Agriculture in the Dutch Delta region: The effects of droughts on the agricultural sector in the Dutch delta region

Agriculture in the Dutch Delta region 

22-12-2019

Koelewijn, Saskia

Business Administration

11640731

Oers, Anne van

Artificial Intelligence

11867515

Stoof, Tamara

Earth Science

11291664

Westerhof, Lisan

Earth Science

11763167

Supervisor

de Jong. V

Wordcount

7953

Future Planet Studies 2019-2020

Interdisciplinary Project

Abstract

This interdisciplinary report researches the consequences of recurring droughts for agriculture in the Dutch delta region in the future. To be able to assess what those consequences will be, insights from three different disciplines are implemented in this research, namely artificial intelligence, earth sciences and business administration. Artificial intelligence selected the most suitable model to analyse droughts in the Dutch delta region and the parameters used by this model were then analysed by the earth science discipline. The most important conclusion of this analysis was that as droughts are increasing, groundwater levels will lower and soil quality will decrease, causing major implications for agriculture. Finally, the effects of those implications on the profitability of potato farmers in the Dutch delta was analysed by the business administration discipline, using Porter’s five forces. At least three of those five forces are very likely to increase, which will cause the profitability of the potato farmers in the Dutch delta region to decrease. Therefore, this paper concludes that drought formation in the Dutch delta region would result in a reduced soil quality and agricultural yield, negatively affecting the profitability of the agricultural sector in the Dutch delta region.

Index

Discussion and recommendations 23

References 25

Appendix 29

Introduction

The droughts of 2018 and 2019 in the Netherlands have been a cause for concern. While water management plays an important role in Dutch society, it is focussing on protecting the land from flooding rather than coping with water deficiency (Lintsen, 2002). A water deficiency occurs when the groundwater levels have dropped below a certain point where water demand exceeds water supply (Peters, Torfs, Van Lanen & Bier, 2003). While the Dutch summers of 2018 and 2019 did not break the drought record of 1976 (KNMI, n.d.), the trend of increasingly warm and dry summers in the Netherlands is concerning (Beersma & Buishand, 2007). The long term effects of periodic droughts on Dutch society is unknown. However, the drought of 2018 has resulted in many problems for agriculture, some of which had long term consequences (Van Duinen et al. _{​, 2015). It is therefore important to analyze} possible threats if the droughts prove themselves to be a persistent feature of the future climate of the Netherlands.

By analyzing the threats it is possible to search for suitable mitigation measures in order to avoid future drought related problems. For example, the importance of water availability and soil salinization on agricultural sustainability should not be neglected. The fertility of arable soils is greatly influenced by water availability and soil quality; two factors that are greatly influenced by groundwater levels (_​Burk & Moench, 2000) _​. A change in groundwater levels can therefore eventually result in both food insecurity as well as financial insecurities in the agricultural sector (Kuenzer & Renaud, 2012).

This report uses existing climate models to predict the consequences of droughts on the agricultural sector in the Netherlands by addressing hydrological changes related to drought, and the subsequent changes in soil quality. Various neural networks regarding prediction of crop growth will be analysed to find one that is most suitable for this particular case. However, due to lack of suitable soft- and hardware to run these models, they will not actually be run, but thoroughly analysed in order to provide the information needed for this research. The parameters used by the best models will function as a guideline throughout the research. Finally, the physical consequences of drought will be linked to the security of the profitability of the agricultural sector in the future based on the expected consequences. This all results in the following research question: “ _{​What are the consequences of recurring} drought for the future of the agricultural sector in the Dutch delta region?”.

As the research question indicates, this paper will focus on the Dutch part of the delta region (figure 1). This paper will address the relevant theories concerning droughts and its subsequent relationship to agriculture from different perspectives. One of these theories is the I/O model, which needs a specific industry to analyse. Since the agricultural sector is a broad industry, this paper will focus on the potato industry for this analysis. Potatoes are agricultural products which are relatively sensitive to water stress (Van Loon, 1981; Panigrahi, Panda & Raghuwanshi, 2001). Therefore, potatoes are useful for this paper in order to get a detailed view of how changes in deltas affect the agricultural sector. After discussing the theories used for the research, the problem definition, the interdisciplinary approach and the methodology will be discussed. This paper will end with the results and a conclusion.

Figure 1. The Delta Region (TKI Dinalog, 2019). This research will focus on the Dutch part of the Delta Region

Theoretical framework

In order to study the possible consequences of recurrent droughts in the Netherlands on the future of agriculture in the Dutch delta region, it is necessary to understand how droughts influence the factors that decide agricultural prosperity. This section will provide an overview of the most important factors contributing to or influenced by droughts and the subsequent consequences on agriculture in the Dutch delta region. The focus will be on the interrelation between these factors in order to emphasize the complexity of the hazard. These influencing factors will be determined by the use of Artificial Neural Networks (ANN) in the analysis. ANN is a tool to solve complex problems, by means of extensive information processing (Basheer & Hajmeer, 2000).

ANN provide an easy way to measure and predict potato yield (Alvarez, 2009; Kaul, Hill, & Walthall, 2005). They have been proven to be more accurate in predicting wheat (Alvarez, 2009), soybean and corn yield (Kaul, Hill, & Walthall, 2005) than models that use linear regression (Mahajan, Das, & Sardana, 2015). However, as is often the case, ANN models are not readily available. Hence, the focus of this research lies on identifying the parameters used by the models to measure and predict the yield, rather than using the models to actually predict potato crop yield in drought scenarios. The identified parameters will be used as predictors for the change in yield due to droughts, by identifying how droughts influence the parameters. Three models have been chosen to study; Fortin et al. (2010), Ahmadi et al. (2014), and Dai, Huo, & Wang (2011). These models have been chosen, because they were deemed most relevant, due to the crops they study (potatoes) and their use of ANN’s. Furthermore, they have in common that they specify which parameters are implemented in the model. An analysis of these models will be made, which results in a list of parameters that will be used in this paper.

Two different categories can be associated with droughts. Precipitation and temperature could be assigned to factors resulting in drought formation, while salinity and a change in soil moisture correspond to the consequences of droughts (Peters, Torfs, Van Lanen & Bier, 2003; Xie et al., 2019). First, the role of precipitation and temperature on drought formation will be discussed. Afterwards, the role of droughts in changing soil salinity and soil moisture will be addressed. Finally, the consequences of a change in salinity and soil moisture will be linked to a subsequent change in the agricultural yield.

Before the influence of precipitation and temperature on drought formation can be discussed, it is important to define droughts. This paper defines droughts as a state where water demand exceeds water availability (Guerrero-Salazar & Yevjevich, 1975; Peters, Torfs, Van Lanen & Bier, 2003). There are different types of droughts, but this paper will only focus on meteorological, hydrological, and agricultural droughts. Meteorological and hydrological droughts are chosen because it is the precursor for all other droughts. Agricultural droughts are chosen because this paper focuses on the effect of droughts on agriculture (Guerrero-Salazar & Yevjevich, 1975; Peters, Torfs, Van Lanen & Bier, 2003).

