1 Introduction

The current and continuing trends of global decline in biodiversity reflect the transgression of six planetary boundaries and the sixth major mass extinction event in the history of life on earth [25, 87]. Research agrees on the fact that the rapid acceleration and the severity of environmental issues are anthropogenic [48]. Specific human actions, such as the construction and operation of reservoirs and dams [72, 101], as well as the ongoing urbanization [73], profoundly impact biodiversity by altering natural habitats and disrupting ecosystems. These issues affect the availability and quality of natural resources, leading to further challenges in conservation efforts. The associated negative consequences for the environment and humankind manifest the urgency of the need for action through a societal switch to sustainable development. The regional level is of particular importance for the concrete implementation of the Sustainable Development Goals (SDGs) on the ground [7, 54, 108]. Here, regionally adapted concepts for the sustainable use of environmental resources must be developed and implemented together with relevant stakeholders [31, 44, 99], as specified by SDGs 17.16 and 17.17 [103]. In the case of environmental resources, these stakeholders are essentially resource users, conservation associations, public authorities, and local environmental academia. However, researchers report a lack of interdisciplinary information exchange and a high frequency of conflicts regarding land use and nature conservation between the stakeholders mentioned above [8, 85]. These issues create obstacles in the way of successfully implementing cooperative projects [84, 110].

To overcome such obstacles, different approaches to “stakeholder management” are becoming increasingly important e.g., different case studies in Cambodia [23]; China [61]; Malaysia [77]; Spain [64]; the United States and Poland [52]. Stakeholder Analyses (SA) can be used to identify relevant regional actors or to map and optimize their resource and information flows [81, 86]. Social Network Analyses (SNA) can reveal gaps in regional collaborative networks, the marginalization of specific groups, or hotbeds of conflict [69, 78, 105]. Combinations of SNA and SA have already been used in various contexts [5, 38, 58, 63] and offer the advantage that data on contextually relevant stakeholder characteristics can be interpreted together with data on stakeholder interactions. The lack of contextual content in the quantitative interaction data of an SNA can thus be circumvented. At the same time, the extensive quantitative data set of the SNA offers a view of the interactions between stakeholders that would be impossible with a solely individual view of the stakeholders, as is the case in a stakeholder analysis. With an individual perspective on the stakeholders and the overarching perspective on the stakeholder social network, a holistic view of the stakeholder landscape of a region can be created, which enables optimally grounded stakeholder management [81]. A combination of SA and SNA was used by Przesdzink et al. [81], to optimize a collaborative network of regional stakeholders with regard to mutual complements in the availability of resources and knowledge. While this approach was mainly focused on social network characteristics, resource optimization and overall ratio-based aspects of promoting stakeholder interactions, others argue to focus on more “human” aspects of stakeholders [9, 74, 92], such as trust [111, 112], values [12, 22, 111], or attitudes [4, 26, 83]. In a case study in Northwest Germany, also, social factors like different ways of operating among organizations or different ideological attitudes toward the environment were named by numerous stakeholders as crucial factors for well-functioning collaborations or profound conflicts [79].

In this paper, we therefore analyze the (1) stakeholder categories conservation associations, land use associations, public authorities and research groups regarding three psychological constructs [(2) Organizational Culture, (3) Environmental Worldview, (4) Environmental Risk Perception] that may influence the existence of collaboration or conflict between the stakeholders. First, we examine these constructs regarding their expression among the stakeholder categories in the present sample [Sect. 3, (2), (3) and (4)]. Second, we analyze correlations with the presence of collaboration and conflict between stakeholders (at the end of the respective construct`s subsection). As discussed in Sect. 4.1, the SNA in this study has limitations in capturing the qualitative aspects of stakeholder communication. To address this, future studies should not only consider the presence of collaborations or conflicts but also their quality. Methodological challenges, such as the need for better validation of questionnaires and the difficulties some participants faced with the ipsative rating scale, suggest that alternative approaches like multi-level Likert scales could be considered. Additionally, to mitigate potential biases in responses, ensuring clear instructions and considering the influence of social desirability is important. Although our results stem from a specific case study, the examined constructs are generalizable to Western societies, as discussed in the Sect. 4.2. We anticipate that further case studies in the future will contribute to solidifying these constructs as a helpful complement to current more resource-oriented stakeholder analysis and—management approaches.

1.1 Theoretical background and hypotheses

  1. (1)

    Social network analysis and homophily

    Social Network Analysis (SNA) is a crucial method for studying relationships and interactions among environmental stakeholders. By applying SNA, researchers can uncover complex network structures, analyze interactions between actors, and develop strategies to improve collaboration. Especially when combined with Stakeholder Analyses (SA), SNA enables the identification of key actors and the analysis of information flows, resource exchanges, and cooperation patterns essential for implementing sustainable practices [78]. For example, Mills et al. demonstrated how SNA can be used to identify gaps in collaborative networks and analyze the marginalization of specific groups. Other studies focus on the influence of social networks on collaborative resource management [15, 21]. SNA/SA insights are particularly relevant for developing strategies tailored to the specific needs and dynamics of regional environmental stakeholders.

    A well-documented occurrence in social networks is the principle of homophily, stating that persons or organizations with similar characteristics are attracted to one another and therefore more likely to positively interact with each other [6, 32, 78, 96]. Regarding organizations, homophily has been referred to factors such as founding date, geographical distance, or institutional preferences [6, 96, 98], regarding persons also to shared attitudes and beliefs [68]. Based on literature presented in the next sections and an interview study with 20 regional stakeholders [79], we focus in this paper on homophily regarding three psychological constructs that seem to be of particular interest in the context of environmental resource stakeholders: Organizational Culture, Environmental Worldview and Environmental Risk Perception. Each construct was analyzed in terms of its expression within the stakeholder categories conservation associations, land use associations, research groups, and public authorities, as well as for correlations with the existence of conflicts and collaboration between the stakeholders using inferential SNA [20].

