (IPCC 2014), attention to adaptation has increased significantly for the past few years. Numerous scholars analyzed climate change adaptation (CCA) and literature grew at a rate of several hundred papers per year (Perry 2015). Although the concept of CCA has got so much attention from scholars all over the world, the limits and barriers in practice are still manifold (e.g. Adger et al. 2007; Dow et al. 2013; Ford and King 2015). Identifying such obstacles (e.g. Biesbroek et al. 2013; Moser et al. 2010) may be a promising approach to overcome them, but we could also question the role of our conceptualization of CCA and if it actually helps in overcoming such limits and barriers. CCA is seen as a “process of adjustment” (IPCC 2014) in response to climatic stimuli (Smit et al. 2000), but this often used definition reflects the “first-generation adaptation” approach that aims at particular adaptation solutions to particular CC problems (Boyd and Cornforth 2013). “Second-generation adaptation” offers a more comprehensive view as it also addresses the context in which hazards occur (Burton et al. 2002) and frames adaptation more in the sense of sustainability and resilience (e.g. Boyd et al. 2008; Ensor 2011; Eriksen and O’Brien 2007). Following an intentional perspective, a very common conceptual differentiation is that between autonomous and planned adaptation (for a more detailed description of different concepts see “Climate change impacts and mountain agriculture” section). Autonomous adaptation is seen as “spontaneous adjustment” (Carter et al. 1994), which “does not constitute a conscious response to climatic stimuli” (Malik et al. 2010). Whereas planned adaptations are “based on an awareness” (Malik et al. 2010) and “require conscious intervention” (Fankhauser et al. 1999). Although both types are recognized by several scholars (e.g. Smit et al. 2000), most empirical studies focus on planned adaptations (e.g. Füssel 2007; Kates et al. 2012; Moser 2009). However, adaptation in practice does not always follow the concept of planned adaptation, it is often reactive (Adger et al. 2005) and non-climatic stressors play a role for adaptability (McDowell et al. 2014).

Many scholars conceptualize adaptation processes as sequences of “ideal” adaptation phases, which usually start with problem definition, followed by planning processes, such as identification, prioritization and selection of actions, implementation and finally monitoring and evaluation (e.g. Cross et al. 2012; Mirfenderesk and Corkill 2009; Wheaton and Maciver 1999). This concept of planned adaptation focuses on climate change as key driver for adaptation and leaves hardly any room for other drivers although interactions between climate change and other driving forces, such as demographic change, globalization or social polarization are plain to see (Simonet and Fatoric 2015).

Against this background, we want to provide an analytic framework for a functional perspective on adaptation that considers all actions—intentional, semi-intentional and non-intentional—that improve a sector’s adaption to climate change. A case study from mountain agriculture in Tyrol (Austria) illustrates the applicability of the analytic framework. We selected the case, because mountains are among the most vulnerable regions in the world (Messerli and Ives 1997) and experience climate change impacts earlier and in a more pronounced manner (Ingold et al. 2010). Furthermore agriculture is a sector particularly affected by climate change. Agricultural production primarily takes place outdoors and directly depends on climatic conditions (Lobell et al. 2008; Vermeulen et al. 2012). Adaptation research in mountain regions has a strong focus on developing countries, especially Peru, Nepal and India (McDowell et al. 2014). Studies about mountain regions in ‘developed’ countries are the exception and often focus on tourism (e.g. Behringer et al. 2000; Hoffmann et al. 2009; Scott et al. 2007). CCA is also important to sustain sensitive mountain ecosystems in ‘developed countries’ and their multifunctional services going far beyond food production, such as water provision, biodiversity, landscape amenities, recreation or identity (Schermer and Kirchengast 2006).

