Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Gaps in Quantitative Decision Support to Inform Adaptive Management and Learning: a Review of Forest Management Cases

Gaps in Quantitative Decision Support to Inform Adaptive Management and Learning: a Review of... Purpose of Review Theoretical frameworks for adaptive natural resource management are quite common, whereas documented examples showing successful implementation of adaptive management and learning through multiple time intervals have remained uncommon. Measures of quality of adaptive natural resource management processes are needed to examine potential factors driving the successful implementation. To address this gap, we developed a multimetric index composed of 22 metrics to assess quality of case studies using quantitative decision support (QDS) to inform adaptive forest management (AFM). Metrics represented three main tasks, including conceptual setup, modeling, and application. We further distinguished these into subtasks: definition of objectives and management options (setup); specifying uncertainty, prediction, and optimization (modeling); and stakeholder involvement along with practice and learning (application). We used a multimetric index to examine temporal and geographic variation in quality of reviewed case studies using QDS to inform AFM. We then conducted a structured literature review of 179 articles, wherein 34 case studies met a priori criteria. Recent Findings When applying the multimetric index to these case studies, we found that over the past decade the index has been intermediate and annual average scores declined by 33% from 4.5 to 3.0 of 10 (where 10 is the highest possible quality score). Aligning with reviews of adaptive natural resource management, reported on-ground application of QDS to inform AFM was rare (n=2). We also confirmed the expectation that there has been a substantial lack of stakeholder engagement during QDS development tasks. Summary Our multimetric index provides a novel tool to examine gaps in the use of QDS for adaptive management in diverse domains including but not limited to forests. . . . . . Keywords Adaptive management Case studies Decision support Forest management Multimetric index Stakeholder engagement Introduction Adaptive management has been recognized as a promising ap- proach to inform and iterate decision-making and promote This article is part of the Topical Collection on Forest Management learning to achieve conservation and natural resource manage- Electronic supplementary material The online version of this article ment objectives under uncertainty in diverse contexts [1, 2]. (https://doi.org/10.1007/s40725-018-0078-3) contains supplementary Although there are many ways in which adaptive management material, which is available to authorized users. has been interpreted [3� , 4� ], a common thread among disparate approaches emphasizes learning by doing rather than managing * Brady J. Mattsson brady.mattsson@boku.ac.at systems as if they were static or so unpredictably dynamic that they preclude the possibility to learn. Theoretical expositions and frameworks for adaptive natural resource management Institute of Silviculture, University of Natural Resources and Life Sciences, Peter-Jordan-Straße 82, 1190 Vienna, Austria (henceforth, adaptive management) are quite common, whereas documented examples showing successful implementation of Present address: Institute of Wildlife Biology and Game Management, University of Natural Resources and Life Sciences, adaptive management and learning through multiple cycles of Gregor-Mendel Straße, 1180 Vienna, Austria decision-making have remained uncommon [3� , 5]. Proposed University of Freiburg, 79106 Freiburg, Germany factors limiting implementation of adaptive management based �� 112 Curr Forestry Rep (2018) 4:111–124 on literature reviews are many and include level of organiza- complex system, as component indices can be examined indi- tional and financial support for the approach [6, 7], appropriate vidually and in aggregate [19]. Investigating a total index application of the approach [3� ], concise and vetted problem score can account for simultaneous yet minor shifts in multi- definition [2], stakeholder involvement in tailoring adaptive ple subcomponents that would otherwise go unnoticed if ex- management plans to local situations [5, 8], acceptance of the amined individually. On the other hand, the quality of a sys- approach by practitioners [9], communication between scien- tem as a whole may remain unchanged while the quality of tists and managers [3� , 4� ], and balance between planning and subcomponents shift in contrasting ways that cancel each oth- real-world action [10]. This leads to the question: Which factors er out. Such component dynamics could be crucial for improv- are most limiting for successful implementation of adaptive ing understanding and decision-making in natural resource management and how can they be overcome? management. Therefore, examining both total scores and sub- Viewing the adaptive management process through the lens of component scores of a multimetric index is important to ob- quantitative decision support (QDS) offers a way to examine and tain a comprehensive understanding of a multipart system. decompose this question. We define QDS as an approach to Developing and using a QDS to inform natural resource man- inform management choices by specifying ultimate management agement is a highly complex and multidimensional endeavor objectives, identifying at least two possible management strate- [13 , 21, 22 ], which warrants such a multimetric index ap- gies, and comparing the strategies using a predictive modeling proach to examining quality of the components and entirety of procedure that includes at least one quantitative variable to rep- the QDS development and application. resent the objectives. This definition complements definitions of Forests offer a particularly interesting study system for de- decision support systems in the field of forest management (e.g., veloping indicators of QDS quality for adaptive management [11, 12]), which have emphasized all-in-one computer-based sys- and formal learning. Forests are managed at multiple spatial tems that usually require quantitative inputs. QDS need not be scales (e.g., stand, forest, and landscape level; [23]) and the entirely computer-based and would therefore be suitable for ap- majority is publically owned, but forests are becoming in- plication in regions of the world where computer resources are creasingly owned by private individuals [24]. Management limited. We propose that QDS offers a transparent means to of forests is challenged due to many sources of uncertainty inform adaptive forest management (AFM) and thereby improve and long turnover rates and time lags of woody productivity learning and achievement of management objectives relative to [25, 26]. Forest management is also a diverse subject that we approaches that do not use QDS. The parameter values for the define as any decision-making process that affects conditions predictive modeling may be all theoretical, all empirical, or some in a forest. The conditions could be abiotic (e.g., temperature, combination of these as long as they are quantified. To increase precipitation, fire), biotic (e.g., tree growth, animal population the chances that QDS will inform real-world decision making, demographics), or a combination of these [27]. Forest man- stakeholders and end users should be involved throughout the agement actions can also be diverse (e.g., timber harvest, tree development process [13 ]. planting, protection against disturbance by recreationists, pro- QDS canbeusedtoinformimplementationof forestman- viding artificial nest cavities). Correspondingly, there are agement, but published examples of cases are rare [6, 14, 15]. many types of QDS, which range from more academically One possible explanation for this rarity is the lack of documen- driven dynamic optimization tools (e.g., [28]) to graphical tation rather than a low rate of true uptake of QDS for AFM, interfaces that managers can readily learn and use on their and certainly better documentation of use vs. nonuse is needed. own (e.g., Heureka; [29]). For reviews of decision support The reasons for the lack of uptake of QDS overlap with the systems available to inform forest management, see forest reasons for the lack of implementation of adaptive manage- DSS Community of Practice [30] and Segura et al. [31]. ment. Lack of sufficient involvement by stakeholders and end Existing reviews have pointed out prospects for QDS and users is a commonly cited reason for the rare uptake of QDS in formal learning to be useful to inform forest