    A first very general research question RQ1 relates to four broad categories (land use associations, conservation associations, public authorities, research groups), which we generated from stakeholders' statements on distinct fields of operation: Can belonging to the same stakeholder category be correlated with the presence of collaborations and conflicts between stakeholders? As hypothesis H1 we assume, that equal categories of stakeholders are positively correlated to collaborations and negatively to conflicts between them.

  2. (2)

    Organizational culture

    In this study, Organizational Culture is interpreted according to Schein [93] as a pattern of shared basic assumptions learned by a group as the correct way to perceive, think, and feel in relation to external problems. Researchers described the construct of organizational culture as shared among organization members [34], influencing decision-making processes, the management style [100], the members’ attitudes and behaviors [95], and consisting of enduring values, beliefs, and assumptions [19, 93]. The construct of Organizational Cultures is based on the “Competing Values Framework” (CVF) [82]. In CVF, organizational values are competitively conceptualized and integrated into a holistic framework. The framework consists of two basic value dimensions (“flexibility vs. stability”, and “internal focus vs. external focus”), which create axes of a matrix and form four quadrants with related organizational values (see Fig. 3). These four quadrants result in a typology of four culture types: Clan (internal focus, flexible; “do things together, collaborate”), Adhocracy (external focus, flexible; “do things first, create”), Market (external focus, stable; “do things fast, compete”), and Hierarchy (internal focus, stable; “do things right, control”). These culture types are characterized by Cameron and Quinn [19] and further described in supplementary material A. Statistical analysis demonstrated that organizations with different types of cultures differ in their organizational strategies, decision-making processes, and structures [19]. However, current research did not yet analyze influences of these organizational differences on the existence of collaboration or conflicts between organizations. However, in a regional study by Przesdzink et al. [79], aspects such as different ways of working, goal definitions, or project management styles, which can certainly be attributed to different organizational cultures, emerged as relevant factors for well or poorly functioning interactions with other organizations, which is why we included the construct Organizational Culture in this study.

    Two research questions arise regarding organizational cultures: RQ 2.1: How do the four stakeholder categories differ regarding their organizational cultures and is there a significant correlation between membership of a stakeholder category and the dominant manifestation of a particular organizational culture? RQ 2.2: Can similar dominant organizational cultures be correlated to the presence of collaborations and conflicts between the stakeholders?

    To the authors’ knowledge, there is no survey of the organizational cultures of different types of environmental stakeholders. However, organizations in the environmental sector of Western societies are generally described as inert and entrenched: For instance, Gunderson et al. [36] and Rogers et al. [91] observed that environmental stakeholders in the public sector often have an incapacity to restructure due to rigid bureaucracies, which lead to a certain resistance to change. They thus lack the competence to adapt to new forms of thinking, functioning, and structuring. Bureaucracy on the side of public authorities in particular is a frequently mentioned major problem in the German nature conservation and land use sector [24, 41, 57], which indicates the organizational culture of hierarchy in German public authorities.

    Considering the previous findings, H 2.1 surmises, that hierarchical cultures, which value stability, consistency, and uniformity will be found especially among public authorities. Overall, the more "conservative" cultures of clan and hierarchy are expected to occur as dominant cultures in a majority of the stakeholders. We therefore do not expect a significant correlation between membership of a stakeholder category and the dominant manifestation of a particular organizational culture. Regarding H 2.2, we assume that equal dominant organizational cultures of stakeholders are positively correlated to collaborations and negatively to conflicts between them.

  3. (3)

    Environmental worldview

    The “advocacy coalition framework” (ACF) suggests, that actors with similar belief systems are more likely to form coalitions [107]. Based on this groundwork, several studies have already demonstrated strong links between similar “policy beliefs” and collaborative interactions of environmental stakeholders in their social networks [45, 46, 67, 107]. In an interview study in Northwest Germany, stakeholders reported that conservation associations and land use associations would support different general viewpoints on sustainability, for example whether the priority of forest use should be the production of timber or the preservation of endangered species and deadwood or whether or not large predators such as the wolf belong in the northern German cultural landscape. Such ideological disputes would lead to conflictual relationships between specific land user groups and conservation associations [63]. Since our project does not refer to a specific policy, but to the interactions between regional environmental stakeholders in general, we decided to survey their general environmental worldview, as conceptualized by the New Environmental Paradigm (NEP) [28]. The revised version of this concept includes the five facets “balance of nature,” “limits to growth,” “antianthropocentrism,” “anti-exemptionalism” (of humankind from nature), and “possibility of an eco-crisis,” which combined indicate a “pro-ecological worldview” [29]. The validity and reliability of the NEP are demonstrated in numerous studies, [29, 43, 76]. Communicating differences in NEP-scores to the resolution of land use and conservation conflicts has already been suggested by Edgell & Nowell [30] and Kaltenborn et al. [50], even though no recent studies addressing this issue could be found by the authors.

    This leads to the next set of research questions: RQ 3.1: How do the four stakeholder categories differ in terms of their environmental worldviews and is there a significant correlation between stakeholder categories and environmental worldviews? RQ 3.2: Can differences between stakeholders’ environmental worldviews be correlated to the presence of collaborations and conflicts between the stakeholders?

    Edgell and Nowell [30], in a study in British Columbia, found high NEP-scores for environmentalists and the general public and low scores for commercial fishermen. Kaltenborn et al. [50], in Norway, found high scores for wildlife managers and research biologists and low scores for sheep farmers. Hawcroft and Milfont [43] reviewed that environmentalist samples score higher NEP values than representative population samples.

    Based on this research, we assume significant differences between NEP scores of the different stakeholder categories (H 3.1). In more detail, relatively high NEP values among conservation actors, research actors, and government agencies, and lower values among land use associations. As H 3.2 we assume differences in NEP scores of stakeholders to be positively correlated to conflicts and negatively correlated to collaborations between them.

