Models of intentional behavior suggest that the performance of environmental behavior is the result of the rational consideration of cognitive factors, like attitudes, beliefs, and norms, and external factors, such as cost and time. While numerous rational choice models have been put forward (e.g., Rogers, 1975; Keshavarz and Karami, 2016), the most common used to research environmental behavior (Klöckner, 2013) are the Theory of Reasoned Action (TRA) (Ajzen and Fishbein, 1980; Fishbein and Ajzen, 1981), and its extension, the TPB (Ajzen, 1985, 1991). These models suggest that most human behavior is goal-directed, and therefore intentional. The TPB proposes that intentions to engage in a behavior are influenced by attitudes, social norms, and perceived behavioral control (PBC). Ajzen (1991) argues that PBC exerts influence on behavior by way of intentions, and directly. PBC includes factors related to self-efficacy, or external practical factors (Armitage and Conner, 2001).

Of the studies included in our review 39 examined the relationship between TPB, TRA, or individual factors included in these models, and the performance of environmental behavior. Numerous others combined these models with other variables, including moral norms and habits, as reported below in relation to integrated approaches. Models of intentional behavior were found to explain examples of behavior from across the six categories of environmental behavior examined in our review.

Habits refer to the “automatic performance of behavioral patterns triggered by context cues” Klöckner (2013, p. 1030). Many habits serve a function and therefore involve intentionality (Verplanken and Aarts, 1999). Three conditions determine habit formation: the degree of involvement, perceived complexity, and the degree of constraints (Jackson, 2005). Behaviors that are easier to perform, are performed more regularly, and face fewer constraints are more likely to become habitual (Bratt et al., 2015). For example, curbside recycling (e.g., Bratt, 1999; Osbaldiston and Schott, 2012) and water saving (Martínez-Espiñeira et al., 2014) are generally considered easy to perform and face few, if any barriers. In both cases, past behavior has been found to predict behavioral intentions (Carrus et al., 2008; Martínez-Espiñeira et al., 2014). Further, the determinants of habits can vary depending on the strength of the habit. Strong habits are triggered predominately by context rather than intentions and goals, while weaker habits are more directly influenced by goals (Neal et al., 2012). Habit strength is determined by the frequency and consistency of past performance (Ouellette and Wood, 1998; Aarts and Dijksterhuis, 2000). While behaviors become automatic over time, habitual behavior often begins with intentions (Verplanken and Aarts, 1999). Over the time the focus of the behavior shifts from performing specific activities, to the goal of performance, as the sequence of activities becomes routine (Thøgersen, 1999).

Of the studies included in our review six considered the influence of past behavior on the performance of pro-environmental behavior. Interestingly, we found that the role of habits has been considered for each type of environmental behavior outlined below, except for agricultural environmental behavior. It is of course possible that we have overlooked some studies of this nature. However, it may be that agricultural environmental behaviors involve complex activities that are unlikely to become habitual. Our discussion of the activities involved in common agricultural environmental behaviors below suggests that this may be at least a partial explanation.

Of 78 papers reviewed, 26 examined the determinants that influence environmental behaviors related to transport choice, particularly curtailing car use, choosing to use other modes of transport compared to car use, and the purchase and use of electric, hybrid, and alternative fuel vehicles. Some studies suggest that travel choice is habitual (Aarts and Dijksterhuis, 2000; Carrus et al., 2008). Further, it is possible to alter travel habits through intervention, such as by subsidizing public transport use (Hunecke et al., 2001), or increasing ease of access to public transport (Bamberg et al., 2003).

Other studies indicate that factors related to self-determination, such as attitudes (e.g., Nilsson and Küller, 2000), and altruism, such as personal norms (e.g., Nordlund and Garvill, 2002; Klöckner and Matthies, 2004; Thøgersen, 2006; Bamberg et al., 2007), explain travel choice. In addition, multiple studies have found that travel choice is explained by the TPB both singularly (e.g., Bamberg et al., 2003) and combined with other factors, such as personal norms (Harland et al., 1999), and habits (Bamberg and Schmidt, 2003).

Numerous factors are related to the adoption and use of electric, hybrid, and alternative fuel vehicles, such as attitudes (Egbue and Long, 2012; Barbarossa et al., 2015; Wang et al., 2016), self-efficacy (Bockarjova and Steg, 2014; Langbroek et al., 2016), moral norms (e.g., Jansson et al., 2011), values (e.g., Carley et al., 2013), and emotions (e.g., Moons and De Pelsmacker, 2012). However, the most common determinants are factors related to cost (e.g., Flamm and Agrawal, 2012; Hahnel et al., 2014; Lai et al., 2015; Barth et al., 2016).