management in natural resource management (e.g., [13, 16]). Developing an the face of uncertainties about climate change [23, 32]. indicator of quality of adaptive management among tasks of Although the topic of learning through actual implementation QDS development and implementation would provide a means of adaptive management has been reviewed [5], a comparative to learn about drivers of successful implementation, but metrics assessment of the quality of decision support to inform to construct such an indicator are currently lacking. adaptive management is needed for a better understanding Recently, multimetric indicators have been developed to of the mechanisms limiting the application of adaptive quantify impacts of management activities on ecosystem func- management. tions and services [17–19], for assessing participatory devel- Our overarching objective is to identify the key gaps in the opment of decision support systems [14] and for evaluating development and application of QDS that allow for the imple- collaborative planning approaches to support natural resource mentation of AFM to improve learning and achievement of management [20]. Using multimetric indices allows for com- management objectives in a changing world. Toward this end, prehensively and consistently characterizing elements in a we develop a multimetric index of quality of applications of �� �� �� �� �� Curr Forestry Rep (2018) 4:111–124 113 QDS to inform AFM. We then conduct a structured literature optimization, < 60 focused on decision support, and < 50 fo- review to evaluate them according to the selected attributes cused on Bayesian updating. and the multimetrics index. Next, we utilize the index and In the second round of searching, we developed a final set component metrics to examine temporal and geographic var- of search keywords (Table 1), which were chosen based on iation in the use of particular methods and approaches, to our main objective, the first round of searches, and delibera- enhance understanding about any changes or patterns in the tion among coauthors. Our goal in this final search was to comprehensiveness and application of QDS to inform AFM. identify approximately 200 papers, which would be a feasible The geographic analysis enables us to identify broadscale number of papers for the coauthors to review. We also had a gaps in the application of QDS and AFM. We then discuss goal of selecting at least 30 case studies, and we believed this the evolution of the interface between QDS and the applica- would be possible from a set of 200 papers identified via the tion of AFM, along with complementary cultures of decision keyword search. When combining the final set of keywords support and forest management that could learn from each with an OR operator, 188 papers were identified. Of these other in terms of best practices. We use regression models to matching papers, nine were classified as review papers in examine a priori hypotheses, and we use an exploratory clus- the Web of Science database and were removed from our list. ter analysis to generate ideas for future research on adaptive We then read the title and abstract from each of the 179 management that complements the vast existing literature. candidate papers identified from the final set of search key- words to determine if it would be selected for inclusion in the analysis. In particular, we determined whether each candidate Methods paper included a case study that developed and described QDS in a forest management context (for definitions of We hypothesized that the quality of case studies using quan- QDS and forest management, see BIntroduction^). In some titative decision support to inform adaptive forest manage- cases, we read the candidate paper if we could not determine ment has varied across time (temporal hypothesis), among its relevance with confidence based on reading the title and global geographies (geographic hypothesis), and among and abstract alone. To examine the recent trends, we included pa- within tasks of QDS development and application (task hy- pers published between 2005 and 2015 in the final analysis. pothesis) or has been constant across time, space, and tasks (null hypothesis). To examine these hypotheses, we identified Attributes and Scoring Selected Case Studies relevant case studies in the literature and developed a method for scoring their quality. To score the quality of QDS to inform AFM in each selected case study, we identified 23 attributes comprising three main Case Study Selection tasks of adaptive management, including conceptual setup Table 1 Final search criteria used to identify relevant papers for the To ensure relevant case studies were included in the analysis literature review on quantitative decision support to inform adaptive and that a suitable number of papers were selected, we used forest management. Asterisks (*) indicate wildcard characters, quotes two rounds of literature search within the Web of Science (B) surround exact phrases, and numbers at the end of each row indicate database on 10 August 2017. In designing the selection pro- the number of papers matching that set of criteria in the Web of Science database on 16 December 2017. Unless otherwise noted with italics, cess, we followed the principles of systematic literature re- search terms were applied to the titles, abstracts, and keywords within view [33, 34], with some simplifications to keep the process the database feasible. In the first round of searching, our goal was to iden- tify a series of keywords that reflected our main objective and Search criteria Number of papers assess the number of matching papers for multiple combina- (forest* or woodland*) AND tions of keywords. Toward this end, we conducted a series of (Badaptive management^ AND learn*) OR 94 exploratory keyword searches with increasing specificity (Badaptive management^ AND Bdecision 25 (Online Resource 1). For example, adaptive forest* manage- support^)OR ment yielded 1429 matches, whereas refining these with AND (bayes* updat* AND manage*) OR 17 Bdecision support^ yielded 54 papers. We aimed for a close fit (cited paper with title or title is BUsing 13 to our main objective while keeping the search as broad as Bayesian belief networks in adaptive possible to allow for any application of adaptive forest man- management^ )OR ((Bdynamic programming^ OR Bdynamic 58 agement, not limited to any particular definitions of adaptive optimization^) AND stochastic) management [3� , 4� ]. During the first round of searches, we found that there were over 1400 papers that appeared to focus Search term was used to limit all sets of search criteria, to maintain a on adaptive forest management, and of these, < 300 focused focus on forest management on modeling, < 200 focused on learning or dynamic Ref. [15 ] �� 114 Curr Forestry Rep (2018) 4:111–124 (n = 5), modeling (n = 10), and application (n = 8; Table 2). Recognizing the importance of balanced emphasis among We further distinguished these into subtasks, including defini- attributes of tasks and the tasks themselves, we assumed equal tion of objectives and management options (setup); specifying weight among attributes within tasks and equal weight among uncertainty, prediction, and optimization (modeling); and tasks when computing scores for tasks and total scores. In stakeholder involvement along with practice and learning (ap- particular, we summed attribute scores within each task, and plication). We recognize that some subtasks can be done si- each sum was standardized on a scale from 0 to 10. Likewise, multaneously (e.g., stakeholder involvement and defining ob- we summed task scores and computed a total score that was jectives), but each subtask on its own represents an important then standardized on the 0 to 10 scale. aspect of adaptive management. We are aware of the many definitions of adaptive management (e.g., [7, 10, 42–44]), which to a large degree represent the inherent flexibility in Statistical Analysis how the approach is applied in a given context [10] but also owing to fundamental differences in approaches [4� ]. We have We used linear regression to examine our hypotheses about chosen the tiered system of tasks and subtasks to account for spatiotemporal variation and differences among tasks and sub- these diverse approaches while facilitating our analysis and to tasks regarding quality of QDS to inform adaptive forest man- help ensure that our framework could be extended to other agement, after confirming that the model assumptions were domains beyond forest management with varying levels of met (i.e., normally distributed residuals and homoscedastici- complexity. The three main tasks we identified are largely ty). Each model included the