  4. (4)

    Environmental risk perception—myths of physical nature

    In the context of the ACF and environmental stakeholder interactions, “Cultural Theory” receives special attention [88, 97]. According to this concept, people follow one of four cultural biases or “myths” concerning human nature and environmental risk perception: The “hierarchical” view of humans is sinful but controllable by rules and that of nature is robust within limits. The “egalitarian” view of humans is caring and sharing and that of nature is ephemeral. The “fatalistic” view of humans is untrustworthy and that of nature is unpredictable. The “individualistic” view of humans is selfish and that of nature is benign [35, 94]. These cultural biases not only represent general worldviews but can also be used to derive individually preferred management styles for the human and natural world in general [104] and environmental resources in specific [40]. Different preferences in the management of environmental resources can easily lead to conflicts between stakeholders [39], which is why numerous studies have already shown connections between Cultural Theory and stakeholder interactions [14, 88, 97, 106]. The biases to human and physical nature are not necessarily correlated and therefore can be considered independently [35]. Therefore, in this study, we have limited ourselves to the myths of physical nature (MoN), which are more relevant in the environmental context.

    This leads to the third set of research questions: RQ 4.1 How do the four stakeholder categories differ in terms of their myths of physical nature and is there a significant correlation between stakeholder categories and their myths of physical nature? RQ 4.2: Can similar myths of physical nature be correlated to the presence of collaborations and conflicts between the stakeholders?

    A classical but oversimplified depiction of society categorizes the passive society as fatalist, the “active society” (e.g., NGOs) as egalitarian, government actors as hierarchist and the private sector as individualist [13, 62]. Grendstad and Selle [35], found predominantly hierarchical instead of egalitarian MoN in Norwegian environmentalists. Peter et al. [75], in a study in rural Germany, found individualist and fatalist MoN in agricultural and quarrying organizations, as well as hierarchical and egalitarian MoN in research groups and conservation associations. Sotirov et al. [97], in a study on forest policy actors in Germany, found individualist MoN in private and industrial actors, egalitarian MoN in research groups and conservation associations and predominantly hierarchical MoN in public authorities. Based on these sources, as hypothesis H 4.1, we assume a predominantly egalitarian MoN in research groups and conservation associations, a predominantly hierarchical MoN in public authorities and a predominantly individualist MoN in land use associations. As hypothesis H 4.2 we assume similar MoN of stakeholders to be positively correlated to collaborations and negatively to conflicts between them.

    In summary, the following statistical considerations are relevant to answering the research questions: RQ 1—Stakeholder Category vs. Social Networks, RQ 2.1—Organizational Culture vs. Stakeholder Category, RQ 2.2—Organizational Culture vs. Social Networks, RQ 3.1—NEP vs. Stakeholder Category, RQ 3.2—NEP vs. Social Networks, RQ 4.1—MoN vs. Stakeholder Category and RQ 4.2—MoN vs. Social Networks.

2 Materials and methods

2.1 Sample

Environmental stakeholder organizations in the study region of Osnabrück in Northwest Germany (see Fig. 1) have been previously assembled by Przesdzink et al. [81] using internet research and a snowball system.

Fig. 1
figure 1

Topographical map of the city and district of Osnabrück [60]. Recognisable in brown are the mountain ranges "Ankumer Höhe" (max. 142m above sea level), "Wiehengebirge" (max. 211m above sea level within the region) and "Teutoburger Wald" (max. 331m above sea level within the region). In green, the flat parts of the district in the North German Plain in the north and in the Münsterland in the far south. Also shown: Municipal boundaries in purple with the respective municipality name in black; motorways in yellow; canals in light blue. The small river Hase, which flows through the region from south-east to north-west, and its tributaries are not shown

Full size image

The city of Osnabrück, located in northwest Germany, is characterized by its well-developed urban infrastructure and the Hase river flowing through it. The surrounding district of Osnabrueck includes broad areas used for agricultural purposes, but also diverse and ecologically significant nature reserves. Notably, the TERRA.vita UNESCO Global Geopark spans the mountain ranges “Teutoburger Wald”and “Wiehengebirge”, highlighting rich geological history. The area also features moor reserves that support rare flora and avifauna, as well as ancient woodlands and river floodplains. Overall, the study area is comprised of diverse ecosystems, making it an ideal location for analyzing diverse environmental stakeholders. In the previous study of Przesdzink et al. [81], 130 organizations were identified, 53 of whom responded to the questionnaire for this study. The classification into four types of stakeholders, namely public authorities (AUT), conservation associations (CON), land use associations (LAN), and research groups (RES), was generated from stakeholders' statements on distinct fields of operation. The organizations surveyed were equally distributed between these categories (AUT 24.5%; LAN 24.5%; CON 26.4%; RES 24.5%). The average organization size was 2381 members, with a minimum of 2 and a maximum of 100,000 (SD = 13,729). From each organization, the most relevant representative for interaction with other organizations on regional environmental issues—in most cases the chairman, professor, or a specific administrator—was surveyed on behalf of his or her respective organization. The average age of the participants was 54 years (SD = 13.5), with the youngest being 26 and the oldest 80 years old. An initial visual representation of the collaborations and conflicts between the stakeholders is shown in Fig. 2. A descriptive examination shows that there are significantly fewer conflicts (15) than collaborations (253). Regarding collaborations, public authorities were located quite in the center of the network, land use associations and research groups could be identified as quite distinct groupings that collaborate more intensively with themselves than with the rest of the network. Conservation associations were involved in no conflicts at all and were distributed between the research groups and public authorities.

Fig. 2
figure 2

Graphical representation of collaborations (A) and conflicts (B) between stakeholders in our sample. Dots represent individual stakeholders (yellow: land use associations, orange: public authorities, green: conservation associations, blue: research groups) and arrows represent collaborations or conflicts between them. Arrows point from the stakeholder who has named an interaction to the stakeholder with whom they have ticked this interaction. The color of the arrows is made up of the colors of the interacting stakeholders. The graphic was created in Gephi using the standard algorithm ForceAtlas2

Full size image

2.2 Questionnaire

This study used two online questionnaires for data acquisition, which were accessible via a self-created online platform: Social network data was collected using a list of all 130 stakeholders identified by Przesdzink et al. [81]. Participants could check whether their organization is in collaboration or conflict with the other organization. For the network analysis, only the interactions between the 53 participating stakeholders were considered, not interactions with stakeholders who did not complete the questionnaire. Organization-specific data was collected using a questionnaire incorporating 48 items: 8 items assessing sociodemographic data, 24 items assessing organizational cultures, 15 items assessing environmental worldviews, and one item assessing environmental risk perception. For the detailed items see supplementary material B and the sections below.