Multiple studies found that the high purchasing cost is a barrier to adoption, while lower use costs compared to conventional vehicles, is a driver of adoption (e.g., Ozaki and Sevastyanova, 2011; Graham-Rowe et al., 2012). Further, pro-environmental values and attitudes, and social norms, are less influential than financial considerations (Hahnel et al., 2014; Barth et al., 2016). For example, Mairesse et al. (2012) observed that, “most people have positive attitudes towards the environment, but they do not prevail over attitudes associated with other car attributes” (p. 568), such as cost.

In comparison to the adoption and use of sustainable cars, the adoption of electric bicycles is positively related to attitudes and norms toward technology, and negatively related to norms about physical activity (Wolf and Seebauer, 2014; Kaplan et al., 2015). Taken together, these findings suggest that a number of factors can influence initial decisions about travel choices that are relatively inexpensive, and without substantive barriers habits may form around those decisions. These travel habits may also be altered through interventions. However, the adoption of behaviors that involve purchasing expensive sustainable technologies, such as electric cars, is influenced substantially by factors related to cost.

Twelve studies included in our review examined the determinants that influence environmental behaviors related to household consumption. Most studies consider multiple purchasing behaviors. Unlike most other environmental behaviors, the purchasing behaviors considered here involve the same activities: switching from conventional items to environmentally friendly items.

Both cognitive and practical factors influence the adoption of pro-environmental household consumption behaviors. However, the degree that each factor influences behavior depends both on the nature of the behavior and the context (Black et al., 1985). Easy behaviors are more likely to be related to cognitive factors, such as attitudes, while difficult behaviors are more likely to be related to practical factors, such as socio-economic context. For example, personal norms were found to be related to behaviors that face few barriers, such as reducing energy consumption (Abrahamse and Steg, 2009) and using energy saver lightbulbs (Thøgersen, 2006).

Both cognitive factors and financial factors were found to influence the adoption of green power (Arkesteijn and Oerlemans, 2005; Pichert and Katsikopoulos, 2008; Ozaki, 2011), and the installation of wood pellet heating systems (Sopha and Klöckner, 2011), smart energy systems (e.g., Noppers et al., 2016), and solar power systems (Claudy et al., 2013; Schelly, 2014; Korcaj et al., 2015). This is not surprizing given that green power typically costs substantially more than traditional power (Roe et al., 2001) and wood pellet heating and solar systems are expensive to install (Sopha and Klöckner, 2011; Schelly, 2014). Similarly, studies have found that simple water saving activities, such as turning off the faucet for teeth brushing and filling the washing machine before doing a load, are related to moral norms (Harland et al., 1999) and habits (Gregory and Leo, 2003; Martínez-Espiñeira et al., 2014). By comparison, purchasing water saving technologies, such as a water efficient washing machine, is more greatly influenced by financial factors, such as income (Martínez-Espiñeira et al., 2014).

Research also suggests that context, such as financial insecurity, also prevent households from adopting pro-environmental consumption behaviors. For example, Martinsson et al. (2011) found that attitudes are more influential for energy saving behavior in high-income homes compared to low-income homes. In summary, cognitive, practical, and contextual factors influence pro-environmental household consumption behaviors. However, those behaviors that are easier to perform are more likely to be influenced by cognitive factors while those behaviors that are more difficult to perform, particularly if high costs are involved, are more likely to be influenced by practical factors. Following the above review of environmental behaviors related to transport, we might also expect that easier and frequent household behaviors, such as reducing energy consumption and turning off the faucet for teeth brushing, are more likely to result in the formation of habits (e.g., Gregory and Leo, 2003; Maréchal, 2009; Binder and Boldero, 2012).

Thirteen studies from our review explored the relationship between determinants and the performance of pro-environmental purchasing behavior. Pro-environmental purchasing behavior involves either substituting traditional products for environmental friendly products, or reducing consumption, such as reducing meat purchasing and avoiding packaging. Most purchasing behaviors are relatively simple to perform, and many, such as purchasing recycled toilet paper, are low-cost. However, some purchasing behaviors face moderate barriers such as cost, effort, access (Davies et al., 2002; Tanner and Wölfing Kast, 2003), and context (Liang, 2016). For example, Tanner and Wölfing Kast (2003) found that the time taken to access fair trade and local produce was a barrier to purchasing behavior. Further, Liang (2016) found that the relationship between attitudes and intention was stronger for instances of low-priced organic foods compared to high-priced organic foods.