index of quality as a response consistent across approaches, whereas the subtasks within variable, along with one of the following predictor variables: modeling are more specific to a decision-theoretic school year (continuous), continent, task, subtask, or attribute. We of thought [4� ]. In addition to attributes for scoring, we used an alpha level of 0.05 for determining statistical signifi- also classified each selected case study according to the cance and Tukey’s honest significant difference test to conduct location and climatic zone (Table 2). These classifications pairwise contrasts of means. We also conducted k-means clus- were not used for the scoring, but rather to examine the ter analysis to explore similarities within groups of case stud- geographic hypothesis by comparing scores among loca- ies across scores for subtasks of adaptive management. To tions and climatic zones. determine the number of clusters for this analysis, we exam- For each attribute describing a task of adaptive manage- ined the frequency of optimal numbers of clusters across 30 ment, we assigned a score from 0 to 10 representing low to independent methods, setting the maximum possible number high quality for that attribute (Table 2). Assignments of qual- of clusters to 10 to ensure feasible interpretation [45].We used ity were subjective, and we used our collective judgment and program R for all statistical analysis [46]. logic in designing the scoring criteria. For example, stakehold- er input on the objective function was assigned 5 points if stakeholders (or literature produced by stakeholders) were consulted on the structure or parameterization. If stakeholders Results were consulted on both structure and parameterization, then 10 points were assigned. We reasoned that structure and pa- Based on a structured literature review, we selected 34 case rameters of the objective function are equally important and studies, each of which used QDS to inform adaptive forest additive in contributing to this criterion. management and was published between 2005 and 2015 The scores were readily assigned based on simple attributes (Online Resources 2 and 3). Across these case studies, there except the quality of recommendations, which was scored were 25 unique lead authors and 1–5 case studies published according to the level of conciseness of the text and graphics per year. Two of the 34 case studies reported that QDS was representing the suggested course of action given the QDS used to inform real-world AFM [15�� , 47�� ]. The majority of inputs. For example, a case study received a higher score if case studies (n = 21, 62%) modeled a single group of ecosys- it presented a clear graph showing how predicted outcomes (in tem services (ESs), and more case studies modeled two or terms of the objectives) change among management options. more groups of ESs (n = 8, 24%) than those that modeled none We also assigned a higher score for quality of recommenda- (n = 6, 18%). Provisioning services were modeled by the ma- tion if there was a clear one to two sentence summary of how jority of case studies (n = 24, 71%), followed by regulating the results could be used to inform decision-making. Such (n = 14, 41%) and cultural services (n = 1, 3%). The most concise descriptions would be interpretable by less technical frequently modeled group within provisioning services was stakeholders and decision-makers, and the quality of the rec- Bbiomass^ (n = 23), and the most frequently modeled group ommendation is an important indicator of successful knowl- within regulating services was Blifecycle, habitat, genetic^ edge transfer between science in practice in the context of (n = 9; Table 3). A cultural service (i.e., recreation access) QDS and AFM. was only modeled in one case study [55]. Curr Forestry Rep (2018) 4:111–124 115 Table 2 Attribute descriptions and scoring schemes for a multimetric the attributes (numbered) within subtasks (lettered) and tasks (Roman index applied to case studies using quantitative decision support (QDS) to numerals) of adaptive management are provided, and wherever none of inform adaptive forest management. Descriptions of points assigned to the conditions were met the score was set to 0 Attribute Scoring Description and justification Location of landscape/region Attributes used to examine geographic patterns in the multimetric index Continent(s) – Continents where the case study was applied; does not imply that the case study is relevant to entire continents Country(ies) – Countries where the case study was applied; does not imply that the case study is relevant to entire countries Climatic zone(s) – Climatic zones where the case study was applied; does not imply that the case study is relevant to entire climatic zones (I) Conceptual setup (A) Objectives 1. Group(s) of ecosystem services 1 ES = 2 points; 2 ES = 5 points; Classification based on CICES [35] version 4.3. Useful for (ES) modeled 3+ ES = 10 points characterizing diversity of objectives, acknowledging calls for multifunctional forest management [36], and attaining sustainable development goals [37] 2. Duration of temporal horizon 10–19 = 1 point; 20–39 = 2 Evaluating applicability of model, considering long for modeled objectives (years) points; … 100+=10 points turnover rates of woody productivity [25, 26] 3. Spatial extent of modeled objectives Specified = 10 points Important to specify the area to which the model has been developed when applying it to real-world forest management [27] (B) Management options 4. Number of time intervals during which 1 point per for each interval 2–10; Evaluating the applicability of the model, considering actions can change for a given 11+ intervals = 10 points the iterativeness and adaptability of forest management unit management practices [27] 5. Specified size of smallest area within Specified = 10 points Evaluating the applicability of the model, recognizing which actions were modeled that forest managers require spatial specificity to implement recommendations [27] (II) Modeling (C) Uncertainty 6. Modeled climate scenario(s) 1 scenario = 2 points; 2 scenarios = 5 Important consideration recognizing strong influences points; 3+ scenarios = 10 points of climatic variables on forest dynamics and interactions with forest management [38] 7. Multiple relationships considered Yes = 10 points Important to consider given the high level of uncertainty when modeling effects of climate about future climate [39]. See also justification for variables BModeled climate scenario(s)^ 8. Multiple relationships considered when Yes = 10 points Important to consider given uncertainties about modeling effects of management options long-term outcomes of forest management strategies [27, 38] 9. Multiple relationships considered 1 uncertainty = 2 points; 2 Important to consider given uncertainties about when modeling effects of factors other uncertainties = 5 points; 3+ influence of other factors having strong influence than climate or management effects uncertainties = 10 points on forest management, e.g., timber prices [27]. (D) Predicting consequences of management options 10. Specified probability distributions Yes = 10 points Useful for explicitly incorporating uncertainty in a (stochasticity)aspartofpredictive predictive model and can be updated using Bayes’ modeling theorem [15�� ] 11. Displayed model as a box-and-arrow Yes = 10 points Visualizations allow forest managers and other relevant diagram stakeholders to understand and provide input to the model structure and parameters [15�� ]. 12. Model parameterized Yes = 10 points Populating model with quantitative information allows for transparent forest management recommendations [27]. 116 Curr Forestry Rep (2018) 4:111–124 Table 2 (continued) Attribute Scoring Description and justification (E) Optimization 13. Type of optimization Nondynamic optimization = 5 points; Optimization allows for management recommendations dynamic optimization = 10 points that are explicit for predefined states of the system, e.g., forest condition [27]. Dynamic optimization is a special case that accounts for alternative future states and management decisions. 14. Multiattribute value function Yes = 10 points Allows for optimization accounting for multiple management objectives [40] 15. Considered all possible Yes = 10 points Useful for providing spatially explicit recommendations combinations of options among needed to inform stand- or forest-level forest multiple spatial units management [27] (III) Application (F) Use of stakeholder input 16. Objectives Yes = 10 points It is important that modeled objectives are relevant to stakeholder wishes and concerns [27]. 