The questionnaires themselves have already been tested and validated in many previous studies and can therefore be regarded as robust data collection instruments. Before starting the survey, the questionnaire was revised to test its practicability and estimate the duration of completion (~ 20 min). Sample questionnaires were completed with stakeholders from all four stakeholder categories and subsequently discussed both with the stakeholders and internally to ensure that the questions did not intrude too invasively into the stakeholders’ internals and did not deviate completely from the stakeholders’ self-perceived organizational cultures and environmental views. During these trial questionnaires, a Thinking Aloud Protocol [53] was conducted and discussed afterwards to ensure that the trial participants understood the content of the questionnaires correctly. For the data set used in statistical calculations, participants were later anonymized. Both questionnaires were completed by the stakeholders on a tablet following an interview data collection for a stakeholder database that is currently being set up in our region.

The study was carried out in accordance with the national and institutional guidelines, the Declaration of Helsinki, the German Research Foundation, and the American Psychological Association. Anonymity of the participants was guaranteed, and the participation was voluntary. All participants had the chance to decline participating in the study at any time and without any consequences. The participants previously signed a declaration of consent that contained information on the main topic of the interview, the study procedure, as well as the recording and further processing of the data. As the research had no medical background, it involved no risks to our participants. Moreover, the research assessed no sensitive personal data. In consequence, an ethics approval was not required.

2.3 Assessing organizational cultures

Organizational Cultures were assessed using the “Organizational Culture Assessment Instrument” (OCAI) [19]. The validity and reliability of the OCAI have so far been confirmed by various empirical studies [19, 51, 56, 109]. Items corresponding to the OCAI were professionally translated from English into German by an external translator using the translation-backtranslation method [16]. OCAI items consider six dimensions of organizational culture (dominant characteristics of the organization, organizational leadership, employee management, “glue” that holds the organization together, strategic focus, success criteria), each containing four statements representing one specific culture type. As the OCAI was primarily developed for business consulting, some statements are formulated in a rather business-oriented sense. However, as the sample analyzed in this study does not only include businesses but mainly consists of associations, public authorities, and scientific institutions, particular market formulations were adapted to the target group (“winning in the marketplace” changed to “achieving targets” and “products” changed to “services”). Nevertheless, this did not change the basic meaning of these statements. To answer the items, participants estimated the extent to which each alternative is similar to their organization in a forced-choice format using an ipsative rating scale, in which 100 points were to be distributed among four possible response alternatives (see supplementary material B, pp. 2–5). The points had to be assigned on a scale from 0 (does not apply at all) to 100 (applies very much) by moving a control button. After the assessment, from all six dimensions closed geometric mean values were calculated for each of the four organizational cultures. Since some participants did not perform the calculations correctly, leading to the four value statements of an item not adding up to 100 points as required, the statements were in these cases adjusted by ratio calculations. One participant who failed to answer all items was excluded from the statistical analysis.

Reliability analyses showed internal consistency for all four organizational cultures (Cronbach’s α of 0.78 to 0.88), indicating a high level of reliability. Cronbach’s α ranges from 0 to 1. Values above 0.7 are generally considered acceptable, above 0.8 are good, and above 0.9 are excellent. The results were plotted graphically using kite-shaped value profiles [19], which visually represent the mean values achieved for each culture (Fig. 3). For the statistical analyses, only the “dominant”, i.e., most pronounced, stakeholder cultures were used. The cross-tabulation of the distribution of dominant organizational cultures within the stakeholder categories and vice versa revealed a weak cell occupancy (75% of cells with an expected frequency below 5). The Fishers exact test was therefore used to determine the significance of this correlation as it is more reliable for small sample sizes and sparse data than the chi-square test. Subsequently, Bonferroni-corrected z-tests were used as a post-hoc test to detect where exactly significant differences between the organizational culture characteristics of the individual stakeholder categories are. The Bonferroni correction was applied to control for Type I errors due to multiple comparisons, ensuring that the overall significance level was maintained. Additionally, Cramer’s V was used to measure the effect size. Values below 0.1 indicate a weak association, 0.1 to 0.3 indicate a moderate association, and above 0.3 indicate a strong association.

Fig. 3
figure 3

Kite-shaped value profiles for each stakeholder category, displaying the individual stakeholder categories’ average culture values (continuous line) compared to average culture values of all stakeholders (dotted line)

Full size image

2.4 Assessing environmental worldviews

Environmental worldviews were assessed using the revised NEP scale [29], which has become the most widely used tool to measure environmental worldviews [43]. It consists of fifteen items; each one is a statement referring to the relationship between humans and nature. Eight of the items, if agreed to by a respondent, ought to reflect an endorsement of pro-environmental beliefs, whereas the remaining seven items need to be disagreed with, to reflect pro-environmentalism [3, 29]. We used the German version of the NEP by Kaiser et al. [49], including 5-point Likert scales encoded from 4 (strongly agree) to 0 (strongly disagree) (see supplementary material B, pp. 6–7). Aside from summing up all items to give a single score, Dunlap et al. [29] present five subscales of the revised NEP, respectively “reality of limits to growth”, “antianthropocentrism”, “fragility of nature’s balance”, “rejection of exceptionalism”, and “possibility of an eco-crisis”. However, different studies disagree on the actual number of dimensions of the NEP, which seems to vary by sample [2, 17, 29, 59]. In our case, when combined into a single factor, the NEP has strong item-total correlations as well as a high internal consistency (Cronbach's α of 0.84). Therefore, due to the exploratory nature and already high number of different constructs examined, this study simplistically considered the NEP as unidimensional. Consequently, mean scores of all fifteen items were generated for each stakeholder as a unidimensional NEP score. Reaching an average NEP score above 2.0 was classified as representing pro-environmental beliefs. Two participants who failed to answer all items were excluded from the statistical analysis. Due to the small sample size, determining the distribution of NEP scores was important to choose appropriate statistical methods. One requirement for the ANOVA is the normal distribution of the data, which can be tested using the Shapiro–Wilk test. Values close to 1 for the test statistic indicate a good fit to the normal distribution.The results showed a normal distribution of NEP scores within the whole sample as well as within the stakeholder categories (see Table 1). To determine variance homogeneity, a Levene’s Test was used, which showed that the variances for NEP scores were not equal, F(3,49) = 0.120, p = 0.948. A single-factor ANOVA was therefore used for the statistical analysis additionally calculating η2 to measure the effect size. A Gabriel post-hoc test was then used to examine the individual stakeholder categories for significant differences.

