Fostering Marine Environmental Stewardship: New Tactics Needed to Engage Millennials

The generation subculture theory asserts that as each generation comes of age and begins to harness political and economic power, their unique value-set will shape social and environmental outcomes. With this in mind, we undertook a preliminary social survey that explored the relationship between birth cohort (the Silent Generation, Baby Boomers, Generation X and Millennials), and marine environmental values, behaviour, and preferences. The findings demonstrate a declining value is placed on the marine environment across the generations, with an associated decrease in pro-environmental behaviours. The results emphasise the importance of targeted conservation outreach, particularly for Millennials.


INTRODUCTION
An exploration of cultural generations illustrates the existence of multiple national subcultures existing within one society, depicting key values that existed during various historical periods (Egri and Ralston 2004;Tung 2007;Cogin 2011). The generational subculture theory gives particular credence to the potential influence of formative experiences on later attitudes, perceptions and behaviours (Strauss and Howe 1991;Inglehart 1997;Thau and Heflin 1997;Egri and Ralston 2004). The broad cultural influences and social characteristics of the various generation cohorts have been extensively researched (Table 1). Yet, little is known about how generation might influence one's relationship to the marine environment.
Understanding how generation may influence environmental perceptions and behaviour allows for greater predictive power regarding trends in the environmental movement, as well as a more nuanced approach to conservation outreach (Gifford and Nilsson 2014). Increased knowledge of generational value differences would help determine each generation's degree of receptivity (or resistance) to environmental interventions (Schultz and Zelzeny 1999;Steg et al.. 2014). Such understanding could help enhance education and awareness programmes.
Furthermore, it has been theorised that a decline in environmental values across generations may be selfperpetuating. The notion of 'shifting baselines' may explain the impoverishment of expectations in natural resource management as an asset (such as a fishery) is gradually degraded (Pauly 1995). Similarly, "environmental generational amnesia" (EGA), first termed by Kahn and Friedman (1995), is broadly conceived as the knowledge extinction that occurs due to each generation's lack of understanding of past biological conditions (Kahn and Lourenco 2002;Papworth et al. 2009). Papworth et al. (2009) describe a process by which individuals generate environmental perceptions based on their own direct experience and fail to pass their observations on to future generations, leaving a generational knowledge gap that is filled by updated experiential perceptions of normality. This supposedly occurs continually as the population cyclesmeaning that past conditions are repeatedly forgotten and environmental norms are repeatedly updated.
The concern with EGA is that the generations responsible for managing the environment in the future will not be able to recall a time when it was not threatened (Kaas Boyle 2011). Thus, degradation and loss of ecological process and function becomes acceptable through time as we lose our environmental baseline or benchmark. This theory is particularly relevant in the Australian context, where rapid environmental change has occurred in the relatively short period since colonisation, accelerating in recent decades due to the 'resources boom.' Many coastal areas have experienced an economic shift from rural activity to resource extraction in recent years, with associated environmental degradation. The generational trends in environmental values captured by this research are likely to represent age groups with vastly, disparate ecological 'baselines.' With this in mind, this preliminary research investigates how birth cohort (i.e., generation) influences marine and coastal environmental values, behaviours and preferences.
The results discussed here comprise part of a larger survey (for details see Paterson de Heer et al. 2017). All survey questions explored here were answered via Likert scales, composed of either reported frequency, agreement, or preference.
Three survey sub-sections are addressed herein: 1. Respondents responded to a version of the New Ecological Paradigm (NEP) scale (Dunlap et al. 2000). The NEP scale is an internally consistent summed rating-scale (ibid.), comprised of 15 questions designed to test affiliation with an intrinsic valuation of the natural environment 2. Respondents were asked to report their frequency of participation in pro-environmental behaviours, representing increasing levels of commitment to environmentalism, and defined as regulated (e.g., the government regulated activity of recycling), passive (e.g., reading environmental materials, which requires personal initiative but is relatively passive), and active (e.g., active participation in an environmental group, requiring mental, physical and political action) 3. A visual component of the survey asked respondents to rank their preferences for coastal landscapes based on levels of urban interference.
Descriptive statistics were used to explore the survey population, stratified by generational cohort (Table 1). To test for significant relationships between factors, one-way analysis of variances (ANOVAs) were conducted using Microsoft Excel. A P-value of 0.05 was used for statistical significance. Whilst Likert-scored scale items such as those used in this survey are technically ordinal in their level of measurement (can be ranked, but distances between responses are not numerically meaningful), it is not uncommon for these responses to converted to numbers, and in some cases a single interval variable, and be treated as continuous for the purposes of data analysis (with the assumption that spaces between the response items are equal). Parametric tests have been found to be suitable for application to Likert-scale responses, even when assumptions of normality are violated (Sullivan and Artino 2013). Due to sufficiently large sample sizes, we have applied parametric tests to these survey results.
The NEP scale and the Marine Environment Value scale were measured for internal consistency using Cronbach's Alpha (α), and demonstrated good internal consistency (NEP α = 0.81; Marine Environment scale α = 0.92).