17. Management options Yes = 10 points It is important that the modeled management options are relevant to forest managers [27]. 18. Predictive model structure and/or Structure = 5 points; parameterization Incorporating stakeholder input on model structure parameterization = 5 points; both = 10 points and parameters increases uptake and use of QDS [13]. 19. Objective function structure Structure = 5 points; parameterization It is important that the functional form of the objective and/or parameterization = 5 points; both = 10 points function represents stakeholder wishes and concerns [27]. (G) Practice (and learning) 20. Management recommendations 0 points = no recommendations; 1–10 Concise management recommendations are important to based on output from QDS points depending on the conciseness: ensure that forest managers can use the output from QDS. least = 1 point, most = 10 points 21. Number of real-world 1 = 5; 2+ = 10 Evaluating the applicability of the model, considering management cycles to which the the iterativeness of forest management practices [27] QDS-based recommendations were applied 22. Demonstrated theoretical updating Yes = 10 points Important to show how new information can be used of model parameters for adaptive to update the model and ultimately informs iterative management forest management decisions and learning [41] 23. Reported real-world updating and Yes = 10 points Represents actual implementation of QDS in on-ground adapting of management actions adaptive management Table 3 Numbers of case studies where ecosystem services (ESs) were modeled to inform adaptive forest management 2005–2015. ES classification is based on CICES [35] version 4.3 Section Division Group Number of case studies Example services Provisioning Materials Biomass 20 Providing timber, pulpwood [48] Provisioning Nutrition Biomass 3 Providing understory mushrooms [49] Provisioning Nutrition Water 1 Providing drinking water [50] Regulation Maintenance of physical, chemical, Lifecycle, habitat, genetic 9 Persistence of birds and mammals [51] biological conditions Regulation Maintenance of physical, chemical, Atmosphere and climate 3 Regulating CO [52] biological conditions Regulation Maintenance of physical, chemical, Pests and diseases 1 Regulating gypsy moth biological conditions (Lymantria dispar) population [53] Regulation Maintenance of physical, chemical, By natural chemical and 1 Regulating fire [54] abiotic conditions physical processes Cultural Human interactions with the Physical and experiential 1 Recreation access [55] environment Curr Forestry Rep (2018) 4:111–124 117 Temporal Trends found that five of six case studies with an overall score ≥ 5 modeled multiple ecosystem services and used stochastic dy- Basedontheattributes of the selected case studies, the overall namic programming to evaluate management options. score of AFM declined by 35% from 4.5 ± 0.7 to 2.9 ± 0.7 between 2005 and 2015 (Fig. 1). The score of the modeling Geographic Patterns task also declined during this period by a similar amount from 4.8 ± 0.7 to 3.3 ± 0.8. We detected no other statistically signif- Case studies were either globally relevant (n = 6) or focused icant temporal trends in scores for the remaining tasks nor sub- on forest management in one of the following four continents: tasks of adaptive management, although the parameter for the Asia (n = 2), Australia (n = 7), Europe (n = 9), or North year effect on score was negative for all subtasks. Notably, the America (n = 10). Place-specific case studies each covered number of case studies published per year (range 1 to 5) neither one (n = 22), two (n = 2) or three (n = 1) of seven climatic increased nor decreased significantly. Scores for prediction and zones, including temperate (n = 20) followed by cold (n =5), management options had the greatest interannual variability (0 arid-steppe (n = 2), and tropical-rainforest zones (n =1; to 9), compared to the other subtasks (range of ca. 5). We also Online Resource 4). With the exception of two papers focused on China and Costa Rica, respectively, place-specific case studies focused on forest management in developed countries (Online Resource 4). All place-specific case studies focused on forest management within a single country, except for one that focused on forests in parts of the USA and Canada [55]. The overall score of adaptive management did not vary significantly among the three continents having ≥ 7casestud- ies (i.e., Australia, Europe, and North America). When exam- ining the raw average scores, none of the average scores for a given geography exceeded those of all other geographies across all subtasks (Fig. 2). Subtask scores for Asia were less than or similar to those of the other continents, except for the management options task. Australia had similar or greater scores for each subtask, except for optimization. Compared to other geographies, case studies in Australia and North America had higher scores for prediction, stakeholder in- volvement, and practice. Scores were most similar for these two continents. Globally relevant case studies had a higher optimization score compared to place-specific case studies. Variation within and among Tasks When comparing scores among the three tasks, the application task (stakeholder involvement and practice) score (1.8 ± 0.7) was nearly 25% less than those of the modeling or setup tasks (4.0 ± 0.7 or 4.7 ± 0.7, respectively). When comparing sub- tasks, the scores for objectives, options, and predictions exceeded those for uncertainty, stakeholder, and implement (Fig. 3). Except for uncertainty, which had a lower score, the subtask scores were balanced between modeling (predictions and optimization) and the conceptual setup tasks (objectives and options). Average scores for individual attributes ranged from 0.3 to 7.4, and we detected significant pairwise contrasts between attributes within each subtask except for uncertainty and Fig. 1 Changes in scores of overall quality of quantitative decision stakeholder (Fig. 4). Within the objectives subtask, the aver- support (a)and qualityofprediction (b) to inform adaptive forest age score for temporal horizon exceeded that of ES classes management in 34 case studies published 2005–2015. The solid line is (attributes 3 vs. 1). Considering the options subtask, the aver- the best fit to the data based on a linear regression, and dashed lines represent the 95% confidence limits age score for management intervals was greater than that of 118 Curr Forestry Rep (2018) 4:111–124 Fig. 2 Scores for 34 case studies (published 2005–2015) using quantitative decision support to inform adaptive management by subtask among focal continents (or global focus) stand level (attributes 5 vs. 4). In the predictions subtask, the scoring attributes. Within the practice subtask, the average average score for model visualize (attribute 11) was lower score for management recommendation (attribute 23) than both stochastic and parameterize (attributes 10 and 12), exceeded that of the remaining attributes (20 through 22), and there was no significant difference between these higher which included applied QDS, applied model update, and dem- onstrated model update. Cluster Analysis The majority of indices indicated that two clusters are optimal when considering the variation in scores among subtasks (Online Resource 5). Average overall score did not differ sig- nificantly between the two clusters, but differences were ob- served when comparing the mean scores for subtasks in the exploratory analysis (Fig. 5). Compared to cluster 2, cluster 1 had higher scores for stakeholder (6.0 ± 1.1 vs. 0.5 ± 0.8) and prediction (9.2 ± 1.4 vs. 4.5 ± 1.1) but a lower score for opti- mize (2.6 ± 1.4 vs. 5.1 ± 1.1). Discussion We developed a multimetric index of quality of applications of quantitative decision support to inform adaptive management and applied this through an intensive structured literature review of case studies on adaptive forest management. The index pro- Fig. 3 Scores for quality of case studies using quantitative decision vides not only a total score for a given case study but also scores support to inform adaptive forest management by subtask in each of the for each of the main tasks, subtasks, and attributes of developing 34 case studies published 2005–2015. Dots are averages, whiskers are and applying QDS to inform AFM. This way the index measures 95% confidence intervals, and differing letters between means indicate a significant difference the multiple dimensions of success of specific applications and Curr Forestry Rep (2018) 4:111–124 119 Optimize Stakeholder (SH) Practice Objectives Options Uncertainty Predictions Fig. 4 Scores for quality of attributes within 7 subtasks (bold x-axis published 2005–2015. Dots are