Table 1 Results of the Shapiro-Wilk test for normal distribution of the NEP score
Full size table

2.5 Assessing myths of physical nature

MoN were assessed using only one item: participants could choose between graphic representations of the four myths (see supplementary B, pp. 8–9), which one most closely corresponds to their worldview. As with organizational cultures, the cross-tabulation of the distribution of myths of physical nature within the stakeholder categories and vice versa revealed a weak cell occupancy (75% of cells with an expected frequency below 5). The Fishers exact test was therefore used in this case to determine the significance of the correlation.

2.6 Correlating psychological data with social network data

The data entered in the SNA questionnaire were converted to two “adjacency matrices” (collaborations and conflicts), in which each stakeholder was assigned a row and a column, and cells contain data on the presence of an interaction (e.g., 1 = collaboration; 0 = no collaboration). These matrices consist of data points that are not independent of each other (the existence of a relationship between A and B and A and C may influence the existence of a relationship between B and C). This complicates the use of conventional correlation or regression analyses, which require independent data points. This problem can be circumvented using inferential SNA methods such as the Mantel test [65, 66], which can calculate Pearson correlations between network matrices. For this purpose, the metric data on the NEP score was converted into a distance matrix, in which cells contain data on the difference between NEP scores of two stakeholders instead of data on interactions between them. For the nominal attributes (stakeholder category, dominant organizational culture, MoN selection) we instead created matrices including information on which stakeholders shared the same expression of the corresponding attribute (0 = different expression; 1 = same expression). These matrices were then analyzed for correlations using the Mantel test.

While we primarily used the Mantel test due to its suitability for our data and research questions, it is worth noting that the Quadratic Assignment Procedure (QAP) is another robust method commonly used in social network analysis to address similar issues of non-independence in network data. The QAP approach, as demonstrated in studies like Muter et al. [70], involves permuting the rows and columns of the adjacency matrices to generate a distribution of correlation coefficients under the null hypothesis, thereby providing an additional layer of reliability to network correlation analyses. Although QAP was not used in this study, its relevance and robustness in the field underscore the importance of choosing appropriate methods for network data analysis.

All methods were performed in accordance with the relevant guidelines and regulations.

3 Results

  1. (1)

    Stakeholder categories

    The Mantel test was able to detect a significant correlation between the affiliation of two stakeholders to the same stakeholder category and the existence of collaboration between them (r = 0.177; p = 0.001). No significant correlation could be detected between the affiliation of two stakeholders to the same stakeholder category and the existence of a conflict between them (r = -0.005; p = 0.754). H1 can therefore only be partially confirmed: Equal categories of stakeholders are positively correlated to collaborations, but not correlated to conflicts between them.

  2. (2)

    Organizational cultures

    The most common dominant organizational culture within conservation associations (85.7%), as well as within research groups (61.5%), is the clan culture. Within public authorities, it is the hierarchical culture (61.5%), and within land use associations clan and hierarchy both score high values, respectively (46.2% each). Across the entire sample, clan culture and hierarchy culture represent the dominant culture for 86.8% of stakeholders (52.8% and 34.0%, respectively). Adhocracy culture (7.5%) and market culture (5.7%) only represent the dominant culture for seven stakeholders combined. Regarding the significance of a correlation between membership in a stakeholder category and the dominant manifestation of a particular organizational culture, the Fishers exact test found it to be significant (two-tailed p = 0.004). The Cramers-V effect size measure shows a moderate to large effect (V = 0.349). Z-test results in Table 2show that the hierarchy culture is significantly more dominant in public authorities than in conservation associations, where the clan culture is significantly more dominant. With regard to H 2.1, it can therefore be summarized that the hierarchy culture is indeed the most common dominant culture within the public authorities and is also significantly more common there than among the conservation associations. It can also be confirmed that the “more conservative” hierarchy and clan culture are the dominant organizational cultures for a large part of the sample (86.8% combined). A significant correlation between membership of a stakeholder category and the dominant manifestation of a particular organizational culture is indeed present for the entire sample and in particular for the differences between public authorities and conservation associations.

    Additionally, the comparison of average culture values in Fig. 3 shows that values for clan culture were the strongest in the group of conservation associations (M = 42.8; SD = 16.96) and the weakest in public authorities (M = 26.59; SD = 9.58). Reversely, average hierarchy culture values were the highest in public authorities (M = 35.98; SD = 16.00) and lowest in conservation associations (M = 42.89; SD = 6.39). Overall, the average values of clan and hierarchy cultures were always above 20 and represent the two strongest cultures in all stakeholder categories. Compared, average adhocracy and market values are all below 25 and were therefore the two weakest cultures in all stakeholder categories.

    The Mantel test did not find any significant correlations between equal dominant organizational cultures of stakeholders and the presence of collaborations (r = 0.059; p = 0.112) or conflicts (r = − 0.012; p = 0.383) between them. H 2.2 can therefore not be confirmed.