Factors related to self-interest, including attitudes and social norms, are related to purchasing behavior including purchasing recycled paper, eco-detergents, organic produce, ozone-friendly products, and avoiding products tested on animals (e.g., Schlegelmilch et al., 1996; Bissonnette and Contento, 2001; Tanner and Wölfing Kast, 2003; Fraj and Martinez, 2006). Personal norms have also been found to be related to packaging waste prevention (Thøgersen, 1999), buying organic milk (Thøgersen, 2006), and choosing not to purchase meat (Dietz et al., 1995). Further, models that combine factors related to both self-determinism and altruism have found to explain a range of purchasing behaviors (Harland et al., 1999; Chen and Chai, 2010; Yazdanpanah and Forouzani, 2015).

Further, purchasing behaviors may also become habitual. Dahab et al. (1995) found that both habits and perceived effort predicted purchasing behavior. Perceived effort influences the initial trial of purchasing recycled materials, and perceptions of the initial trial influence how likely a behavior is to become habitual. In summary, the initial act of purchasing pro-environmental products can be motivated by a range of factors. How the initial trial of a new behavior is perceived may determine whether the behavior becomes habitual, and thus whether the behavior is likely to be maintained.

Of 22 studies about pro-environmental waste management, all examined recycling behavior, and two examined composting behavior. Both the TPB (Taylor and Todd, 1995) and personal norms (Thøgersen, 2006) have been found to explain composting behavior.

The determinants of recycling vary depending on the activities involved and convenience. Common barriers include the cost of drop-off recycling, time, and limited storage space for items (e.g., De Young, 1990; Guagnano et al., 1995; Martin et al., 2006). Some studies suggest that curbside recycling behavior is explained by the TPB (e.g., Chan, 1998; Tonglet et al., 2004), and TPB with the inclusion of moral norms (Chan and Bishop, 2013; Botetzagias et al., 2015) and identity (Nigbur et al., 2010). Others have found that individual factors associated with rational choice, such as attitudes (e.g., De Young, 1990; Vining and Ebreo, 1992; Schultz et al., 1995; Schultz and Oskamp, 1996; Vicente and Reis, 2008), and PBC (e.g., Taylor and Todd, 1995) determine recycling when there are few barriers to practice. When recycling is inconvenient, such as in the case of poorly organized curbside recycling services, social-economic factors (Martin et al., 2006) and demographic factors (Lansana, 1992) are better predictors of behavior than cognitive factors like attitudes. It has also been suggested that the relationship between cognitive constructs and recycling behavior is multidirectional. For example Werner et al. (1995) found that engaging in recycling behavior strengthens positive attitudes toward recycling.

Multiple studies of recycling behaviors support Schwatz's norm activation model of altruistic behavior (e.g., Allen and Ferrand, 1999; Thøgersen, 1999), and Schwatz's model in combination with variables associated with the TPB such as PBC (e.g., Davies et al., 2002; Valle et al., 2005). Unlike factors related to rational choice, the relationship between altruism and recycling behavior is weaker when the activities involved are easy to perform (e.g., Bratt, 1999). For example, Guagnano et al. (1995) found that Schwatz's model only explained recycling in the absence of convenient curbside recycling services.

Past behavior, and therefore the formation of habits, also explains recycling when recycling activities are relatively easy to perform (Carrus et al., 2008). Similarly, González-Torre and Adenso-Díaz (2005) found that the frequency of recycling increased as distance from recycling facilities decreased. In summary, self-determination may influence the initial performance of waste management behavior in the absence of constraints, while altruism is likely to influence behavior in the presence of constraints. Habits form when behavior is easily repeated. Barriers such as cost, time, can prevent the performance of waste management behavior, while altruistic factors, such as moral norms, may positively influence behavior despite the presence of barriers.

Only eight of the studies included in our review considered the relationship between behavioral determinants and agricultural environmental behavior. Multiple other studies, particularly from the disciple of rural research, were identified that examined the relationship between common psychological behavioral determinants and agricultural environmental behavior (e.g., Maybery et al., 2005; Minato et al., 2010). However, these studies did not meet our selection criteria, and many suffer from common errors (Burton, 2004), such as the misuse of scales and measures, the simplification of models, and the tendency to over-represent the attitude construct compared to other factors widely acknowledged in behavioral literature, such as social norms (Burton, 2004).