Meteorological droughts occur when the evapotranspiration rates exceed precipitation rates (Peters, Torfs, Van Lanen & Bier, 2003; Beersma & Buishand, 2007). This means the water input is less than the water output. If this water deficit is large enough, it could result in a hydrological drought. A hydrological drought can be summarized as a

decrease of the groundwater table below a certain threshold (Peters, Torfs, Van Lanen & Bier, 2003). This threshold depends on the perspective from which you study the water levels. After all, it is rather subjective as to when a drought occurs (Guerrero-Salazar & Yevjevich, 1975; Peters, Torfs, Van Lanen & Bier, 2003). A population or sector that needs less water will not suffer a drought at certain water levels whereas other populations or sectors might already be undergoing the consequences of drought. This difference in experiencing droughts can be appointed to the differing water demands of each population or sector. In other words, the threshold for experiencing droughts depends greatly on the water demand of a particular population or sector.

Lastly, the agricultural drought occurs when there is insufficient precipitation and the groundwater table has sufficiently dropped that it limits the agricultural yield (Peters, Torfs, Van Lanen & Bier, 2003; Beersma & Buishand, 2007). If the precipitation rates are less than the evapotranspiration rates, a meteorological drought forms. If persistent, this could result in a hydrological and/or agricultural drought. It must be noted that it takes a considerable amount of time for an aquifer to recharge once it has been depleted, especially when combined with a meteorological drought (Peters, Torfs, Van Lanen & Bier, 2003; Wilhite, 2012; van Duinen, Filatova, Geurts & van der Veen, 2015). Both the precipitation as well as evapotranspiration rates are influenced by temperature.

Precipitation and temperature

Precipitation occurs when the air is supersaturated with water vapor (Christopherson & Birkeland, 2015). The air cannot function as a water storage to the excess water and precipitation forms. The amount of water that can be stored in the air depends directly on the temperature. The maximum specific humidity increases with a rising temperature. The maximum specific humidity is the amount of water that can be stored in the air at any given temperature (Christopherson & Birkeland, 2015). In other words, at higher temperatures, the air contains more water and can therefore result in higher precipitation rates.

Evapotranspiration and temperature

Potential evapotranspiration is the maximum amount of evapotranspiration that can occur under ideal circumstances (i.e. enough water, ideal temperatures for the vegetation, and sufficient sunlight) (Van den Hurk et al., 2007; Christopherson & Birkeland, 2015). Evapotranspiration can be divided into evaporation and transpiration. Evaporation (in mm) is the amount of water that is evaporated from a system (i.e. from bodies of water and soil). The evaporation rates increase with temperature (Beersma & Buishand, 2007; Christopherson & Birkeland, 2015).

Transpiration (in mm) is the amount of water that is transpired by the vegetation in the system. Transpiration occurs due to photosynthesis processes in the plant as carbon dioxide is transformed into oxygen. Water is a byproduct of this process and leaves the vegetation through the stomata (Kruijt, Witte, Jacobs & Kroon, 2008). Transpiration is inversely related to temperature. As the temperature increases and passes a certain threshold, plants will close their stomata to preserve their water storages. However, it is unclear to what extent this process contributes to reducing the transpiration rates, although it is likely that it is insignificant compared to the increased evaporation rates due to the higher temperatures (Kruijt, Witte, Jacobs & Kroon, 2008).

In short, higher temperatures lead to less precipitation and more evapotranspiration. This will result in an increased chance of drought formation. Drought formation is often combined with soil salinization. The evaporation rates are higher than water accumulation rates in the soil (Xie et al., 2019). As a result, water-soluble salts precipitate in the soil, causing the salinity of the soil to increase (Amezketa, 2006). This process is called soil salinization. Soil salinization is one of the most important processes that causes soils to degrade due to drought (Rengasamy, 2006). As a result of soil salinization and the associated lower water-uptake capacity, crops like the potato will have physiological and biochemical responses at the cellular and organism level (Anjum et al., 2011). Responses of the potato plant to drought and high salinity are primarily caused by water stress. Water stress is further increased due to a diminished soil moisture in absence of enough available water in the soil.

Water stress causes a higher rate of reactive oxygen species production (oxidative stress) and a low internal carbon dioxide concentration, due to stomatal closure, inhibiting photosynthesis in the potato plant (Vos & Groenwold, 1989; Zandalinas et al., 2018). Moreover, the environmental adversity affects osmotic potential (hyperosmotic stress), alters nutrient uptake and consequently creates ionic imbalances causing potential death of the potato plant (Eker et al., 2006; van der Zaag, 1973). Water stress is one of the most adverse factors of plant growth and productivity (Blom-Zandstra, 2014; Anjum et al., 2011). Since water is fundamental at every stage of plant growth, deficiency of water could be a severe threat to potato crop production if droughts prove themselves to be persistent (Turner, 1991; Anjum et al., 2011).

Summarizing, ANN models thus can provide knowledge on the different parameters relevant to potato crop growth, which can give a general prediction in the change in yield due to droughts and, subsequently, can affect the profitability of the agricultural sector. The next section will explain Porter’s five forces, a means of measuring changes in profitability. This section will also include the main factors that profitability in the potato sector is affected by.

Porter’s five forces

It is important to realize that farms are like any other company; they need to be economically viable in order to last ( _{​Bagheri, 2010)​. This paper therefore focuses on the drought-related} factors that influence the profitability of the farmers. The profitability of potato farmers in the Dutch delta region have been used as a proxy for all agriculture in the region to determine these factors, because of their sensitivity to droughts (Van Loon, 1981; Panigrahi, Panda & Raghuwanshi, 2001). Porter’s five forces consists of five forces that affect the profitability of potato farmers: rivalry among competitors, the threat of new entrants, the bargaining power of suppliers, the threat of substitutes, and the bargaining power of buyers (Porter, 2008; Volberda et al., 2011). These five forces will be discussed next, combined with insights from the current potato industry.

Rivalry among competitors

The first force, rivalry among competitors, entails that, in order to compete, farmers strife to reduce their costs, and therefore, prices, to become the cheapest and most favoured potato supplier among customers (Porter, 2008). This is the so called ‘cost leadership strategy’. Rivalry among competitors is, per example, influenced by circumstances that force them to focus on their own company, diminishing the competition, and therefore allowing the farmers to invest more time, money and/or effort in their own company (Volberda et al., 2011; Hitt,

Ireland & Hoskisson, 2016). This also works the other way around; when farmers can focus more on further improving their production, the competition will strengthen, which will increase the rivalry among competitors.