averages, whiskers are 95% confidence labels) and 3 tasks (gray boxes) of quantitative decision support to intervals, and differing letters between means within a subtask indicate a inform adaptive forest management in each of the 34 case studies significant pairwise difference can reveal important contrasts and synergies among these tasks to the poorly developed QDS). Our multimetric index is designed as they contribute to quality of the application as a whole. For to capture such contrasts within a given case study. We therefore example, AFM could be implemented based on a poorly devel- argue it is important to consider task-specific scores in addition to oped QDS, and therefore, the effort could be considered success- total scores to have a more complete picture of QDS for ful at implementation but having a low potential for learning (due informing adaptive management. Fig. 5 Scores for quality of quantitative decision support to inform adaptive forest management by subtask in each of the 34 case studies published 2005–2015. Radar graph depicts average scores for each cluster identified by cluster analysis, average scores across all case studies, and scores for individual case studies with the highest and lowest scores across all attributes 120 Curr Forestry Rep (2018) 4:111–124 Temporal Trends Variation Within and Among Tasks We discovered that over the past decade, overall quality of Ecosystem services relevant to management objectives QDS to inform AFM declined by 33% from 4.5 to 3.0 of 10, being modeled in the selected case studies lacked diver- which supports the temporal hypothesis. Quality of modeling sity, as the strong majority of case studies have focused (especially optimization and prediction) associated with QDS on biomass as a provisioning service. Case studies fo- for AFM has also declined during this period, while metrics cusing on regulating and especially cultural services are related to uncertainty have remained low (< 5 of 10). Declines needed to demonstrate the potential of QDS to inform in quality of modeling within QDS for AFM have not been AFM addressing a broader suite of ESs, which is rele- balanced with improvements in other tasks, for which there vant for the United Nations Sustainable Development are at least three possible explanations. First, the decline may Goals [37]. be related to reduced per capita funding for developing QDS We confirmed the expectation that there has been a for AFM. This hypothesis stems from observations of stable to substantial lack of stakeholder engagement during QDS declining funding available for environmental research in gen- development tasks and on-ground implementation of eral within the USA [56, 57], along with an increasing number adaptive forest management. Aligning with the reviews of publications in the field of adaptive management [4� ]. As a of adaptive natural resource management [3� , 5], report- result of this decoupling, environmental research institu- ed on-ground application of QDS to inform AFM was tions and researchers may have shrunken the time and re- rare (2 out of the 34 case studies). We were therefore sources they devote toward application-oriented modeling unable to examine relationships between successful im- activities. Another hypothesis is that the per capita funding plementation and other tasks such as stakeholder in- for research supporting adaptive management has volvement. The question of which factors are most lim- remained stable, but researchers in this area are developing iting for successful implementation of adaptive forest models that are increasingly complex but less comprehen- management remains an important area of future work, sive when it comes to developing QDS to inform AFM. A and examining patterns in scores within our multimetric third hypothesis is that there may be a growing publication index helps to generate hypotheses about the limiting bias toward publishing novel tools and approaches, which factors. For example, the spatial scale of management has pushed researchers to publish early iterations of QDS options considered by the reviewed case studies was rather than invest in demonstrating actual implementation. often broader than stand level, which could limit the Whatever the mechanism, our findings indicate that contem- real-world application of QDS. Furthermore, many case porary QDS to inform AFM is less intensive and more qualita- studies lacked a visualization of the structure of the tive than their predecessors. We argue that this comes at a cost modeling framework(s) embedded within the QDS. of less transparent decision-making for natural resource man- Successful uptake of QDS by forest managers likely agement and undermines the ability of researchers to contribute depends on their understanding of it, which would be a critical task for adaptive management. Despite a large interest supported by clear diagrams illustrating the logic and in adaptive management, as evidenced by many review papers key parameters. on the subject (10 published 2006–2014; cited in the first par- Our finding that stakeholder involvement and practice agraph of the BIntroduction^), we hypothesize that research occurs in a minority of decision support applications is funding priorities have drifted away from comprehensive supported by previous literature reviews (e.g., [5]). Of the modeling needed to inform adaptive management. Such a shift 108 case studies on adaptive comanagement reviewed by could have been spurred by a belief that rare implementation of Plummer et al. [8], only 17 were focused on forest man- adaptive management is the fault of the approach itself rather agement. They found that stakeholder participation was the than how QDS is developed to inform it [10]. As such, the most commonly cited factor contributing to the potential underlying mechanism for this hypothesized decline in per success of adaptive comanagement, whereas this was one capita funding may be that the number of such researchers in of the less commonly cited factors contributing to actual this field has increased faster than total funding for such efforts. success. They also reported that conflict of interest among We are unaware of any peer-reviewed study that has examined participating stakeholders was the most commonly cited trends in funding and researchers in the field of environmental factor leading to failures of adaptive comanagement. science (including QDS and AFM), but there is some When asked about suggested guidelines on developing de- supporting evidence from gray literature on recent trends in cision support systems to inform forest management, funding for environmental science in the USA [56, 57]. We stakeholder involvement was one of the most frequently are unaware of any database that tracks funding nor numbers voted suggestions according to experts as part of another of researchers in the realm of QDS or AFM, although this study [58]. Although caution must be taken when applying would be a useful tool for future research. adaptive management in contexts with stakeholder Curr Forestry Rep (2018) 4:111–124 121 conflicts, resolution of conflicts was the most commonly resources may be limiting the merging of these schools of cited outcome of these processes. They also found that adaptive management (sensu [1]) to enable both high- learning was the most common emergent property of adap- quality optimization and high-quality application to over- tive comanagement. These findings highlight the impor- come the current ceiling of moderate-quality QDS to in- tance of future research that compares and evaluates ap- form AFM. Additional resources could allow for a new proaches to stakeholder involvement in adaptive manage- school of adaptive management that mitigates these ment with respect to outcomes for learning and manage- tradeoffs in a more balanced approach that improves qual- ment objectives. ity of all subtasks. Our scoring method provides a trans- On a more positive note, the quality of considered man- parent way of monitoring existing and emerging schools of agement options, predictive modeling, and optimization in adaptive management. QDS has been high in many cases. There has also been a general balance between the quality of the conceptual set- Bayesian Decision Networks up task and modeling task of QDS to support AFM. Although examples of implementation were lacking, sev- We argue that Bayesian decision networks (i.e., Bayesian belief eral (n = 5) of the selected case studies demonstrated theo- networks containing at least one decision node and one utility retical updating of model parameters for adaptive manage- node) [15] are useful yet underused for QDS. A Bayesian de- ment. Such illustrations can facilitate the use of QDS