A common example of what Burton (2004) refers to as the, “failure of agricultural geography to develop theoretically” (p. 361) is the use of very broad measures of social factors, rather than measures that differentiate specific constructs, such as social norms (Burton, 2004). For example, Greiner and Gregg (2011) concluded that landholders are more motivated by stewardship values, and less motivated by “social considerations” on the basis of Likert scale responses to a single survey item that asked landholders to rate how motivated they are by being appreciated by society and colleagues.

Behavioral research that met our selection criteria indicates that factors such as attitudes (Wauters et al., 2010) and moral norms (Raymond et al., 2011) influence the adoption of conservation behavior by farmers. For example, Wauters et al. (2010) found that the decision to adopt soil erosion mitigation strategies, such as reducing tillage, planting cover crops, and planting buffer strips was explained by landholder attitudes.

A number of studies emphasize the importance of PBC and constraints for environmental agricultural behavior. For example, Lynne et al. (1995) found that PBC influenced both the initial decision of farmers to invest in water saving technology, and subsequent investment. Similarly, Borges et al. (2014) found that the adoption of strategies to improve natural grassland, including planting new species and using fertilizer, was influenced by attitudes, norms, PBC, and actual control. In particular, PBC was determined by knowledge, skills, and technical assistance, while actual control related to the cost of new technologies. In terms of constraints, (Keshavarz and Karami, 2016) found that farmers that experienced greater drought severity were less likely to adopt conservation behavior compared to those that experienced less drought severity.

Much less research has been conducted about the behavioral determinants of agricultural environmental behavior compared to common domestic behaviors such as recycling. However, the studies reviewed here suggest that agricultural behaviors are influenced by attitudes, PBC and constraints, particularly related to drought conditions, and the cost and skills required to adopt and maintain new behavior. In contrast to the other environmental behaviors included in this review, the studies considered here do not suggest that environmental agricultural behaviors are habitual.

Numerous other environmental behaviors have been considered in the behavioral literature, notably including multiple forms of political activism. However, many of these were excluded from our review because they did not meet the criteria specified. Two that did meet our criteria and do not fall into the above categories were visiting green hotels and signing a petition. Chen and Tung (2014) found that choosing to visit green hotels was related to the factors outlined in TPB and moral obligation. Hamid and Cheng (1995) tested a modified model of the TRA, incorporating past behavior and personal control. They found that signing an antipollution petition was related to past behavior and attitudes.

The findings of this review are consistent with the theoretical perspectives outlined above. Behaviors that are easier to perform are more likely to be explained by cognitive factors, while behaviors that are more difficult to perform, particularly those related to high costs, are more likely to be explained by practical factors, such as income. Few studies consider the maintenance of new behaviors. However, given that the initial experience with a new behavior is positive, and the ongoing activities involved are easy to perform, many pro-environmental behaviors that involve continued activities are likely to become habitual. Environmental behaviors related to transport, household consumption, purchasing, and waste management involve either a one-off cost, or ongoing behavior that is relatively easy to perform. By contrast, agricultural environmental behaviors are influenced to a greater degree by barriers and constraints, particularly in relation to drought. While the initial adoption of a new agricultural environmental behavior may be related to cognitive factors, over time the additional costs and effort associated with the changing activities involved for maintenance may result in the discontinuation of behaviors. Given the changing nature of agricultural environmental behaviors, and the influence of factors related to climate on the nature and amount of maintenance required, they are less likely to become habitual.

Fundamental to the success of environmental behaviors for mitigating problems like climate change is that new behaviors, such as reduced energy consumption and reduced car usage, persist over the long-term. For the purpose of designing effective interventions, policy makers should understand which behaviors are likely to be adopted and maintained relatively easily, compared to those behaviors that are likely to require additional incentives over the long-term. In the following we classify behaviors as one-off, continuous, or dynamic, depending on the activities involved for adoption and for maintenance, and discuss the implications for research and intervention.

Most behaviors that we define as “one-off” behaviors involve a single investment to replace an existing item with an environmentally friendly alternative, such as an energy efficient washing machine. The determinants of one-off behaviors vary depending on the nature of the activity involved. If a large cost is involved, or logistical barrier, such as the amount of effort required to access renewable technologies, psychological motivations likely to be unrelated to decision (Black et al., 1985; Steg, 2008). Examples of high-cost items from our review include household energy and water saving items, such as insulation, water conservation technologies for agricultural production, and renewable technologies. A number of other items also fall into this category, including water tanks. If a small cost is involved, such as for purchasing a small energy-efficient item (e.g., a kettle or lamp), cognitive factors may be more influential.