Another factor that influences the intensity of rivalry is industry growth. The potato industry growth is relatively fast compared to other EU countries. However, in the last 30 years, many firms were taken over, and many farms formed alliances in order to keep up with the increased competition and demand (Rademakers & McKnight, 1998). This has also influenced the rivalry among competitors (Volberda et al., 2011).

The last factor influencing rivalry is the presence of exit barriers. When exiting an industry gives a greater loss than proceeding production, companies tend to stay in the industry. Due to companies that do not exit the industry, small market shares stay in the hands of these companies (Volberda et al., 2011). This increases rivalry among competitors. If companies would exit the industry, rivalry would decrease (Volberda et al., 2011).

The threat of new entrants

The potato industry, as well as the agricultural sector, has one main entry barrier that makes sure the threat of new entrants is low, which is the so called ‘economies of scale’. In the potato industry, large-scale businesses are most present, which makes it possible to divide the fixed cost and thus make these relatively low (Rademakers & McKnight, 1998; Bechdol, Gray & Gloy, 2010). For a new company, it is difficult to divide these fixed costs, because the production will be lower in the beginning (Bechdol, Gray & Gloy, 2010). This results in higher costs, and therefore higher prices for a small company that enters the industry. This subsequently leads to a lower threat of new entrants.

However, economies of scale is not the only entry barrier for new entrants in an industry. Another barrier to enter the potato industry are capital requirements, in which deltas have an important role. _​Since deltas provide attractive features for industries, the available land has decreased in the last decade (Bucx et al., 2010). 55% of the area in the Delta region is used for agriculture _​,_​which contributes significantly to the Dutch economy (Bechdol, Gray & Gloy, 2010). However, since the area of the Delta region is densely populated, it will be difficult to find a large and good enough place for a new company (Bucx et al., 2010). Existing companies thus have an advantage, because they possess highly productive arable land, which is rich in fertile sediment and organic material for agricultural cultivation _​(Kuenzer & Renaud, 2012).

The bargaining power of suppliers

The third force, the bargaining power of suppliers, is influenced by demand and supply (Volberda et al., 2011). In the case of the potato industry, the bargaining power of suppliers increases when farmers are in need of a specific type of seed or fertilizer. This power is strengthened by the number of suppliers (Rademakers & McKnight, 1998; Bechdol, Gray & Gloy, 2010).

For commercial production, high-quality seed potatoes are preferred. These potatoes are free of plant diseases, have stable characteristics and enable high production. The suppliers of these high-quality seeds are a few agribusinesses, which have acquired smaller seed companies in order to maintain a larger market share (Rademakers & McKnight, 1998). This gives these large suppliers some power, because farmers want the high-quality seeds. However, there are hundreds of registered varieties of seed potatoes, which have a lower

quality. If prices become to high, farmers can switch to these smaller suppliers (Howard, 2009; _​Rademakers & McKnight, 1998 _​). This gives suppliers a limited amount of power, because potato farmers can switch to substitute seed potatoes (Howard, 2009; Bechdol, Gray & Gloy, 2010). Seed suppliers thus have a moderate power. Suppliers of for example fertilizer and equipment have a similar case as the seed suppliers. Potato farmers need the resources, but there are plenty of substitutes available (Howard, 2009; Bechdol, Gray & Gloy, 2010). However, the farmers need to cope with switching costs when acquiring their resources elsewhere. Therefore, the suppliers keep a certain amount of power.

The threat of substitutes

The threat of substitutes considers the economical backlash of a reduced yield, or the economic benefits of an increased yield. When the harvest of potatoes decreases, the prices go up and other crops become relatively cheaper (Bechdol, Gray & Gloy, 2010). The potato then faces the risk to be substituted for the cheaper crop. This also works the other way around. The threat of substitutes decreases when yield increases, because prices decrease, as discussed in the case of economies of scale.

The bargaining power of buyers

Potato farmers sell their products both to stores like Albert Heijn directly as well to processing companies to make products that have potato as an ingredient (Rademakers & McKnight, 1998; Nederlandse Akkerbouw Vakbond, 2017). The number of buyers present in the potato industry is large, and these buyers all buy a small part of the total output of the products (Rademakers & McKnight, 1998; Bechdol, Gray & Gloy, 2010). Both these factors decrease the bargaining power of buyers. In the case of a small number of buyers, the bargaining power would increase, because farmers would be dependent on these buyers.

Another factor that decreases the bargaining power of buyers, is the ability of farmers to meet the large demand of potatoes. Large-scale potato businesses are able to meet this demand because of their size, and since they have multiple buyers, they are not dependent on a few buyers. However, available substitutes and low switching costs increase the bargaining power of buyers, because these factors make it easier for a buyer to switch to another company (Volberda et al., 2011).

These threats have one thing in common: the severity of it lies in the loss of yield, and therefore profit, of the farmers (Van Duinen et al., Rademakers & McKnight, 2998; Bechdol, Gray & Gloy, 2010). It is therefore necessary to study how droughts impact the agricultural yield. The next section will elaborate on this.

Problem definition

In this section, the importance of this research will be discussed using the perspectives of different disciplines in an interdisciplinary way, providing insight into the complexity of the research problem. Additionally, an indication of the research gap on the effects of drought on agriculture in delta areas will be given.

Delta’s provide many favourable conditions for agriculture (Kuenzer & Renaud, 2012). The flat topography and the fresh and salt water availability by the sea and rivers attract farmers to grow their crops in those types of areas. However, the Dutch delta has a long history of water defense strategies in order to create and maintain those conditions (Meyer, 2009). The construction of dikes and dams and drainage technologies date back to Roman times and since then, the development of those technologies has drastically increased. This development is still in progress and requires knowledge of experts in different fields, such as civil- and hydraulic engineering and economics and is therefore considered a complex, layered system (Meyer & Nijhuis, 2013). Taking into account that developing and preserving an area such as the Dutch delta is so complex, the conclusion can be drawn that the alteration of conditions in this area is a complex problem as well, that needs to be analysed from different perspectives.

It is important to tackle a complex system from different perspectives, because of the intrinsic interconnectivity between certain factors within such a system. By only tackling one individual part of a problem, the other problems are neglected and will not be taken into account when implementing certain solutions (Menken & Keestra, 2016). The solution to one part of the problem can in turn negatively affect another factor in the system that is linked to the solved problem, thus worsening or creating a new problem within that system. If all factors in a complex system are taken into account when studying the problem area and analyzing possible solutions, the chances of implementing a solution that will destabilize the system will be reduced (Menken & Keestra, 2016). Interdisciplinary studies are therefore a necessity.