    Table 2 Absolute and relative frequency of the organizational cultures as dominant culture (rows) within the stakeholder categories (columns)
    Full size table
  3. (3)

    Environmental worldview

    Conservation associations had the highest mean NEP scores (M = 3.17; SD = 0.44), followed by research groups (M = 3.08; SD = 0.35) and public authorities (M = 2.98; SD = 0.48). Land use associations achieve the lowest scores (M = 2.60; SD = 0.47), which nevertheless reflect a pro-environmentally attitude. Results of the single-factor ANOVA showed a significant effect of the stakeholder category on NEP scores, F(3,49) = 4.365, p = 0.008. η2 = 0.211 suggests this effect to be moderate to large. As shown in Table 3, the Gabriel post-hoc test showed significant differences between the NEP scores of the land use associations and the scores of the conservation associations and research groups.

    With reference to H 3.1, a significant correlation between NEP score and stakeholder category can thus be confirmed. Furthermore, it can be confirmed that the NEP scores of the land use associations are significantly lower than those of the conservation associations and research groups. However, this does not apply to the public authorities. The Mantel test did not find any significant correlations between differences in NEP scores of stakeholders and the presence of collaborations (r = -0.006; p = 0.569) or conflicts (r = 0.012; p = 0.303) between them. H 3.2 can therefore not be confirmed.

    Table 3 Results of the Gabriel post-hoc test in a cross table. The stakeholder category in the first column is compared with the category in the second column
    Full size table
  4. (4)

    Environmental risk perception

    As shown in Table 4, 81% of respondents (n = 43) voted for the hierarchical myth of physical nature, which interprets nature as robust within certain limits. The Fisher's exact test failed to detect a significant relationship between stakeholder categories and preferred myths (p = 0.495). Cramers-V shows a small effect (V = 0.244).

    Regarding H 4.1, a significant correlation between myth of physical nature preference and stakeholder category cannot be confirmed. Also, the hypotheses for egalitarian myths in research groups and conservation associations as well as for individualist myths in land use associations cannot be confirmed. Only in public authorities, a prevalence of the hierarchical myth was expected. The Mantel test was able to detect a significant correlation between the preference for the same physical myth of nature by two stakeholders and the existence of collaboration between them (r = 0.086; p = 0.028). No significant correlation could be detected between the preference of the same physical myth of nature by two stakeholders and the existence of a conflict between them (r = 0.0459; p = 0.226). H 4.2 can therefore only be partially confirmed: Equal preferences regarding myths of physical nature are positively correlated to collaborations, but not correlated to conflicts between them.

    Table 4 Cross-tabulation of preferences in the myths of physical nature and stakeholder categories
    Full size table

4 Discussion

The correlation between two stakeholders belonging to the same stakeholder category and the existence of collaboration between them indicates a homophily effect, as suspected. In addition to variously investigated factors such as founding date, geographical distance, or institutional preferences [6, 96, 98]; differences that arise from working in one of the specific fields of land use, conservation, administration, and research also appear to be relevant for inter-organizational collaborations. Apart from the three psychological constructs examined, which we discuss below, we suggest for further research to characterize relevant differences between stakeholder categories more clearly, for example through qualitative interviews. With regard to the existence of conflicts between stakeholders, no correlations with any other variable were observed in our sample. However, statistical analyses of conflicts should be viewed with caution in our sample, as the small number of existing conflicts (15) reduces the statistical power of these results. More on this in the Limitations. The dependence of all conflicts in our sample on just two stakeholders is also striking. In work on the reduction of land use and conservation conflicts in the Osnabrück region, a special focus could be placed on these stakeholders.

The dominance of the organizational cultures of clan and hierarchy confirms the statements of Gunderson et al. [36] and Rogers et al. [91] on the conservative attitude of environmental stakeholders. The dominance of the hierarchical culture at the public authorities is consistent with the diverse statements on the bureaucratic, rule-oriented nature of German authorities in general [24, 41, 57]. The fact that there were no significant correlations between dominant organizational cultures and the presence of collaboration and conflict is not consistent with the assumptions we generated from qualitative interviews [79]. Due to the domination of clan and hierarchy culture in our sample, the characteristics 'flexibility and discretion' represented by the clan culture as well as 'stability and control' represented by the hierarchy culture oppose each other most commonly in stakeholder interactions. Typical values represented by clan cultures are attachment, affiliation, membership, as well as support [42]. Teamwork, participation, and open communication are typical behaviors associated with those values. These attributes stand in contrast to hierarchy cultures, where precise communication, routines, formal structures, and consistency are valued and members tend to act conform and predictable [19]. Thus, in our sample, most public authorities are used to strict routines and communicating very precisely, whereas other stakeholders may want to communicate more openly and are more flexible. Evidence from Northwestern Germany on apparent differences regarding the organizational cultures of public authorities and the other stakeholder categories can also be found in the qualitative interview study by Przesdzink et al. [79]: For instance, one land use stakeholder reported that he would like the public authorities to see themselves more as partners and less as controllers of requirements. Likewise, he wishes them to be more willing to find common solutions to problems. Typical characteristics of clan culture become evident in this statement. There is a desire for collaboration and open communication and an aversion to controlling and rigid patterns of behavior—typical for hierarchy cultures. So, even if collaborations between clan and hierarchy culture exist frequently in our sample, such differences may create barriers regarding the quality of these collaborations. We see a weakness here in our data collection, which only focused on the presence or absence of collaborations, but not on their quality. More on this in the limitations. According to Sulich et al. [100], changing organizational cultures is a lengthy process and may prove difficult for public authorities who are bound by overarching directives to specific structures and courses of action. However, how public authorities communicate with surrounding stakeholders could be a starting point for improvement. Communication aids by impartial third parties or through trained personnel could be one way to allow more compromises and make communication more equitable.