RESULTS
A total of 541 people participated in the surveys. Response rates (proportion of those invited to participate that completed the entire questionnaire) varied across the sample sites, but fell within the expected range for this type of research (average response rate of 55.6%; Baruch 1999).
The generic demographic data collected indicates that the sample group was reasonably representative of the Australian population as a whole. Survey variables were analysed against cultural generation (birth year cohort). A large proportion (40.8%) of survey respondents were born after 1981 (Millennials), whereas only 13.5% of the Australian population falls within this age bracket. This discrepancy represents a key limitation of the findings. However, sufficient sample numbers, both overall and across generation groups, were obtained for meaningful analysis of differences. Table 2 provides response  numbers and Table 3 compares the survey sample frame against the Australia population (Australian Bureau of Statistics 2016) to provide a measure of representativeness of the sample.

New ecological paradigm
A one-way ANOVA was calculated on generation and NEP score (Figures 1 and 2). The analysis was significant (F[1, 1072] = 906.2, P-value < 0.001). A higher score on this scale indicates greater affiliation with the NEP and thus a tendency to value the environment for its intrinsic (rather than utilitarian) worth. Baby Boomers scored highest on the scale (M = 15.32, SD = 0.68), followed by Generation X (M = 13.30, SD = 0.78), Millennials (M = 12.25, SD = 0.50) and the Silent Generation (M = 11.48, SD = 0.91).
Overall, the sample group subscribed to the ideas of the NEP. Affiliation with the NEP scale reportedly correlates with pro-environmental attitudes and a "pro-environmental orientation" (Huang 2016). This result complements the majority of prior research that have used this particular scale, including those focussed on different stakeholder groups such as general public, special interest groups, farmers and students in numerous countries (e.g., Albrecht et al. 1982;Pierce et al. 1987Pierce et al. , 1992Bechtel et al. 1999;Caron 1989;Edgell and Nowell 1989;Gooch 1995;Widegreen 1998).
The oldest (Silent Generation) and the youngest (Millennial) generations were more likely than the middle generations to have lower NEP scores, indicating that a potential shift in values is occurring across the generations. The NEP scores from this study are relatively consistent with general findings (or stereotypes) about generational value differences from other studies (e.g., Craig and Bennett 1997; Egri and Ralston 2004;Yu and Miller 2005;Crampton and Hodge, 2009). For example, the Baby Boomer generation is considered to be the most socially progressive and environmentally conscious of the generation cohorts, and in this study, were much more likely than respondents from other cohorts to indicate a very high level of affiliation with the NEP.
Having a tendency toward lower scores on the NEP scale suggests that the Silent Generation and the Millennials have a weaker affiliation with the NEP compared to the respondents from the other two-generations. This finding may be due to the widespread conservatism of the Silent Generation (Strauss and Howe 1991), and the reported apathy of Millennials (Bennett 2000;Dahl et al. 2018).
There is a substantial body of research linking proenvironmental values, such as those demonstrated by affiliation with the NEP, and positive environmental behaviours (Stern 2000;Schultz and Zelzeny 2003;Steg and Vlek 2009). However, situational factors (Gardner and Stern 1996;Kolmuss and Agyeman 2002;Harland et al. 2007), perceptual frameworks (Davis and Stroink 2016), goals (Steg et al. 2015), and personality traits (Brick and Lewis 2014) are thought to mediate this relationship. The implications of a generational value shift thus need to be teased out further, to determine whether this shift in values is likely to result in a decline in pro-environmental behaviour, and if so, to determine what the most promising means by which to foster pro-environmental values amongst Millennials and younger cohorts. The influence of generation on the frequency of Passive Environmental Behaviour (reading environmental magazines and websites), showed no statistically significant trend (F[1, 1072] = 3.5, P-value > 0.05).
The most commonly engaged pro-environmental behaviour is the regulated activity of recycling. This is unsurprising as Australian local waste management authorities often collect recycling materials from people's homes and provide local waste stations where there is no cost associated with leaving items to be recycled.
Of concern is that small proportions of the Silent Generation, Generation X and Millennials report 'never recycling.' There are several possible reasons for the distinctly lower rates of recycling behaviour amongst the Millennial respondents. This generation is characterised by their reluctance to move out of home (Szamosi 2006). In Australia, nearly 1 in 3 people aged from 18 to 34 years still live in their parental home (Australian Bureau of Sensitive 2013). This extension of time spent living in the family home (relative to previous generations) can be considered an extended adolescence, in which young people avoid taking on adult responsibilities whilst still under the  Born prior to 1945 (aged 68+) 11.8 Born prior to 1948 (aged 65+) 14.4 Born 194614.4 Born -1964 26.8 Born 1949-1958(aged 55-64) 11.8 Born 1965-1979 19.7 Born 1959Born -1988 42.4 Born 1980 onward (aged 33 and younger) 40.8 Born 1989 onward (aged 15-24) 13.5 economic and practical auspices of older family members (e.g., Twenge et al, 2012). Members of the Millennial generation tend to score high on scales of over-reliance on parents (Wilson and Gerber 2008). In these circumstances, it may be that recycling decisions and actions are made and carried out by older members of the household. Similar trends to those regarding passive engagement were detected for active engagement in pro-environmental activities. Millennials and the Baby Boomers both had medium levels of active participation in environmental groups. These results are somewhat unexpected, as generation stereotypes characterise Millennials as apathetic and the Baby Boomers as having a revolutionary outlook and willingness to challenge authority (Parment 2013), as well as high moral priorities regarding environmental issues (Smith and Clurman 2007). These strong values, coupled with formative experiences during a time of great public activism (the World War II, Korean War, and Vietnam War eras), would be predicted to lower potential barriers to active participation in environmental groups. In the case of the youngest generation, despite their reputation for apathy and inaction, it seems that Millennials are as active in environmental groups as the Baby Boomers, and indeed both are much more active than the Silent Generation. One possible reason for the relative lack of active participation in environmental groups by members of the Silent Generation is the practical issue of age and mobility.