Performing an interdisciplinary research on a topic is one way in which a problem will be studied from multiple perspectives, and thus acknowledging most, if not all, factors in a system (Menken & Keestra, 2016). It should be noted that a complex system, as the name implies, is oftentimes too complex to understand completely. There will therefore always remain some uncertainty as to how the system functions and how an alteration in one factor will affect the other factors, and thus the functioning of the system as a whole. Interdisciplinary research approaches a complex system from multiple different disciplines and integrates pre-existing bodies of knowledge within these different disciplines and combines it to give a more accurate and complete explanation of the system (Menken & Keestra, 2016).

The aim of this research is to connect natural- and social sciences in order to research what the consequences of recurring droughts will be for the agricultural sector in the future. By integrating pre-existent parameters used by ANN to predict potato growth as a guideline for the earth sciences, the focus of this research was on the influence of droughts on the hydrology and soil quality of the delta. The hydrology focussed on a change on the precipitation, evapotranspiration and groundwater levels. These factors are either directly or indirectly based on the parameters of the ANN. Soil quality will be studied by addressing the

relationship between droughts and soil salinization. Subsequently, business administration will use those results to determine in what way the profitability of potato farmers will change in the future. In doing so, the aim is to eventually bridge the knowledge gap between the physical changes of the delta and the changes in profitability of the farmers that rely on those physical conditions. In what way the individual disciplines are integrated and how they all contribute to answering the research question of this interdisciplinary research, will be discussed in the next section.

Interdisciplinary approach

Three different disciplines have been used in order to answer the research question: Artificial Intelligence, Earth Science, and Business Administration. The question was initially tackled in a singular way, where one discipline provided the information needed by another discipline in order to continue onwards to answering the research question. However, this is not a truly interdisciplinary approach. In order to integrate the information in an interdisciplinary manner, it was important to define how each discipline contributed to the research as a whole. This resulted in the integrated theoretical framework. A visualization of the framework can be found in figure 2.

Figure 2. Visualization of the integration of the three disciplines in order to answer how droughts influence the agricultural sector in the Dutch delta region in the future.

Incorporating artificial intelligence in the interdisciplinary framework proved challenging for several reasons. The ANN models predicting potato growth were need publicly available. This research could therefore not directly use the models and therefore had to resort to literature reviews only. However, by focussing not on the limits of the models, but on the information provided by these models that was available, it became clear that artificial intelligence became the cornerstone of the research as the parameters used in the models were also the most important factors that were used by the earth science disciplines.

The earth science discipline was divided into two different perspectives, namely Hydrology and Soils. The earth sciences focused more on the relationship between drought and agricultural process. The hydrological perspectives focussed more on the probability of the occurence of drought formation in the Netherlands by analyzing the parameters that would influence drought formation as provided by the artificial intelligence models. The soil perspective focussed on the qualitative changes in the soil that would affect the agricultural efficiency of the region. These changes were also based largely on parameters indicated by the models.

The parameters, as singled out by artificial intelligence, were compared by the qualitative process that influence the hydrology and soil quality of a region due to drought related circumstances. The model parameters matched the previously defined processes and it became clear that a deeper understanding of these processes along with the usage of the models, could provide a helpful tool in predicting the possible consequences of drought on the agricultural yield. Thus, instead of separating the natural sciences from the artificial models, combining the two areas was deemed more useful. This combination explained how a change in certain hydrological, biological or soil scientific processes would influence the agricultural sector. The earth sciences provided a qualitative explanation and the models formed a quantitative confirmation. Even if the models themselves have not been used in this paper, previous studies have confirmed that a certain change in one of the parameters/processes had a certain agricultural impact.

Business administration plays a crucial part in answering the research question. The agricultural sector can only be sustainable if they are profitable. Business administration is therefore a necessary discipline to incorporate into the framework. By using Porter’s five forces, business administration was able to use the information provided by both earth sciences and artificial intelligence to analyse whether or not a drought related change would result in a loss of profitability for the agricultural sector.

As the profitability of the agricultural sector in the Dutch delta region is key to answering the research question, the artificial models were chosen based on their focus on the agricultural yield. In other words, the models that form the cornerstone of the research were chosen specifically to provide the necessary information for the business administration discipline.

In summary, the chosen processes to qualitatively describe the influence of droughts on the agricultural sector were identified by using the most important parameters used in models to predict agricultural yield. The focus of the models on yield was chosen specifically because the research question can be answered by studying a change in the profitability of the agricultural sector in the Dutch delta region.

Method

In this section, the methods for data gathering and data analysis will be discussed. For data gathering, this research started with quantitative research, in order to select the most suitable parameters. Hereafter, a literature study was used. This research thus used a mixed method approach: both quantitative and qualitative research methods were used. This mixed method has been chosen because quantitative research explains the physical processes that lead us to understand drought and why it changes certain variables the way that it does. Meanwhile, qualitative research explains the connection between physical processes and the changes in strategies for companies, and thus the influence of droughts on the potato crop industry in the Dutch delta.

First of all, the selection criteria for the parameters will be discussed. Next, it will be explained how data on the selected parameters was gathered and why this method was chosen. In the last section, the analysis of the gathered data will be thoroughly explained and discussed.

Selection of the parameters

The primary research method for this study is a literature review and will thus mostly focus on analysing secondary data in the form of research papers. This is the quantitative research method used for this paper. To find what parameters are most important in predicting potato crop yield, a comparative literary analysis has been conducted. In this part, three models that predict crop yield through applying artificial neural networks have been compared on several fronts. First of all, it is important that the change in crop yield is predicted as closely as possible. Hence, the models were ranked based on their RMSE (root mean square error). The lower the RMSE is, the lower the difference between the predicted crop yield and the actual measured crop yield. A model with a low RMSE thus has a high chance of accurately predicting new crop yield and is thus favorable over a model with a higher RMSE (Dai et al., 2014, Fortin et al., 2010).

Furthermore, considering the focus of this research on droughts, it was paramount to choose models that incorporated parameters that were affected by droughts. After all, if the parameters accurately predict crop yield, but foregoes certain parameters, such as precipitation, that would be affected by droughts then it is impossible to analyze the influence of droughts on the potato yield. As the parameters of the models are important, a list of the used parameters in each model was composed. The overall best combination of RMSE and drought-affected parameters was chosen to select what parameters this research was going to focus on (table 1).

Selection of relevant literature

This research has focused on literary research and results of field work conducted by other studies. For each of the chosen parameters (salinity, soil moisture, precipitation and temperature) that were considered to be most relevant with respect to droughts, several possible relevant theories were acquired (appendix 1). Then for each theory, it was established that at least two peer-reviewed papers had to substantiate its relevance for this particular case study. If this were the case, databases such as _{​Web of Science ​}and_​Google

Scholar_​, as well as the library of the University of Amsterdam were systematically searched. The aim of this part was to find papers providing an in-depth explanation as to why these particular parameters are important for changes in potato crop yield and how they relate to changes in drought. This study was not limited to specific disciplines, thus literature was sought in different disciplines, such as hydrology, soil science and biology.