to cision network (BDN) is a parameterized influence diagram (i.e., inform actual implementation of AFM. visual representation of a predictive model that links manage- ment objectives to decision alternatives and possibly factors be- Geographic Variation yond direct control of managers), allowing for explicit represen- tation of uncertainty in the relationships between factors that are Our ability to evaluate the geographic hypothesis was limited, as usually represented by conditional probability tables [15 ]. The none of the reviewed case studies were in developing regions. expanded use of BDNs could form the basis of an emerging This finding is consistent with a literature review on adaptive school of thought that would bring together the more comanagement, which included case studies that incorporated application-focused resilience school with the more learning by multiple actors collaborating on implementing adap- optimization-focused decision-theoretic school of adaptive man- tive management in an environmental context [8]. In contrast agement. We postulate that harmonizing these two schools of with our sample that only included one case study from Asia, thought could help to slow or reverse the declining quality of published case studies on adaptive comanagement across all QDS to inform AFM. Another advantage is that BDNs can be ecosystem types (published between years 2000 and 2010) were designed in a way that is intuitive for nontechnical stakeholders relatively common in Asia, some were in Africa and South to visualize and understand how their inputs are used to form the America, and a small minority was in Australia. Further work structure and content of QDS that informs decision-making is needed to understand and address geographic gaps of pub- ([59]; B.J. Mattsson, unpublished data). lished case studies using QDS to inform AFM. Despite the known advantages of this tool, only one of the Based on our analysis, applications of QDS to inform AFM 32 case studies explicitly used a BDN as part of QDS [15�� ]. in Australia and North America were more successful with As with any computer-based decision support method, using a practice compared to those in other continents. By contrast, BDN carries the risk of a management team getting bogged case studies from Asia were strong in defining management down in planning and computing rather than on-ground action options. This could provide a basis for intercontinental work- and learning. BDNs do, however, provide a user-friendly and ing groups to exchange knowledge and support improved ap- customizable platform for planning and learning in the context plications of QDS to inform implementation of AFM. of adaptive management that can be applied to many possible contexts within and beyond forests [15�� , 60, 61]. There are Cluster Analysis many applications that users can use to construct simple BDNs free of charge (e.g., Netica, Norsys Software Corp, Our exploratory cluster analysis revealed two clusters of Vancouver, British Columbia, Canada; http://www.norsys. case studies. One cluster performed relatively well on op- com). While having the ability to serve on its own as QDS, timization, but it performed relatively poorly on prediction a BDN can also be embedded in more complex decision and stakeholder involvement. This cluster fits well under support systems that include a geographic information the decision-analytic school of adaptive management, with system or that offer multicriteria decision analysis tools (e.g., a greater emphasis on modeling compared to application Ecosystem Management Decision Support; http://emds. and practice. Despite their differences in quality of sub- mountain-viewgroup.com/explore). A BDN can also serve tasks, the two clusters were similar in terms of overall as a starting point for dynamic algorithms to inform iterative decision-making [6, 62]. Increasing interest and demand index of quality. An emergent hypothesis, then, is that �� �� 122 Curr Forestry Rep (2018) 4:111–124 Open Access This article is distributed under the terms of the Creative should lead to a community of practice that will allow for Commons Attribution 4.0 International License (http:// the development of open-source BDNs that are freely creativecommons.org/licenses/by/4.0/), which permits unrestricted use, available for more widespread and advanced use (see for distribution, and reproduction in any medium, provided you give example [38]). Four case studies used a Bayesian belief appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. network (i.e., BDN without decision or utility nodes) [6, 47��, 59, 63], and 10 presented an influence diagram to represent a stochastic optimization (e.g., [64]). Extending References these existing applications to specifying a BDN would re- quire a rather small additional time investment, with a dis- Papers of particular interest, published recently, have been tinct advantage of providing a more transparent means of highlighted as: linking QDS to real-world decision-making. � Of importance �� Of major importance Conclusion 1.�� Williams BK, Brown ED. Adaptive management: from more talk to real action. Environ Manag. 2014;53(2):465–79. https://doi.org/10. Our study represents the first comprehensive assessment of the 1007/s00267-013-0205-7. The authors demonstrate a quality of QDS to inform adaptive management, and our frame- framework for adaptive management and point out needs of work is extensible such that it can be applied to decision con- balancing technical specificity with stakeholder involvement texts outside of forest management. We have also demonstrated while addressing challenges of sustainable development and maintaining ecosystem services. for the first time how cluster analysis and radar charts can be 2. Greig LA, Marmorek DR, Murray C, Robinson DCE. Insight into useful for exploring patterns according to the seven subtasks of enabling adaptive management. Ecol Soc 2013;18(3). https://doi. QDS development and application for adaptive management. org/10.5751/es-05686-180324. We found these techniques to be useful for generating new 3.� Westgate MJ, Likens GE, Lindenmayer DB. Adaptive management hypotheses about how the multidimensional quality of QDS of biological systems: a review. Biol Conserv. 2013;158(Supplement C):128–39. https://doi.org/10.1016/j. for AFM has evolved and might continue to vary over time biocon.2012.08.016. Based on a structured review of literature and space. This approach would be useful to inform future through 2011 on adaptive management for biodiversity and research agendas to ensure that key knowledge gaps are being ecosystem integrity, the authors provide suggestions for filled, for example addressing how QDS can better integrate improvement such as better collaboration between scientists and managers and better communicating risks of non- stakeholder input and be implemented on the ground as part adaptive approaches. of AFM. An expanded literature review would enable sufficient 4.� McFadden JE, Hiller TL, Tyre AJ. Evaluating the efficacy of adap- sample sizes to formally evaluate these hypotheses. This further tive management approaches: is there a formula for success? J work is important to quantify relationships between scores for Environ Manag. 2011;92(5):1354–9. https://doi.org/10.1016/j. jenvman.2010.10.038. Based on a structured literature review actual implementation and tasks related to QDS development, (2000–2009), the authors examine temporal trends in which will provide insight into the usefulness of the index to frequency of publications and degrees of success originating examine drivers of successful implementation of AFM. To min- from two adaptive management schools of thought. imize bias, scoring of attributes should be conducted indepen- 5. Fabricius C, Cundill G. Learning in adaptive management: insights from published practice. Ecol Soc. 2014;19(1). https://doi.org/10. dently by multiple experts and then discussed before assigning 5751/es-06263-190129. a final score. Another area of future research is a controlled 6. Rumpff L, Duncan DH, Vesk PA, Keith DA, Wintle BA. State-and- experiment where adaptive management teams are assigned transition modelling for adaptive management of native woodlands. restrictions as to the levels of emphasis on particular subtasks Biol Conserv. 2011;144(4):1224–36. https://doi.org/10.1016/j. (e.g., stakeholder involvement and predictive modeling). We biocon.2010.10.026. 7. Williams BK. Adaptive management of natural resources—frame- encourage any research that critically examines what works work and issues. J Environ Manag. 