Switching energy companies from a traditional company to a green company also involves a financial investment. However, in this example the investment is for the fee associated with terminating a contract with the current company, and establishing a contract with the new company. Green energy is more expensive than traditional energy and therefore involves an elevated cost over the long-term. In the case of both purchasing expensive items, and switching energy companies, once the initial investment or effort has been made no further effort is required. The new item or contract is utilized in the same way as the old item or contract. The use of the new item, or contract falls into the same habitual pattern as the previous option (Bratt et al., 2015).

Some one-off behaviors involve a single activity that does not include a financial cost, such as signing a petition or attending an environmental rally. Signing a petition requires little effort and may be determined by cognitive factors (e.g., Hamid and Cheng, 1995). By contrast, the effort required and the determinants of attending an environmental rally are likely to vary depending on the nature of the rally and the context. Rallies that involve prolonged disruption to traffic and access to infrastructure, such as the anti-nuclear rallies in Melbourne, Australia during the 1990s to early 2000s often resulted in arrests (A. Dempsy, personal communication, March 11th, 2009). Anti-pollution rallies in China are known to result in violence (Lee and Ho, 2014). Therefore, the decision to participate is likely to be determined by different factors compared to the recent peaceful anti-climate change marches that have taken place in cities across democratic regions, including Australia (Cornish, 2017) and the United States (Cummings, 2017). Thus, while all one-off behaviors involve the performance of a single activity at a given point in time, the determinants vary.

The behaviors that we define as “continuous” involve the same, or very similar, activities, and commitments for adoption and maintenance. Common examples of continuous behaviors include domestic recycling, water and energy saving, and choosing to take public transport to work rather than driving a car. Unlike one-off behaviors, many of the environmental behaviors we classify as continuous are relatively low-cost and low-effort to perform. This is because these behaviors involve the modification of pre-existing activities, and behavior modification is achieved more easily than the adoption of an entirely new behavior (Binder and Boldero, 2012). For example, most people brush their teeth regularly. Therefore, turning off the faucet while teeth brushing involves the modification of an existing behavior. However, some continuous behaviors do involve replacing an existing behavior with an entirely new behavior, and thus involve a moderate degree of effort. For example, switching from driving a car to using public transport is likely to require more effort than turning off the faucet while teeth brushing.

While the determinants of continuous behaviors vary, there are some common patterns. The initial adoption of a continuous behavior may be influenced by a range of factors related to altruism and self-determination, as well as practical factors. Behaviors that are easier to perform, such as purchasing products with eco-friendly packaging (Thøgersen, 1999) and domestic energy curtailment (Black et al., 1985), are more likely to be influenced by cognitive factors, while those that are more difficult to perform are more likely to be influenced by practical factors (Stern, 2000). For example, Davies et al. (2002) found that personal norms predicted curbside recycling behavior and argued that the, “task of putting household rubbish in a bin is a decision that employs a minimal amount of mental and physical effort on behalf of the householder.” (Davies et al., 2002, p. 31). By comparison, practical factors such as cost are more influential than cognitive factors when recycling involves transporting materials to a drop-off point and paying a fee (Guagnano et al., 1995).

With the exception of drop-off recycling, the continuous behaviors listed in Table 1 are performed with regularity, and therefore are likely to become habitual over time (e.g., Aarts and Dijksterhuis, 2000; Davies et al., 2002; Martínez-Espiñeira et al., 2014). Most energy and water curtailment behaviors occur daily, while choosing pro-environmental transport choices and eco-friendly purchasing are likely to occur weekly or daily. Curbside recycling involves separating waste on a daily basis, and putting a recycling container on the curb weekly or fortnightly. By contrast, drop-off recycling may occur with much less regularity, depending on household storage space and distance from a drop-off point. Reduced tillage was the only example of a continuous agricultural environmental behavior identified from our review, however there may be others that we are not aware of. Given the infrequency of planting seasons, it is possible that reduced tillage may become habitual only after years of repeating the new behavior.

The continued maintenance of environmental behaviors is rarely discussed in academic literature. However, studies suggest that while the determinants of the behaviors we term “continuous” vary, over time these behaviors are likely to become habitual. Thus, continuous behaviors should be maintained over the long-term. By contrast, dynamic behaviors present a much greater challenge for policy makers.