Because this research also highlights the industrial changes that might occur due to a changing potato yield, papers on profitability of the potato and the agricultural sector in the Dutch delta region were used. This was completed with a qualitative research approach in which the industry was analysed according to the Industrial Organization model (I/O model), specifically focusing on Porter’s five forces. The needed data for this part of the research included the current state of profitability in the Delta (based on Porter’s five forces), the characteristics that make this area attractive for the industry, the relationship between yield and each of the forces and the possible changes that might occur on the yield due to droughts.

The application of Porter’s five forces-framework is part of a qualitative analysis. This consisted of three steps. Using the I/O model, it is of importance to specify the industry that is analysed, because boundaries between industries are often overlapping (Porter, 2008; Volberda et al., 2011). Therefore, for the first part of the Porter’s five forces analysis, a clear definition of the industry was necessary. This research decided to focus on the potato industry, specifically potato farmers, because potatoes are sensitive to droughts, and farmers are the stakeholders in the industry that are expected to be influenced by droughts the most _{​(Van Duinen et al., 2015; Van Loon, 1981; Panigrahi, Panda & Raghuwanshi,} 2001_​).

The I/O model addresses the importance of the external environment before making a strategy (Volberda et al., 2011). How the external environment of the potato industry was analysed for this paper, which contains the last two steps of the Porter’s five forces analysis, will be discussed in the analysis of the results.

Analysis of the results

The primary goals of this research were to identify what type of droughts the Dutch delta region might suffer from in the near future, how and why this affects potato crop yield, and how the profitability of the potato industry in this case study might react to this. In order to identify what type of droughts the Dutch delta region might suffer from, climatic data of the IPCC (2014) was acquired and used to predict the type of droughts that are most likely to occur in the Netherlands in the future. Based on these findings, and the gathered data for each of the theories, an analysis could be conducted in three steps.

First of all, an overview on drought-related changes to hydrology, nutrients and salinity was established. This overview included the expected changes in the soil characteristics and the groundwater levels of a region suffering from drought.

These changes could then be implemented in step two of the Porter’s five forces analysis. For each of the changes, it was reasoned which of the five forces they could influence and how. This was done by extensive research of historical changes that have led to stronger or weaker forces, and comparing them to events expected to happen should drought formation occur in the Netherlands. If it was determined that a force would increase, it followed that the profitability would decline.

Subsequently, each of the natural science theories was thoroughly discussed to determine whether the processes were relevant for this particular case study.

Lastly, the different parameters were linked together. This started with creating an interdisciplinary visualization that explains the interconnectivity of the parameters (figure 2). Next, the relevant theories were linked together to find an underlying system of feedback loops that could explain the interconnectivity of the different parameters. This could then be used to answer the research question in an interdisciplinary way.

Analysis

Since it is expected that droughts will occur more often, this research has analyzed the scenario of recurring droughts in the Netherlands. In order to understand the effects of recurring droughts, this research first analyzed the three main factors that contribute to drought formation; precipitation, temperature and evapotranspiration. These factors are influenced by climate change. Secondly, the effects of recurring droughts were analyzed regarding soils and the profitability of the potato industry. The potato industry is used as a representative of the agricultural sector, since potatoes are agricultural products which are relatively sensitive to water stress.

First of all, three artificial neural networks were compared in terms of parameters and predictability power (table 1). A low predictability power was desired, as well as parameters that are likely to change due in value due to droughts. Fortin et al. (2010) constructed a model with precipitation, solar radiation and temperature as parameters for potato crop growth. This generated a RMSE of 0.364. Ahmadi et al. (2014) used salinity, soil moisture deficit and soil structure as parameters, and generated a significantly better RMSE of 0.101. Lastly, Dai et al. (2014) built a model with only salinity and soil moisture as its components, and had the best RMSE of all three, namely 0.1. Since Ahmadi et al. (2014) and Dai et al. (2014) are similar in RMSE, as can be seen in table 1, both can be considered as suitable for this research. The soil structure that the former model included primarily indicated the ratio of clay, sand and silt particles. Although changes of soil structure can influence water transport in the soil (Li, Chang & Salifu, 2014), changes in soil structure is not the most important consequence of drought for soil quality (Schimel, 2018). Therefore, this parameter was excluded from this research. However, other soil characteristics have a high influence on soil quality and are thus relevant parameters to analyse in relation to potato yield. Specifically changes in water levels (because of the importance of soil moisture) and salinity have been chosen to research.

Table 1. Overview of used parameters in the artificial neural networks of model 1 (Fortin et al., 2010), model 2 (Ahmadi et al., 2014) and model 3 (Dai et al., 2014). A check means that the selected parameter is indeed included in the model that yielded the best RMSE. RMSE is the root mean squared error, and is a way of measuring how much the predicted yield deviates from the actual measured yield. A low RMSE is desirable, as it means the model has a high predictability power.

According to the IPCC (2014), changes in precipitation are season dependent. The mean annual precipitation in the Netherlands will know a 3.5-7% winter increase, and a decrease of equal size in the summer, resulting in dry summers. The IPCC models predict with a high certainty that the global temperature will rise _{​1.5-2°C above the pre-industrial} period (IPCC, 2014; Van den Hurk _{​et al., 2007; table 2). However, there is an uncertainty in} how this translates directly to the Netherlands. It is expected that regions with considerable oceanic influences will have a temperature increase of 1.5°C, whereas 2°C is predicted to occur inland. Therefore, the Dutch delta region will most likely end on the side of a 1.5°C temperature increase. This would result in the reduced replenishment of groundwater in the aquifers due to higher evaporation and lower transpiration rates (Van den Hurk et al., 2007). This change in the water balance is an important aspect of soil salinization (Xie et al., 2019). When a region experiences a period of drought, the potential evaporation will be higher than the rainfall rate and upward seepage of groundwater (Amezketa, 2006). This causes water to evaporate, leaving water-soluble salts behind in the top layer of the soil (Rengasamy, 2006).

Table 2. The changes in potential evaporation as a percentage and temperature in the Netherlands based on the IPCC climate models (control and intermediate scenario) and the KNMI regional model (dry scenario) (Kruijt, Witte, Jacobs & Kroon, 2008).