2011;92(5):1346–53. https:// and what does not work in the field of QDS and adaptive doi.org/10.1016/j.jenvman.2010.10.041. management, so that we can more quickly learn and achieve 8. Plummer R, Crona B, Armitage DR, Olsson P, Tengö M, Yudina O. conservation and natural resource management goals. Adaptive comanagement: a systematic review and analysis. Ecol Soc. 2012;17(3):11. https://doi.org/10.5751/ES-04952-170311(3):11. Funding Information Open access funding provided by University of 9. Gregory R, Ohlson D, Arvai J. Deconstructing adaptive manage- Natural Resources and Life Sciences Vienna (BOKU). ment: criteria for applications to environmental management. Ecol Appl. 2006;16(6):2411–25. https://doi.org/10.1890/1051- 0761(2006)016[2411:damcfa]2.0.co;2. Compliance with Ethical Standards 10. Allen CR, Fontaine JJ, Pope KL, Garmestani AS. Adaptive man- agement for a turbulent future. J Environ Manag. 2011;92(5):1339– Human and Animal Rights and Informed Consent This article does not 45. https://doi.org/10.1016/j.jenvman.2010.11.019. contain any studies with human or animal subjects performed by any of 11. Borges JG, Nordström E-M, Garcia Gonzalo J, Hujala T, the authors. Trasobares A. Computer-based tools for supporting forest Curr Forestry Rep (2018) 4:111–124 123 management. 2014. http://pub.epsilon.slu.se/id/eprint/11417. 23. Yousefpour R, Jacobsen JB, Thorsen BJ, Meilby H, Hanewinkel M, Oehler K. A review of decision-making approaches to handle un- Accessed 2 April 2018. certainty and risk in adaptive forest management under climate 12. Muys B, Hynynen J, Palahi M, Lexer MJ, Fabrika M, Pretzsch H, et change. Ann For Sci. 2012;69(1):1–15. https://doi.org/10.1007/ al. Simulation tools for decision support to adaptive forest manage- s13595-011-0153-4. ment in Europe. For Sys. 2010;19:86–9. https://doi.org/10.5424/fs/ 24. Whiteman A, Wickramasinghe A, Pina L. Global trends in forest 201019S-9310. ownership, public income and expenditure on forestry and forestry 13.�� McIntosh BS, Ascough JC, Twery M, Chew J, Elmahdi A, Haase employment. For Ecol Manag. 2015;352:99–108. https://doi.org/ D, et al. Environmental decision support systems (EDSS) develop- 10.1016/j.foreco.2015.04.011. ment—challenges and best practices. Environ Model Softw. 25. Radke N, Yousefpour R, von Detten R, Reifenberg S, Hanewinkel 2011;26(12):1389–402. https://doi.org/10.1016/j.envsoft.2011.09. M. Adopting robust decision-making to forest management under 009. As authors of the paper, 27 developers of environmental climate change. Ann For Sci. 2017;74(2):43. https://doi.org/10. decision support systems (EDSSs) identified ways of overcom- 1007/s13595-017-0641-2. ing challenges of EDSS adoption and uptake such as the use of a 26. Pasalodos-Tato M, Mäkinen A, Garcia-Gonzalo J, Borges JG, champion to promote the EDSS within target user groups, en- Lämås T, Eriksson LO. Review. Assessing uncertainty and risk in sure EDSS elements are easy and inexpensive to use and up- forest planning and decision support systems: review of classical date, build capacity within target groups to ensure they can methods and introduction of new approaches. For Sys. 2013;22(2): efficiently use the EDSS, and develop metrics of success of 282–303. https://doi.org/10.5424/fs/2013222-03063. EDSS applications. 27. Bettinger P, Boston K, Siry JP, Grebner DL. Forest management 14. Valls-Donderis P, Ray D, Peace A, Stewart A, Lawrence A, Galiana and planning. New York: Academic; 2009. F. Participatory development of decision support systems: which 28. Moore CT, Conroy MJ. Optimal regeneration planning for old- features of the process lead to improved uptake and better out- growth forest: addressing scientific uncertainty in endangered spe- comes? Scand J For Res. 2014;29:71–83. https://doi.org/10.1080/ cies recovery through adaptive management. For Sci. 2006;52(2): 02827581.2013.837950. 155–72. 15.�� Nyberg JB, Marcot BG, Sulyma R. Using Bayesian belief networks 29. Wikström P, Edenius L, Elfving B, Eriksson LO, Lämås T, in adaptive management. Can J For Res. 2006;36(12):3104–16. Sonesson J, et al. The heureka forestry decision support system: https://doi.org/10.1139/x06-108. Through general arguments and an overview. Math Comput For Nat Res Sci. 2011;3(2):87. a case study, this paper provides a clear demonstration of how 30. forestDSS. List of forest decision support systems. 2017. http:// influence diagrams and BDNs can be useful to inform formal www.forestdss.org/wiki/index.php?title=Category:DSS. Accessed adaptive natural resource management in collaborations 2 April 2018. between scientists and stakeholders. 31. Segura M, Ray D, Maroto C. Decision support systems for forest 16. Lawrence A, Stewart A. Sustainable forestry decisions: on the in- management: a comparative analysis and assessment. Comput terface between technology and participation. Math Comput For Electron Agric. 2014;101:55–67. https://doi.org/10.1016/j. Nat Res Sci. 2011;3(1):42. compag.2013.12.005. 17. Blattert C, Lemm R, Thees O, Lexer MJ, Hanewinkel M. 32. Keenan RJ. Climate change impacts and adaptation in forest man- Management of ecosystem services in mountain forests: review agement: A review. Annals of Forest Science. 2015;72(2):145–67. of indicators and value functions for model based multi-criteria https://doi.org/10.1007/s13595-014-0446-5. decision analysis. Ecol Indic. 2017;79:391–409. https://doi.org/ 33. Collaboration for Environmental Evidence. Guidelines for system- 10.1016/j.ecolind.2017.04.025. atic review and evidence synthesis in environmental management. 18. Irauschek F, Rammer W, Lexer MJ. Evaluating multifunctionality Version 4.2. 2013. http://www.environmentalevidence.org/wp- and adaptive capacity of mountain forest management alternatives content/uploads/2014/06/Review-guidelines-version-4.2-final.pdf. under climate change in the eastern alps. Eur J For Res. Accessed 2 April 2018. 2017;136(5–6):1051–69. https://doi.org/10.1007/s10342-017- 34. Chai J, Liu JNK, Ngai EWT. Application of decision-making tech- 1051-6. niques in supplier selection: a systematic review of literature. 19. Mina M, Bugmann H, Cordonnier T, Irauschek F, Klopcic M, Expert Syst Appl. 2013;40(10):3872–85. https://doi.org/10.1016/j. Pardos M, et al. Future ecosystem services from european mountain eswa.2012.12.040. forests under climate change. J Appl Ecol. 2017;54(2):389–401. 35. Common International Classification of Ecosystem Services https://doi.org/10.1111/1365-2664.12772. (CICES). Cices v4.3 spreadsheet. 2016. https://cices.eu/ 20. Vacik H, Kurttila M, Hujala T, Khadka C, Haara A, Pykäläinen J, et resources/. Accessed 2 April 2018. al. Evaluating collaborative planning methods supporting 36. Cubbage F, Harou P, Sills E. Policy instruments to enhance multi- programme-based planning in natural resource management. J functional forest management. Forest Policy Econ. 2007;9(7):833– Environ Manag. 2014;144:304–15. https://doi.org/10.1016/j. 51. https://doi.org/10.1016/j.forpol.2006.03.010. jenvman.2014.05.029. 37. UN General Assembly. Transforming our world: the 2030 agenda 21. Acosta M, Corral S. Multicriteria decision analysis and participato- for sustainable development. Report no. A/res/70/1. 2015. https:// ry decision support systems in forest management. Forests. sustainabledevelopment.un.org/post2015/transformingourworld/ 2017;8(4) https://doi.org/10.3390/f8040116. publication. Accessed 2 April 2018. 38. Seidl R, Lexer MJ. Forest management under climatic and social 22.�� Vacik H, Lexer MJ. Past, current and future drivers for the devel- opment of decision support systems in forest management. Scand J uncertainty: trade-offs between reducing climate change impacts For Res. 2014;29:2–19. https://doi.org/10.1080/02827581.2013. and fostering adaptive capacity. J Environ Manag. 2013;114:461–9. Based on a review of development of decision support 39. Collins M, Knutti R, Arblaster J, Dufresne J-L, Fichefet T, sy Friedlingstein P et al. Long-term climate change: projections, com- stems (DSSs) to inform forest management, the authors sug- mitments and irreversibility. 2013. gest that evolving toolkits of decision-support approaches rath- 40. Huang IB, Keisler J, Linkov I. Multi-criteria decision analysis in er than stand-alone DSSs can help balance wishes for scientific environmental sciences: ten years of applications and trends. Sci rigor with diverse demands by forest managers and policy- Total Environ. 2011;409(19):3578–94. https://doi.o makers for concise and user-friendly tools that fit within their rg/10.1016/j. normal work routine. scitotenv.2011.06.022. 124 Curr Forestry Rep (2018) 4:111–124 41. Yousefpour R, Jacobsen JB, Meilby H, Thorsen BJ. Knowledge 53. Bogich T, Shea K. A state-dependent model for the optimal man- agement of an invasive metapopulation. Ecol Appl. 2008;18(3): update in adaptive management of forest resources under climate change: a Bayesian simulation approach. Ann For Sci. 2014;71(2): 748–61. https://doi.org/10.1890/07-0642.1. 