With a broader focus on dominant and average organizational cultures in the entirety of the sample, it can be stated that the stakeholders mainly focus on internal maintenance and integration [19]. If the goal is to make a network more adaptable to changes and facilitate proactive responses to environmental challenges, diversity within the network is seen as a prerequisite by various researchers studying social networks [71, 78]. Diversity especially has the potential to expand the scope of vision, which is necessary for collective decision-making and can be advantageous for the realization and efficiency of sustainable development projects. The underrepresentation of adhocracy and market cultures indicates a lack of external focus and differentiation [19], which could expand the scope of vision and benefit the stakeholder network. Adhocracies in particular could break down rigid structures through their innovative spirit and dynamic nature and could possibly collaborate well with the flexible clan cultures, as they are not as averse to change as hierarchy cultures. Market cultures could benefit the network by sticking to demands and being results-oriented. Yet, the competitive nature of these organizations could stand in the way of valuable cooperative projects. Organizing meet and greet events for organizations with adhocracy and market culture characteristics, such as youth associations or green startups can be taken into consideration. Adhocracy cultures are expected to be found in younger organizations, as the earliest stages of an organizational life cycle tend to be dominated by an adhocracy culture [19]. Because of their flexibility and open communication skills, clan cultures could function as first collaboration partners and pave the way for the entry of new organizations into the regional network.

The significantly lower NEP scores of land use associations compared to research groups and conservation associations are in line with our expectations based on Edgell and Nowell [30], Kaltenborn et al. [50] and Hawcroft and Milfont [43]. However, it should be emphasized that land use associations were still generally classified as pro-environmental. This pro-environmental orientation could be used to initiate greater involvement of land use associations in conservation projects. In addition, the general pro-environmental worldview of almost all stakeholders can be used to emphasize the shared worldview of all participants in cooperative projects and act as a groundwork to generate common visions for back-casting methods [89, 90] or similar interventions. The lack of correlations between NEP scores and the presence of collaborations and conflicts between stakeholders does not correspond to expectations based on the Advocacy Coalition Framework and existing publications on the influence of policy beliefs on collaborations between environmental stakeholders [45, 46, 67, 107]. The interview study by Przesdzink et al. [79] suggested a connection between the different environmental worldviews of land use associations and conservation associations and the conflicts that would frequently occur between these groups of actors. However, it should be noted that these interviews related to the general situation of environmental stakeholders in the region of Osnabrück. Of these, we have already identified over 180 during the writing of this article. Our sample therefore only represents part of the regional network—possibly not the part in which the conflicts mentioned take place. Furthermore, the lack of qualitative data on collaboration and the statistically low power of correlations of the few identified conflicts should be pointed out here again. Different worldviews can trigger emotional and ideological discussions, which can in turn be harmful to inter-organizational interactions, especially when conservation actions and land use are discussed [18]. To sum up, regarding conservation conflicts in the region, we recommend communicating different environmental worldviews—if present—to resolve conflicts, as suggested by Edgell and Nowell [30] and Kaltenborn et al. [50].

The public authorities' preference for the hierarchical myth is in line with both the "classical" depiction of society [13, 62] and the data of Sotirov et al. [97] on forest policy actors in Germany. However, the fact that all other stakeholder categories also prefer the hierarchical myth does not match most of the sources considered [13, 62, 75, 97]. This is particularly interesting as the last three studies were also conducted in Germany so one might have expected a similar picture. Only Grendstad and Selle [35] also found a hierarchical myth in environmentalists, otherwise the groups were always assigned to other categories. The stakeholders in our sample therefore seem to have a very "diplomatic" environmental worldview—the graphic representation of the hierarchical myth is essentially between those of the individualistic and fatalistic myth. This could indicate a social desirability bias: Some stakeholders may have tried not to give too "extreme" an answer in the sense of "the world is fine" or "the world is on the brink" and therefore chose the hierarchical myth as a middle ground. For future cooperative projects with stakeholders, however, as already explained regarding environmental worldviews, there is also the option of starting with the similarities between the stakeholders. At the beginning of an intervention, it could be communicated that all participants share a rational view on sustainable land management and that nature conservation and land use can therefore approach each other cooperatively. The correlation between myths of physical nature and the existence of collaboration between stakeholders is in line with expectations based on the existing literature on connections between aspects of cultural theory and environmental stakeholder interactions [14, 88, 97, 106]. However, the statistical power of this correlation must be questioned due to the low preference for three of the four myths. In addition, involving stakeholders with different myths in environmental projects may improve their results, as the different myths are associated with different rationales and management styles that can complement each other [37, 39]. In such cases, communication of the different world views can be encouraged by stakeholder managers to avoid potential ideological disputes.

4.1 Limitations

This study has certain limitations that should be acknowledged. One primary limitation is the sample size, particularly regarding the conflict network. The number of conflict ties reported is relatively small, which inherently limits the power of the associated statistical analyses. However, this limitation highlights a critical aspect of studying conflict networks in environmental contexts: conflict ties are typically less frequent than collaborative ties, which should be anticipated in future research on conflict networks. So, a focus on conflicting stakeholders would be interesting for future studies. A larger-scale, supra-regional social network analysis could be used to identify an identical number of conflicting and non-conflicting actors and only conduct an analysis between them. A similar approach could also be taken with actors in collaboration. This would produce a more robust statistical result.

Another limitation concerns the scope of the Social Network Analysis (SNA) employed in this study. While SNA provides valuable quantitative insights into the relationships between stakeholders, it inherently reduces the richness of these relationships to numerical data. This does not undermine the value of SNA but rather emphasizes that it should be used in conjunction with qualitative methods to provide a more comprehensive understanding. SNA is not intended to replace qualitative data but to complement it by revealing patterns and structures that may not be immediately apparent through qualitative approaches alone. The qualitative interviews conducted by Przesdzink et al. [79] suggest influences of all psychological constructs examined on the quality of communication between stakeholders that could not be captured by our SNA. Thus, we suggest that future studies integrate SNA with qualitative data to capture the full complexity of stakeholder interactions. Beugelsdijk et al. [11], for example, specifically emphasize “success” of a collaboration to be negatively affected by different organizational cultures.

Although our German-language version of the OCAI was created to the best of our judgment using the translation-backtranslation method, no separate validation of the scale took place prior to this study. Also, the ipsative rating scale of the NEP questionnaire proved to be problematic in that most participants did not manage to divide 100 points arithmetically among four statements. In this study, the values were adjusted by ratio calculations, which is not the intention of the original OCAI. The cause of the problem could be an inaccurate reading of the instructions or simple calculation errors on the part of the respondents. As an alternative to the forced-choice format, a multi-level Likert scale would also be conceivable but a lower degree of accuracy level must be expected in this case.