Climate scenario Control scenario Intermediate scenario Dry scenario

Soil salinization is not only caused by an increase in evaporation. As groundwater tables in the Netherlands are quite shallow, there is a strong interaction between groundwater and surface water (Kroes & Supit, 2011). This combined with sea level rise which causes groundwater supplies to increase in salinity, will change soil properties (Kroes & Supit, 2011; Hu & Schmidhalter, 2005). Various nutrients, such as Organic Carbon and Nitrogen, decrease significantly when soil salinity is increasing. However, not all nutrients available in the soil are decreasing. _{​Phosphate, Sodium, Calcium and Potassium are shown} to increase with increasing soil salinity (Pan et al., 2013). This will lead to reduced plant growth, due to toxicity of the levels of those nutrients available in the soil (Hu & Schmidhalter, 2005; Machado & Serralheiro, 2017). Using Porter’s five forces, four possible threats have been identified that could negatively affect the profitability of the potato farmers in the face of droughts, and thus a decrease in the fertility of the soil: rivalry among competitors, the threat of new entrants, the threat of substitutes, and the bargaining power of suppliers (Bechdol, Gray & Gloy, 2010). The bargaining power of buyers is expected to remain virtually unchanged and therefore does not contribute to a change in the profitability of the agricultural sector (Bechdol, Gray & Gloy, 2010; Rademakers & McKnight, 1998). If a force increases, the profitability of the agricultural sector decreases. Conversationally, a decrease of a force will lead to an increased profitability (Porter, 2008; Volberda et al., 2011).

Since the high fertility of soils in the delta region decreases, which provide favorable features for potato growth, potato yield will decrease. This leads to a decrease in barriers to enter the potato industry, such as the barrier economies of scale, because the existing companies in the Dutch delta region lose their advantage over new companies. This allows these new companies to settle in the area and become competitors, concurring over a shared market or space (Porter, 2008; Volberda et al., 2011; Hitt, Ireland & Hoskisson, 2016). This will lead to an increase in the force threat of new entrants. Furthermore, the bargaining power of suppliers will increase, because when farmers are in need of a specific type of seed or fertilizer in order to offset the negative consequences of drought on their yield, suppliers of these seeds and fertilizer can increase their prices (Van Duinen et al., 2015).

Besides nutrient supply, which is an important factor in determining soil fertility, other soil properties are affected by salinity as well. Enzyme activities, which are important for organic matter decomposition and nutrient cycling in the soil, are also sensitive to changes in soil salinity (Pan et al., 2013). _{​When nutrient cycling and organic matter decomposition rates} change, this can cause major implications for agriculture (Manlay, Feller & Swift, 2007). It does not only affect the soil in the present, but it influences the ecosystem sustainability of farmland as well (Omar et al., 1993). Especially soil organic carbon and Nitrogen cycles, who are influencing the litter production in the soil, are important factors in soil fertility on the long term and are closely related to salinity levels in the soil (Pan et al., 2012; Omar et al., 1993). This thus also influences Porter’s five forces through the decline in soil fertility, because a decrease in soil fertility influences the threat of new entrants and the bargaining

Potential Evaporation

power of suppliers, as discussed above (Bechdol, Gray & Gloy, 2010).

Increasing salinity does not only affect the fertility of the soil. The salinity threshold of most vegetable crops is quite low (Machado & Serralheiro, 2017). High concentrations of salt in the soils cause crops to reduce in productivity and quality (Van Duinen et al., 2015). As a result of the decreasing crop growth, the agricultural sector will experience economic losses. These economic losses will lead to higher prices, making substitutes more attractive. Therefore, the threat of substitutes will most likely also increase, because this threat considers the economical backlash of a reduced yield (Porter, 2008). When the harvest of potatoes has significantly shrunk, the prices go up and other crops become relatively cheaper. The potato then faces the risk to be substituted for the cheaper crop (Rademakers & McKnight, 1998; Bechdol, Gray & Gloy, 2010).

Most companies in the potato industry have a cost leadership strategy, which will be harder to maintain because of the higher prices (Bechdol, Gray & Gloy, 2010). Farmers will have to turn away from a cost leadership strategy if circumstances force them to focus on their own company, diminishing the competition, and therefore allowing the farmers to invest more time, money and/or effort into their own company (Volberda et al., 2011; Hitt _{​, Ireland &} Hoskisson, 2016_{​). Rivalry among competitors will therefore decrease. It is possible that} companies will seek opportunities in differentiating their product by for example branding or making specialized products. However, this is hard to achieve in the agricultural sector, because branded or specialized products are quickly copied by competitors (Bechdol, Gray & Gloy, 2010; Volberda_{​ ​}et al._​,​ 2011).

From the four forces that are expected to significantly change, only the rivalry among competitors shows the possibility of a positive effect on the profitability of the potato farmer. As droughts decrease the water availability, farmers will need to turn away from a cost leadership strategy and focus more on the sustainability of their farm in order to adapt to a less than optimal situation. The rivalry between competing potato farmers will decrease, because of this shift in focus. A decrease in the force results in a positive effect on the profitability of the farmers. However, due to a decreased water availability, their yield decreases as well, which in turn has a negative effect on the profitability. This is reflected in the three forces that are expected to increase; the threat of substitutes, the threat of new entrants, and the bargaining power of suppliers. It is uncertain which effect will turn out to be more significant. However, according to Porter (2008) each force has the same effect on the profitability of an industry. The _{​refore, the profitability of the potato industry is expected to} decrease.

Conclusion

According to the climate models, summers are expected to increase in temperature (1.5-2_{​°C) ​and decrease in precipitation (3.5-7%). Higher temperatures and reduced} precipitation, if severe enough, could result in meteorological and hydrological drought formation. Climate models therefore predict that climate change will force atmospheric conditions to lean towards drought forming scenarios. If the severity of these changes is intense enough to immediately result in an agricultural drought, is impossible to say. However, if the summer droughts persist with sufficient frequency, the delta region might not receive enough water during winter to recover from the summer droughts. After all, aquifers take a long time to recover from droughts. If this is the case, then the water table will be lowered each year until the threshold of the agricultural drought in the Dutch delta region has been reached.

Higher temperatures and less precipitation in the summers will result in higher evaporation than precipitation rates, resulting in soil salinization. The soil quality is expected to decrease based on a combination of an increase in salt water intrusion due to sea level rise, a changed nutrient balance of the soil, and a change in enzyme activity. The consequences of soil salinization apply to a wide variety of soils, however, specific analyses of changing soil properties due to salinization in the Dutch delta region that this report is focussing on, are lacking. For a greater understanding of the consequences of droughts on the soil quality, it is of great importance to research what soil properties will change and how. Future research on drought consequences in the Dutch delta will therefore benefit greatly from gathering an in-depth database on the different soils in the delta region and study how these different soil types react to soil salinization and a lack of water,and whether or not the soil quality is affected.

A lowered soil quality, after all, results in a reduced agricultural yield. Subsequently, the profitability of the agricultural sector will be affected due to these changes as four of Porter’s five forces are expected to change. Three of these forces will increase, namely the bargaining power of suppliers, the threat of new entrants and the threat of substitutes. Only one force could possibly decrease; the rivalry among competitors. The bargaining power of buyers is expected to not significantly change. As each of these forces have the same weight in determining the profitability, it is most likely that the profitability of the agricultural sector in the Dutch delta region will decrease.