With this paper, we aim to offer an analytic framework for an enhanced and extended CCA approach. After summarizing the main climate change impacts in mountain agriculture and CCA concepts, we present the “Study area and methods” section. In a next step, we discuss CC as one of several drivers in mountain agriculture. Then, we present the analytic framework that acknowledges hidden adaptations. We use the case study to illustrate the framework’s applicability. From the case study application, we derive some general conclusion on the conceptualization and empirical analysis of CCA.

Although CCA received so much attention from academic scholars, we cannot automatically assume this is also reflecting in farmers’ discourses. Several drivers steer agricultural production and marketing and many of them are much closer to farmers’ perceived priorities than climate change. However, CCA literature usually focuses on narrow conceptualizations of CCA (see “Conceptualizing climate change adaptation” section) that basically ignore other constraints and drivers of the sector.

The analytic framework acknowledging hidden climate change adaptation

The following framework should support not only the analysis of planned and anticipatory CCA, but also the identification of all actions that can contribute to CCA. This model divides CCA actions into three different categories: explicit, multi-purpose and hidden CCA actions (see Fig. 2). The basic idea is to distinguish between the (main) intentions motivating every action. In general, the influence of climate change as motivator for adaptation decreases from explicit to hidden actions (explicit CCA—strong influence, hidden CCA—weak/no influence). The other way round, the influence of non-climatic drivers increases from explicit to hidden actions (see Fig. 2). Explicit actions are typical CCA measures, which are initiated in direct response to climate change and which are addressed by common CCA definitions. There is no doubt that these are CCA efforts since they are usually also labelled as CCA. Besides climate change, other drivers, such as changing prices or policies influence multi-purpose actions. In contrast to explicit CCA, these actions are not only driven by climate change, although the change of climatic conditions may be of great importance for the initiators. The CCA labelling varies—some are clearly labelled as CCA, while others are not. Hidden actions are only motivated by non-CC drivers, such as policies, markets, or life-style changes. CCA was not an intention, but occurs as an unplanned side-effect. Of course, these actions are not labelled as CCA. Although these actions do not fit within common CCA definitions, they can significantly contribute to an agricultural sector well adaptable to climate change.

Of course, the transitions between the three categories are fluid and a clear assignment is sometimes difficult. The main purpose of this analytic framework is to support the empirical analysis of actions contributing to CCA, although they were motivated by other drivers. This integrative analysis is important to include all different types of measures contributing to CCA. Furthermore, those actions that were motivated by other drivers hold potential for learning on how CCA strategies can be better tied up to farmers’ everyday life. The framework can be extended to include the different effects of climate change: incremental changes, extreme weather events and global effects of CC (see Fig. 4). This further differentiation can also focus on different policy level, different sectors or actors. The differentiation in CCA degrees and CC effects (as shown in Fig. 4) helps identifying CCA gaps and to check if all sectors, impacts etc. are targeted or if the allocation of activities to different organizations or policy levels is unbalanced. We apply the analytic framework to the practical example of Tyrolean mountain agriculture to test its applicability and usefulness.

Table 2 gives an overview of all CCA actions identified from interviews and document analysis. As pointed out in the “Methods” section, we focused on actions within the Federal Province of Tyrol, which means that national and EU level actions were not considered. Austrian provinces have a relatively broad scope of action regarding CCA since many relevant competences are allocated to the nine regional governments in Austria (e.g. Spatial Planning Law, Nature Protection Law, social policies). Furthermore, the provinces are responsible for the implementation of national and EU policies on agriculture, water or soil protection.

As expected and conceptualized, the connections of the above actions to CCA are often “hidden” and not at all explicit. For example, the association of farm women supports the development of humans and organizations in agriculture and in rural areas in the face of changing production and living conditions. This association strengthens social capital within and social learning among the farmers’ community and therefor enhances adaptation processes and the sector’s resilience. Better connected farmers are better able to learn from each other and to help each other—also in case of extreme weather events. Additionally, knowledge and experiences in dealing with climate change impacts and adaptation can be shared.