The behaviors that we term “dynamic” involve different activities for adoption and maintenance, therefore adoption and maintenance are likely to be influenced by different determinants. The examples considered below indicate that the amount of cost or effort required for maintenance is often greater than that require for the initial adoption of dynamic behaviors. Unlike continuous behaviors, maintenance is likely to be infrequent, or irregular, and may involve different activities from one instance to the next. Thus, dynamic behavior are less likely than continuous behaviors to become habitual.

Of the behaviors listed in Table 1, only those associated with agriculture, and purchasing expensive sustainable technologies that require ongoing maintenance, including switching to sustainable vehicles, are classified as dynamic. However, we are aware of others that are not addressed in our review, such as joining an activist group. Some forms of activism involve performing the same or similar activities with regularity, such as letter writing (Joireman et al., 2001). In other cases, the level of effort and the nature of activities vary considerably. Initially, joining an activist group is likely to involve an administrative process and attending group meetings. Over the long-term group members are called on to invest time and effort into the organization of direct actions, including rallies, education campaigns, and more elaborate and illegal protests. Organizations such as Greenpeace do not disclose the dates and activities involved for illegal direct actions, such as blockading an oil company, until close to the event (A. Dempsy, personal communication, March 11th, 2009). To participate group members may be required to take a day off work at short notice, or risk being arrested. The initial commitment of attending group meetings may well be related to cognitive factors such as attitudes toward the fossil fuel industry. The decision to engage in illegal activities at short-notice is likely to be related to numerous factors, including job security and legal record.

Dynamic behaviors that involve the purchase and continued use of sustainable technologies include switching to electric or hybrid vehicles. The initial adoption involves a high upfront cost, while maintenance involves the continued use of the purchased vehicle. The most common determinants of these behaviors are related to financial cost (e.g., Flamm and Agrawal, 2012; Hahnel et al., 2014; Lai et al., 2015; Barth et al., 2016). However, while the initial adoption of sustainable vehicles is higher than the cost of conventional vehicles, the continued use is generally less expensive than conventional vehicles (see section The Cost and Effort Involved for Adoption and Maintenance below). Thus, the upfront cost of purchase may deter adoption, while the long-term savings associated with use are a powerful driver both of adoption and continued maintenance (Ozaki and Sevastyanova, 2011; Graham-Rowe et al., 2012).

Agricultural environmental behaviors that we classify as dynamic include revegetation, either to provide a buffer between agricultural runoff and streams or to increase native species on riverbanks. Initially, revegetation involves planting seedlings. The maintenance of revegetation sites includes repairing damage from floods, storms, and animals, and extensive weeding (Moore and Rutherfurd, 2017). An additional example of dynamic agricultural behavior is excluding stock from riverbanks to improve water quality (Price and Leviston, 2014). Initially this involves constructing fencing along the riverbank, revegetation of the riverbank, and installing machinery for off-stream watering, such as troughs and pumps. Maintenance involves maintaining the revegetation site and repairing damage to infrastructure. For both revegetation and stock exclusion, the need for maintenance, and the nature of maintenance is influenced by factors that are unpredictable, such as flooding that promotes the growth of invasive weeds. Few studies have been conducted about the determinants of agricultural environmental behaviors that we term “dynamic,” and those that have only consider the initial adoption of new behavior. The adoption of these behaviors is influenced by factors related to both altruism (e.g., Raymond et al., 2011) and self-determination (e.g., Borges et al., 2014). However, rural research about landholder practices suggests that over time farmers struggle to maintain new conservation behaviors (Wilson et al., 2003, 2008).

In summary, agricultural environmental behaviors and switching to sustainable vehicles often involve different activities for adoption and maintenance, and therefore are dynamic. Sustainable vehicles are likely to be used regularly, and thus, the activities involved for maintenance may become habitual over time. By comparison, agricultural behaviors face unique barriers. Maintenance is less likely to become habitual because the activities required tend to be less routine, compared to other environmental behaviors such as recycling or choosing to take public transport rather than driving. Further, most environmental behavior research focuses on the performance of pro-environmental behavior in the daily lives of individuals, rather than behavior that effects their employment or income. In the agricultural context, environmental behavior is both performed in daily life by individuals, and is a component of landholder livelihoods. Environmental behavior tends to involve a shift from traditional agricultural practices, to practices that involve more time, effort, and cost. Therefore, these behaviors are susceptible to changing circumstances, such as drought or financial insecurity. For this reason, achieving permanent behavior changes, and thus permanent environmental improvements, may be more challenging for policy makers compared to relatively easy and low-cost domestic behaviors.

Below we consider the cost and effort involved for the adoption and maintenance of the environmental behaviors listed in Table 1.