Thus, droughts result in a reduced water availability and soil quality. This in turn influences the profitability of the agricultural sector. In the Netherlands, summer droughts are expected to increase in frequency and severity. Climate models functioned as a guideline throughout this research to predict whether or not it is probable to assume that drought formation will occur in the future in the Netherlands.

Discussion and recommendations

The used climate models function as a guideline to predict the general direction of future drought formation in the Netherlands. However, the periodicity and exact severity remain uncertain due to the inherent uncertainties of climate models. In order to increase the accuracy of the climate predictions, better climate models are needed. Furthermore, there is a need for a better understanding of a change in the atmospheric processes that affect the Netherlands in particular. However, as these models are yet to be refined, this paper has settled on giving only a general prediction.

Additionally, for more accurate predictions concerning a change in the hydrological cycle in the Dutch delta region, a hydrological model of the delta region -along with entire catchments of all important contributing rivers- is necessary. No such model currently exists that focuses on droughts in particular. Such a model could also further predict the rate of salt-water intrusion from the sea and soil salinization. The absence of such a specialized model limits this paper to broad statements concerning changes in the hydrological cycle as well as changes in soil salinization in the delta region.

As for the ANN models, it should be noted that the experiments did not take place in the Netherlands and thus may not acquire the same predictability power here. It might be the case that the types of potatoes that grow in the other study areas, have entirely other characteristics than the ones in the Netherlands. However unlikely, this might mean that parameters that are useful at predicting crop growth in the research areas of the models studies might not return the same result in the Netherlands. Possible adaptations, mutations and alterations in gene-structures, man-made or not, have not been considered. Thus, there is no way to exclude the possibility that longer periods of drought might lead to (artificial) adaptations in potatoes, which might decrease the chance of yield-decline.

The uncertainties of the climate and ANN models and the lack of a hydrological model specialized in droughts limit the accuracy and analysing power of Porter’s five forces. Each of the five forces have the same weight. It is therefore important to be able to precisely quantify the changes in each force in order to quantify the change in profitability. However, since the lack of models makes it impossible to quantify these changes, this paper will only discuss these changes qualitatively.

It must be noted that the research question is answered using the results from the profitability analysis conducted by Porter’s five force. Considering this analysis only considers a general analysis of the predicted direction of these changes, the research question can only be answered qualitatively and is therefore limited in its usability. However, by providing a general overview of the possible consequences of droughts on the agricultural sector in the Dutch delta region in the future, this paper can be used as a further encouragement to invest in suitable countermeasures in order to protect the agricultural sector from future droughts.

Another limiting factor is the usage of potatoes as a proxy for all agricultural activities. In order to quantitatively and accurately measure a change in agricultural yield and profitability, it is insufficient to only study one crop. This research has chosen potatoes as a proxy, because of its sensitivity to soil salinization. However, not all crops have the same threshold for soil salinity. By picking the most sensitive crop, quantitative research would describe the worst case scenario. However, since this paper has used no quantitative data,

and the qualitative relationships between climate change and soil quality exist are described broadly, the qualitative effect of droughts on agriculture, as described in this paper, holds up.

So, droughts are expected to increase in the Netherlands based on climate models. However, the severity and frequency of these droughts remain uncertain and several aspects of this research require supplementary research. However, based on this research the conclusion can be drawn that drought formation in the Dutch delta region would result in a reduced soil quality and agricultural yield, negatively affecting the profitability of the agricultural sector in the Dutch delta region.

References

Ahmadi, S.H., Sepaskhah, A.R., Andersen, M.N., Plauborg, F., Jensen, C.R., Hansen, S. (2014). Modeling root length density of field grown potatoes under different irrigation strategies and soil textures using artificial neural networks. _{​Field crops research, 162,} 99-107

Alvarez, R. (2009). Predicting average regional yield and production of wheat in the Argentine Pampas by an artificial neural network approach. _{​European journal of agronomy,} 30_​(2), 70-77.

Amezketa, E. (2006). An integrated methodology for assessing soil salinization, a pre-condition for land desertification. _{​Journal of Arid Environments, 67(1), ​}594-606

Bagheri, A. (2010). Potato farmers’ perceptions of sustainable agriculture: the case of Ardabil province of Iran. _{​Procedia-Social and Behavioral Sciences​}, _​5​, 1977-1981.

Basheer, I. A. & Hajmeer, M. (2000). Artificial neural networks: Fundamentals, computing, design and application. _{​Journal of Microbiological Methods, 43(1), ​}3-31

Bechdol, E., Gray, A., & Gloy, B. (2010). Forces affecting change in crop production agriculture. _{​Choices​}, _​25​(4), 12-17.

Beersma, J. J., & Buishand, T. A. (2007). Drought in the Netherlands–regional frequency analysis versus time series simulation. _{​Journal of hydrology​}, _​347​(3-4), 332-346

Bucx, T., Marchand, M., Makaske, B., & van de Guchte, C. (2010). Comparative assessment of the vulnerability and resilience of 10 deltas: work document (No. 1). _{​Deltares​}.

Christopherson, W. R., & Birkeland, G. (2015). _{​Geosystems: An Introduction to Physical} Geography:[Julien Moreau: editing the questions of each chapter of the book] _​. Pearson Education Limited.

Dai, X., Huo, Z., & Wang, H. (2011). Simulation for response of crop yield to soil moisture and salinity with artificial neural network. _{​Field crops research, 121​}(3), 441-449

Fortin, J.G., Anctil, F., Parent, L., & Bolinder, M.A. (2010). A neural network experiment on the site-specific simulation of potato tuber growth in Eastern Canada. _{​Computers and} Electronics in Agriculture, 73_​(2), 126-132.

Guerrero-Salazar, P. L. A., & Yevjevich, V. M. (1975). Analysis of drought characteristics by the theory of runs. _{​Hydrology papers (Colorado State University​}); no. 80.

Hitt, M. A., Ireland, R. D., & Hoskisson, R. E. (2016). Strategic management: Concepts and cases: Competitiveness and globalization. _{​Cengage Learning.}

Hu, Y. & Schmidhalter, U. (2005). Drought and salinity: A comparison of their effects on mineral nutrition of plants. _{​Journal of Plant Nutrition and Soil Science, 168(4), ​}541-549. IPCC. (2014) _{​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

Kaul, M., Hill, R.L., & Walthall, C. (2005). Artificial neural networks for corn and soybean yield prediction. _{​Agricultural systems, 85​}(1), 1-18

KNMI (n.d.). _{​Droogte.​} Retrieved November 26, 2019, from

https://www.knmi.nl/kennis-en-datacentrum/uitleg/droogte

Kroes, J. G. & Supit, I. (2011). Impact analysis of drought, water excess and salinity on grass production in The Netherlands using historical and future climate data. _{​Agriculture,} Ecosystems and Environment, 144(1), _​370-381.