The answers of participants may have been biased by several factors, such as the context, the wording of items, and the instructions in the questionnaire, as well as social desirability. The latter is especially to be considered in environmental psychology research, as it can be the case that respondents with pro-environmental beliefs might be biased upwards, while respondents being less pro-environmental may underreport their beliefs in studies openly dedicated to sustainability issues. Survey results were incorporated into a regional stakeholder database where they are visible to stakeholder managers, which might have affected responses.

It should be noted that surveying all members of an organization would have been statistically more powerful for the characterization of a whole organization. However, this work input would have been beyond the scope of this study. For this reason, only one representative of each stakeholder organization was surveyed. Cameron and Quinn [19] report that each organization may consist of different subcultures, as there are often different departments in organizations. However, they also state that even subunit cultures of one organization contain core elements of the entire organization’s culture. Additionally, the social environment influences other people, and the behavior of a person is therefore heavily influenced by the behavior and expectations of others. Underlining, Friedkin [32] states that the attitudes and opinions of one member reflect those of his or her surrounding members. It is therefore expected that members of an organization share certain core values regarding the aspects assessed by our questionnaire.

A general criticism of the concept of the myths of physical nature should be noted: Since the perceived state of nature is dependent on the spatial or temporal scale under consideration [35], different interpretations of the stakeholders—for example, a focus on the state of nature in the region or in the world—could have influenced their statements about their myth.

Finally, it must be considered suboptimal that our constructs and their correlations to the social network of stakeholders are only considered individually and not weighted in relation to each other. As the constructs originate from completely different branches of research (sociology, organizational psychology, environmental psychology), we have not yet succeeded in developing a theoretical derivation for such a weighting. This should be a focus of future studies in order to develop an integrative methodology for psychological stakeholder analysis. Such a methodology could be particularly interesting for active interventions in a stakeholder network in the sense of regional stakeholder management [80].

4.2 Transferability

Regarding the transferability of our approach to other cases and regions, depending on the respective framework conditions, it may make sense to select other stakeholder categories than in our case. For example, indigenous clans, public–private partnerships for renewable energy or planned urban expansions may not fit into our scheme and require the expansion or modification of our four categories, which we ourselves have chosen based on our individual situation. The psychological constructs of Organizational Culture, Environmental Worldview and Environmental Risk Perception have been established by numerous studies in “western societies”, as outlined in the respective theoretical background. Although “western societies” are not uniformly defined in the literature on our psychological constructs, we assume that the methodology used in this study can be considered easily applicable to other regions in Western societies, as defined by Kröger et al. [55]. Apart from the identity of the stakeholders, no other prior knowledge is required to conduct a comparable study. Outside of Western societies, the applicability of our methodology is questionable, as the constructs used are not sufficiently established there in some cases and non-Western societies differ in the following aspects relevant to our constructs:Regarding Organizational Culture, organizational structures tend to be more hierarchical and authority-driven compared to the flatter, participative structures often seen in western contexts [47]. Collectivist values in non-western cultures emphasize group harmony over individual autonomy, which can impact organizational dynamics [102]. Regarding Environmental Worldviews, Traditional Ecological Knowledge (TEK) plays a significant role in shaping environmental worldviews in non-western societies, differing from the scientific-rational approaches emphasized in the NEP [10]. In general, deeply rooted spiritual or religious views of nature are common in non-western societies and can significantly influence environmental attitudes and behaviors [33]. Environmental Risk Perception and the management of environmental risks are influenced by cultural norms and traditional practices in non-western societies [27]. Community-based risk management is common, reflecting a collective approach rather than the individualistic perspectives more typical in western societies [1].

5 Conclusion

In this study, we investigated (1) four stakeholder categories conservation associations, land use associations, public authorities and research groups, as well as the psychological constructs (2) Organizational Culture, (3) Environmental Worldview, and (4) Environmental Risk Perception to understand their influence on collaboration and conflict among environmental stakeholders in Northwest Germany. Our findings suggest that homophily in social networks plays a significant role in facilitating collaborations among stakeholders, aligning with the principle that entities sharing similar characteristics are more likely to interact positively. This effect was notably seen in the alignment of stakeholder categories with collaboration, although no significant correlation with conflict was observed.

Key findings of our study are that (1) Equal categories of stakeholders are positively correlated to collaborations, but not correlated to conflicts between them. (2) Stakeholder categories significantly differ regarding their organizational cultures, with Conservation associations and research groups predominantly exhibiting Clan culture, emphasizing collaboration and flexibility, while public authorities predominantly display a Hierarchy culture, highlighting stability and control. In addition, the study highlights (3) significant differences in environmental worldviews between conservation stakeholders, but a generally pro-environmental orientation across all stakeholders as captured by the New Environmental Paradigm (NEP). This indicates a shared foundation for enhanced collaborative efforts. (4) Analysis of Environmental Risk Perception through the lens of Cultural Theory’s myths of physical nature indicated a strong preference for the hierarchical myth across all stakeholder categories. This preference suggests a shared belief in the resilience of nature within certain limits and highlights an opportunity for collaborative engagement based on common environmental risk perceptions.

Our study highlights the complexity of stakeholder interactions in environmental conservation efforts and underscores the importance of considering psychological and social factors in the development and implementation of effective conservation strategies. By recognizing the role of Organizational Culture, Environmental Worldview, and Environmental Risk Perception in shaping stakeholder interactions, conservation efforts can be better tailored to address the nuances of stakeholder dynamics.

This research contributes to the broader understanding of environmental stakeholder management and provides a foundation for future studies to explore relationships between psychological constructs and stakeholder interactions. It calls for integrated approaches that combine different sociological and psychological considerations to foster cooperation and mitigate conflicts in regional sustainable land use. As environmental challenges continue to evolve, such an approach will be crucial for achieving sustainable development and conservation goals, not only in Northwest Germany but also in other regions facing similar environmental and societal complexities.