Figure 3 Influence of Generation on Frequency of Environmental Behaviours Notes: Orange bars depict regulated behaviour, blue bars depict passive behaviour and purple bars depict active behaviour
Generation X respondents demonstrate the highest incidence of active participation in environmental groups of the four generations surveyed. Generation stereotypes suggest that this generation is environmentally conscious (Craig and Bennett 1997), but very individualistic (Yu and Miller 2005). However, whilst Generation X are often characterised as sceptical and focused on their own interests, prior research into participatory citizenship amongst this cohort recorded a high percentage of respondents having participated in volunteering and activist activities (Vromen 2003).
The majority of respondents across all groups reported that they preferred a beach landscape with mangroves present (lowest level of urban interference). The preference for a beach with mangroves was most pronounced amongst Millennial respondents, and least pronounced in the Silent Generation.
Mangroves were historically perceived as a 'pest' in Australia, given that they are often assumed to be breeding grounds for mosquitoes and other insects. Furthermore, in the early stages of coastal development, mangroves were often considered to be a 'wasteland,' with no particular utility to humans (Krauss et al. 2008). Members of the older generations may have absorbed these beliefs about mangroves and their disadvantages, making them more likely to favour the more 'managed' re-vegetation landscape. Meanwhile, in recent years much research has been published about the myriad ecosystem and human benefits of mangrove forests (Hogarth 1999;Gilman et al. 2008;Nagelkerken et al. 2008;Walters et al., 2008). To some extent, this information has trickled down to the mainstream, including environmental education programmes within schools, with members of the younger generations potentially valuing mangrove systems for their buffering, nutrient absorptive and biodiversity qualities. This held belief is in stark contrast to the public perception of earlier decades, when mangroves were viewed as having little human utility (e.g., Hursthouse 1857).
A small proportion of respondents from Generation X and the Millennials that selected the industrial image as their favourite coastal landscape (highest level of urban interference). None of the Silent Generation or Baby Boomers chose this image. It is possible that for the older generations, an image such as this conjures up notions of pollution and economic hardship. In earlier decades, industrial labour was a predominantly workingclass occupation (Pini et al. 2012), and communities living close to industrial infrastructure have recorded increased incidence of many environmental diseases (Castleden et al. 2011). Thus, the industrial landscape images may have created negative connotations within this generation cohort. Since the resources boom in Australia, many more young people are employed in the extractive/industrial sector, and wages for this type of work are among the highest of all occupations (Pini et al. 2012; http:// www.rba.gov.au). Subsequently, it is possible that the young people that selected this image as their preferred landscape because they perceive it as representing economic prosperity, rather than evaluating it for its environmental, health, or aesthetic qualities (e.g., Paterson de Heer et al. 2017).

CONCLUSION
When historically considering the results of this survey, the salient trend appears to be a diminishing value placed on the marine environment by the youngest generation surveyed (Millennials). This decline in pro-environmental values is apparent in terms of the Millennials' lower affiliation with the NEP, lower levels of participation in some pro-environmental behaviours, and increased tolerance for industrial intrusion into coastal landscapes. Thus, it seems that in order to enhance marine conservation outcomes, we must specifically target young adults. Millennials in this study have already demonstrated a preference for 'wild' coastal landscapes, and research indicates that the environment is a top concern for Millennials in Australia (Deloitte 2017). Furthermore, younger generations have previously been shown to be more post-materialist in their values than older generations, and post-materialism is a predictor of support for the environment (Inglehart 1995). There appears to be great potential to activate these values in Millennials to promote pro-environmental behaviour.
More broadly speaking, in response to this particular survey, each generation group demonstrated a unique response pattern, indicating that conservation messaging could be targeted more specifically to each of the cohorts. What is required is further exploration to develop targeted conservation outreach for various subgroups within society, in order to reverse recent trends towards decreasing pro-environmental values, preferences and behaviour.