Kruijt, B., Witte, J. P. M., Jacobs, C. M., & Kroon, T. (2008). Effects of rising atmospheric CO2 on evapotranspiration and soil moisture: A practical approach for the Netherlands. Journal of Hydrology_​, _​349​(3-4), 257-267.

Kuenzer, C., & Renaud, F. G. (2012). Climate and environmental change in river deltas globally: expected impacts, resilience, and adaptation. In The Mekong Delta System (pp. 7-46). _{​Springer, Dordrecht.}

Li, X., Chang, S. X. & Salifu, F. (2014). Soil texture and layering effects on water and salt dynamics in the presence of a water table: a review. _{​Environmental Reviews, 22(1), ​}41-50 Lintsen, H. (2002). Two Centuries of Central Water Management in the Netherlands. Technology and Culture, 43(3), _​549-568

Machado, R. M. A. & Serralheiro, R. P. (2017). Soil Salinity: Effect on Vegetable Crop Growth. Management Practices to Prevent and Mitigate Soil Salinization. _{​Horticulture, 3(2),} 30. doi:10.3390/horticulturae3020030

Mahajan, S., Das, A., & Sardana, H.K. (2015). Image acquisition techniques for assessment of legume quality. _{​Trends in Food Science & Technology, 42​}(2), 116-133

Manlay, R. J., Feller, C. & Swift, M. J. (2007). Historical evolution of soil organic matter concepts and their relationships with the fertility and sustainability of cropping systems. Agriculture, Ecosystems & Environment, 199(1), _​217-233.

Menken, S., & Keestra, M. (2016). _{​An introduction to interdisciplinary research: Theory and} practice_​. Amsterdam University Press.

Meyer, H. (2009). Reinventing the Dutch Delta: Complexity and Conflicts. _{​Built Environment,} 35(4), _​432-451

Meyer, H. & Nijhuis, S. (2013). Delta urbanism: planning and designing in urbanized deltas - comparing te Dutch delta with the Mississippi River delta. _{​Journal of Urbanism, 6(2),} 160-191

Omar, S. A., Abdel-Sater, M. A., Khalil, A. M. & Abd-Alla, M. H. (1993). Growth and Enzyme Activities of Fungi and Bacteria in Soil Salinized with Sodium Chloride. _{​Folia Microbiologica,} 39(1), _​23-28.

Panigrahi, B., Panda, S., & Raghuwanshi, N. (2001). Potato water use and yield under furrow irrigation. _{​Irrigation Science​}, 20(4), 155-163.

Pan, C., Liu, C., Zhao, H. & Wang, Y. (2013). Changes of soil physico-chemical properties and enzyme activities in relation to grassland salinization. _{​European Journal of Soil Biology,} 55(1), _​13-19.

Peters, E., Torfs, P. J. J. F., Van Lanen, H. A. J., & Bier, G. (2003). Propagation of drought through groundwater—a new approach using linear reservoir theory. Hydrological processes, 17_​(15), 3023-3040.

Porter, M. E. (2008). The five competitive forces that shape strategy. Harvard business review, 86(1), 25-40.

Rengasamy, P., (2006). World salinization with emphasis on Australia. _{​Journal of} Experimental Botany, 57(5), _​1017-1023

Schimel, J. P. (2018). Life in Dry Soils: Effects of Drought on Soil Microbial Communities and Processes. _{​Annual Review of Ecology, Evolution, and Systematics, 49(1), ​}409-432

TKI Dinalog (2019). _{​Delta Regio publiceert economische koepelvisie: focus op Biobased,} Logistiek en Maintenance_​. Retrieved on 10 september 2019, from https://www.dinalog.nl/delta_regio_publiceert_economische_koepelvisie__focus_op_biobase d__logistiek_en_maintenance-1/

Van Duinen, R., Filatova, T., Geurts, P & Van der Veen, A. (2015). Coping with drought risk: empirical analysis of farmers’ drought adaptation in the south-west Netherlands. _​Regional Environmental Change, 15(1), _​1081-1093

Van den Hurk, B., Tank, A. K., Lenderink, G., van Ulden, A., van Oldenborgh, G. J., Katsman, C., ... & Komen, G. (2007). New climate change scenarios for the Netherlands. Water science and technology_​, _​56​(4), 27-33.

Van Loon, C. D. (1981). The effect of water stress on potato growth, development, and yield. American Potato Journal, 58(1), 51-69.

Van der Zaag, D. E. (1973). _{​Potatoes and their cultivation in the Netherlands ​}._{​The Hague,} Netherlands: Dutch Information Center for Potato

Volberda, H.W., Morgan, R.E., Reinmoeller, P., Hitt, M.A., Ireland, R.D., Hoskisson, R.E. (2011). Strategic Management Concepts and Cases Competitiveness and Globalization. Cengage Learning Emea

Vos, J., & Groenwold, J. (1989). Characteristics of photosynthesis and conductance of potato canopies and the effects of cultivars and transient drought. _{​Field Crops Research​}, 20_​(4), 237-250.

Wilhite, D. A. (Ed.). (2012). Drought assessment, management, and planning: theory and case studies: theory and case studies (Vol. 2)._{​ Springer Science & Business Media.}

Xie, X., Pu, L., Zhu, M., Xu, Y. & Wang, X. (2019). Linkage between soil salinization indicators and physicochemical properties in a long-term intensive agricultural coastal

reclamation area, Eastern China. _​Journal of Soil and Sediments,

https://doi.org/10.1007/s11368-019-02333-3

Zandalinas, S. I., Mittler, R., Balfagón, D., Arbona, V., & Gómez _{-Cadenas, A. (2018). Plant} adaptations to the combination of drought and high temperatures. _{​Physiologia plantarum​}, 162_​(1), 2-12.

Appendix

Appendix 1. Data management table

Table 3. Data Management Table.

Discipline and theory

Concepts Assumptions/Method Insight in problem

Earth Science_​- Hydrology

Droughts occur when water demand exceeds water supply in a region over a prolonged period of time. Droughts Water demand Water supply Varies from region to region

When water demand exceeds water supply, there is a shortage of water.

There are four types of drought: meteorological, hydrological, agricultural and socio-economic droughts. Higher temperatures favor drought formation Temperature Evapotranspiration Streamflow Runoff Replenishment water storages Hydrological drought Temperature is directly proportional to evapotranspiration, inversely proportional to streamflow and runoff. Higher temperatures will therefore favor drought formation

Higher temperatures result in an increase in the evaporation and transpiration rates of water. This means less water is available as soil moisture, runoff or streamflow. This reduces the

replenishment of water storages, eventually -if prolonged- resulting in a hydrological drought.

Less precipitation in summer favors drought formation Precipitation Replenishment water storages Meteorological drought A precipitation deficit means there is less water input into the system and there is less water available to replenish the water storages.

A meteorological drought could, over a longer time series, result in the formation of a hydrological drought.