Figure 4 shows the classification of the CCA actions from Tyrolean mountain agriculture. Some CCA actions can contribute to more than one CC effect; these are located directly on the transition line (e.g. action no. 9, 15, 20). It has to be added, that the classification of the CCA measures to a CCA category and to one of the three CC impacts was not completely straight forward in every case due to some scope for interpretation. The allocation was guided by the objective to provide an overview on all CCA actions identified for the Tyrolean agriculture sector and to identify focal areas as well as neglected areas.

The division into the three categories of CCA shows that eight out of totally 29 measures can be considered as explicit CCA—actions in direct and only response to CC. Only two actions are categorized as multi-purpose CCA and a clear majority of 19 actions belong to hidden CCA. This concentration of hidden CCA deserves particular attention as these actions do not count as CCA in the common view. The second differentiation is relating to the different effects of CC. A clear allocation to one specific field was neither possible nor meaningful for those 14 CCA actions that address two CC effects and are thus located on the transition line between those two. 15 actions were directly allocated to one field of which five were associated with incremental changes. Global effects of CC and extreme weather events got five actions each. It is striking that there are no explicit and even no multi-purpose adaptations which address global CC, which means all eleven CCA relevant actions categorized in the field of global effects of CC (exclusively and on the transition line to the other fields) are hidden adaptations. This shows the exclusion of these effects in planned and anticipatory CCA although some strategies highlight their relevance.

The results from the case study show that regional adaption actions are very broad and occur in various ways. As in other studies (e.g. Comoé and Siegrist 2015), interviewees expect that most Tyrolean farmers are aware and already have observed impacts of climate change. Nonetheless, only a few explicit CCA were identified. More than half of the identified actions were not motivated by climate change. Such hidden or multi-purpose did not gain much recognition in past CCA literature. Even autonomous CCA gained only little attention in literature (Thornton and Manasfi 2010), as most scholars focus on planned CCA (e.g. Füssel 2007; Geneletti and Zardo 2016). Bonzanigo et al. (2014) state “it is crucial to include an assessment of farmers’ autonomous adaptation into the design and evaluation of rural policy measures.” It can be assumed that hidden and multi-purpose CCA action do not only play a role in Tyrolean agriculture, but will also be relevant in other sectors and elsewhere. These “hidden CCA” can be considered as autonomous from the perspective of climate change, but might be clearly planed and motivated from other intentions, such as market stabilization or nature conservation. Tompkins et al. (2010) modified an IPCC definition by adding “whether or not motivated by climate” for their assessment of adaptation in the UK and thus considered hidden CCA. It is remarkable that the majority of hidden adaptations from the case study promise an increase of the sector’s resilience by strengthening the farmer’s adaptability through decentral, self-organized social networks for autonomous co-learning and experimentation, while many adaptation measures in literature focuses on technical solutions such as climate information systems or irrigation (Smit and Skinner 2002).

The categorization of CCA actions in the Tyrolean agriculture sector highlight according to the analytic framework highlight two key gaps: firstly multi-purpose actions and secondly actions in the field of global effects of climate change. The underrepresentation of multi-purpose actions reflects the current reductionist approaches to CCA which hardly considers sector specific drivers or constraints. It also shows that the idea of mainstreaming adaptation which aims at integrating adaptation into established or on-going policies, development strategies or management plans (Agrawala 2006) is not yet established. Although it is obvious that global climate change will also impact the regional level (IPCC 2014), adaptation actions in this field are rare. In the liberalized Common Agricultural Policy of the EU, farmers are heavily affected by volatile world market prices (both on markets for production inputs as animal feed, energy as well as output markets for their products) and thus vulnerable to global CC, too. In this line of thought, farmers might be more aware of this indirect effects by focusing on policy and market changes. In contrast, the Austrian Strategy for Adaptation discusses the influence of global conditions on agriculture is discussed, but proposes no specific actions.

It may be pointed out that it cannot be totally clarified to what extent the listed actions actually contribute to CCA. As some contexts to CC and adaptation are very broad, the relevance of the listed actions in Table 1 for CCA may not be clear at first sight. There are already some experiences with ex-ante evaluations of CCA measures, but such endeavors are complex as indicators alone are not enough and specific contexts have to be taken into account (Klostermann et al. 2015). A comprehensive evaluation was outside the scope of this analysis, which focused on testing the applicability of the analytic framework based on a first inventory of actions that can be directly or indirectly connected to climate change impacts.

The presented analytic framework and its underlying methods were applied to the case of Tyrolean agriculture, but experienced researchers could also easily adapt it for application in any other region or for other sectors, such as tourism or forestry. Such an integrative, sector and regional context specific assessment of CCA actions that also takes into account hidden adaptations could be useful for the identification of gaps in already existing adaptation strategies to better focus scarce public resources on those activities with the highest leverage. One of the strengths of the presented framework is its ability to show the full spectrum of the sector’s adaptation practice which includes intentional as well as unintentional adaptations to CC and other drivers. For an application of this approach for other regions and sectors sufficient resources are necessary as well as insights into and sensitivity for the specific geographical and sectoral context. Interview partners should be selected carefully as they have a big influence on the results. The qualitative interviews need interviewers experienced in qualitative research who are able to generate the atmosphere of a trustful, reflective and honest exchange. About 60–90 min per interview plus time for preparation and follow-up processing have to be calculated. Furthermore, transcriptions and analysis of the results are very time-consuming. But such a context-sensitive and integrative CCA assessment provides much needed insights for a more effective and efficient allocation of scarce public resources for CCA.

Effective adaptation actions should be tailored to regional conditions (Reyer et al. 2012), and consider regional actors’ needs, perceptions and motivations to address the limited “intent to implement” as a notable barrier (Öhlmer et al. 1998). Adger and Barnett (2009) state that “adaptive capacity will not necessarily translate into action”, it needs an initial impulse which could be given through motivation.

The concept of hidden CCA can be connected to the concept of second generation adaptation which does not primarily focus on predicted biophysical threats, but considers special contexts and underlying vulnerabilities (Boyd and Cornforth 2013). The idea of mainstreaming adaptation takes a similar direction and aims at integrating adaptation issues into ongoing (policy) processes to consider sector specific contexts (UNEP 2011). The common idea is to see CCA in the bigger context of sector-specific drivers and motivations, which may help to push the sector’s adaptability. Poly-rational multi-purpose adaptations, as also highlighted by the framework, directly consider climate change next to other sector-specific drivers.

The focus on technical solutions and adaptation measures (esp. in agriculture, e.g. Bird et al. 2016; Esteve et al. 2015) also appears to be problematic. We are more and more often challenged with infrequent and unpredictable extreme events which may result in power failure, but technical solutions mostly depend on electrical power. Many of the hidden CCA are “soft” measures, focusing on long term learning processes, self-organization, local knowledge and adaptive governance, which are considered as key for CCA (Adger et al. 2005; Adger 2010; Wolf 2011).

An increased focus on hidden adaptations, supported by the concepts of adaptive governance and resilience could guide the way to agricultural systems with a higher degree of adaptability that is supported by self-organized networks for social-learning, experimentation and exchange of available resources for coping with unexpected situations. Including these aspects into CC scenarios could also improve the scenarios’ validity and create closer links between CC scenarios and the every-day life of their potential implementers and users.

It is hardly possible to quantify the role of hidden adaptations in numbers. However, the Tyrolean case showed that hidden actions, which have been implemented for diverse motivations, do exist whereas explicit CCA action is scarce. It may be assumed that such hidden CCA actions can be found in other regions and in other sectors as well. This article showed that the suggested framework extending the CCA concept to multi-purpose and hidden adaptive practices can support an integrative perspective on CCA that better ties up to farmers’ realities.