301–12. https://doi.org/10.1007/s13595-013-0320-x. 54. Chades I, Bouteiller B. Solving multiagent markov decision pro- 42. Walters CJ. Adaptive management of renewable resources. New cesses: a forest management example. Modsim 2005: International York: Macmillan; 1986. Congress on Modelling and Simulation: Advances and 43. Lee KN. Chapter 1. Appraising adaptive management. Biological Applications for Management and Decision Making. 2005:1594– diversity: balancing interests through adaptive collaborative man- agement. Boca Raton: CRC; 2001. p. 1–26. 55. Tomberlin D. Endangered seabird habitat management as a partially 44. Rist L, Campbell BM, Frost P. Adaptive management: where are we observable markov decision process. Mar Resour Econ. now? Environ Conserv. 2012;40(1):5–18. https://doi.org/10.1017/ 2010;25(1):93–104. S0376892912000240. 56. Jahnke A. Who picks up the tab for science? 2017. http://www.bu. 45. Charrad M, Ghazzali N, Boiteau V, Niknafs A. Nbclust: an R pack- edu/research/articles/funding-for-scientific-research/. Accessed 2 age for determining the relevant number of clusters in a data set. J April 2018. Stat Softw. 2014;61:1–36. 57. Neubauer SJ. Foundation funding trends for environmental & con- 46. R Core Team. R: a language and environment for statistical com- servation efforts 2015. https://www.calandtrusts.org/wp-content/ puting. Vienna: R Foundation for Statistical Computing. 2017. uploads/2015/03/Foundation-Funding-Trends-for-Environmental- https://www.R-project.org/. Accessed 2 April 2018. and-Conservation-Efforts.pdf. Accessed 2 April 2018. 47.�� Howes AL, Maron M, McAlpine CA. Bayesian networks and adap- 58. Marques A, Ficko A, Kangas A, Rosset C, Ferriti F, Rasinmaki J, et tive management of wildlife habitat. Conserv Biol. 2010;24(4): al. Empirical guidelines for forest management decision support 974–83. https://doi.org/10.1111/j.1523-1739.2010.01451.x. The systems based on the past experiences of the expert’s community. authors provide a case study demonstrating how an influence Forest Systems. 2013;22:320–39. diagram and Bayesian belief network are constructed based on 59. Marcot BG, Hohenlohe PA, Morey S, Holmes R, Molina R, Turley inputs from land managers and then updated based on new MC, Huff MH, Laurence JA. Characterizing species at risk II: using data to inform adaptive management of feral grazing and Bayesian belief networks as decision support tools to determine prescribed burning regimes to maintain avian diversity within species conservation categories under the northwest forest plan. woodlands of subtropical eastern Australia. Ecol Soc. 2006;11(2). 48. Yousefpour R, Temperli C, Bugmann H, Elkin C, Hanewinkel M, 60. Dalyander PS, Meyers M, Mattsson B, Steyer G, Godsey E, Meilby H, et al. Updating beliefs and combining evidence in adap- McDonald J, et al. Use of structured decision-making to explicitly tive forest management under climate change: a case study of incorporate environmental process understanding in management Norway spruce (Picea abies l. Karst) in the Black Forest, of coastal restoration projects: case study on barrier islands of the Germany. J Environ Manag. 2013;122:56–64. https://doi.org/10. Northern Gulf of Mexico. J Environ Manag. 2016;183:497–509. 1016/j.jenvman.2013.03.004. https://doi.org/10.1016/j.jenvman.2016.08.078. 49. Ferreira L, Constantino MF, Borges JG, Garcia-Gonzalo J. A sto- 61. Thorne KM, Mattsson BJ, Takekawa J, Cummings J, Crouse D, chastic dynamic programming approach to optimize short-rotation Block G, et al. Collaborative decision-analytic framework to max- coppice systems management scheduling: an application to euca- imize resilience of tidal marshes to climate change. Ecol Soc. lypt plantations under wildfire risk in Portugal. For Sci. 2012;58(4): 2015;20(1):30. https://doi.org/10.5751/es-07018-200130. 353–65. https://doi.org/10.5849/forsci.10-084. 62. Golovin D, Krause A, Gardner B, Converse SJ, Morey S. Dynamic 50. Spring DA, Kennedy JOS, Mac Nally R. Optimal management of a resource allocation in conservation planning. Proceedings of the forested catchment providing timber and carbon sequestration ben- twenty-fifth AAAI conference on artificial intelligence. 2011. efits: climate change effects. Glob Environ Chang. 2005;15(3): http://www.aaai.org/ocs/index.php/AAAI/AAAI11/paper/ 281–92. https://doi.org/10.1016/j.gloenvcha.2005.04.002. download/3617/4079/. Accessed 2 April 2018. 51. Sabbadin R, Spring D, Bergonnier E. A reinforcement-learning 63. Rumpff L, Duncan D, Vesk P, Wintle B. Development of a state- application to biodiversity conservation in costa-rican forest. and-transition model to guide investment in woodland vegetation Modsim 2007: International Congress on Modelling and condition. Landscape Logic: Integrating Science for Landscape Simulation: Land, Water and Environmental Management: Management. 2012:157–72. Integrated Systems for Sustainability. 2007:2189–95. 52. Haim D, Plantinga AJ, Thomann E. The optimal time path for 64. Shahi S, Pulkki R. A simulation-based optimization approach to carbon abatement and carbon sequestration under uncertainty: the integrated inventory management of a sawlog supply chain with case of stochastic targeted stock. Resour Energy Econ. 2014;36(1): demand uncertainty. Can J For Res. 2015;45(10):1313–26. https:// 151–65. https://doi.org/10.1016/j.reseneeco.2013.11.006. doi.org/10.1139/cjfr-2014-0373. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Forestry Reports Springer Journals

Gaps in Quantitative Decision Support to Inform Adaptive Management and Learning: a Review of Forest Management Cases

Download PDF
14 pages

Loading next page...
 
/lp/springer-journals/gaps-in-quantitative-decision-support-to-inform-adaptive-management-jH7WJIAOwr

References (66)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Environment; Sustainable Development; Environmental Management; Nature Conservation; Forestry; Forestry Management; Ecology
eISSN
2198-6436
DOI
10.1007/s40725-018-0078-3
Publisher site
See Article on Publisher Site

Abstract

Purpose of Review Theoretical frameworks for adaptive natural resource management are quite common, whereas documented examples showing successful implementation of adaptive management and learning through multiple time intervals have remained uncommon. Measures of quality of adaptive natural resource management processes are needed to examine potential factors driving the successful implementation. To address this gap, we developed a multimetric index composed of 22 metrics to assess quality of case studies using quantitative decision support (QDS) to inform adaptive forest management (AFM). Metrics represented three main tasks, including conceptual setup, modeling, and application. We further distinguished these into subtasks: definition of objectives and management options (setup); specifying uncertainty, prediction, and optimization (modeling); and stakeholder involvement along with practice and learning (application). We used a multimetric index to examine temporal and geographic variation in quality of reviewed case studies using QDS to inform AFM. We then conducted a structured literature review of 179 articles, wherein 34 case studies met a priori criteria. Recent Findings When applying the multimetric index to these case studies, we found that over the past decade the index has been intermediate and annual average scores declined by 33% from 4.5 to 3.0 of 10 (where 10 is the highest possible quality score). Aligning with reviews of adaptive natural resource management, reported on-ground application of QDS to inform AFM was rare (n=2). We also confirmed the expectation that there has been a substantial lack of stakeholder engagement during QDS development tasks. Summary Our multimetric index provides a novel tool to examine gaps in the use of QDS for adaptive management in diverse domains including but not limited to forests. . . . . . Keywords Adaptive management Case studies Decision support Forest management Multimetric index Stakeholder engagement Introduction Adaptive management has been recognized as a promising ap- proach to inform and iterate decision-making and promote This article is part of the Topical Collection on Forest Management learning to achieve conservation and natural resource manage- Electronic supplementary material The online version of this article ment objectives under uncertainty in diverse contexts [1, 2]. (https://doi.org/10.1007/s40725-018-0078-3) contains supplementary Although there are many ways in which adaptive management material, which is available to authorized users. has been interpreted [3� , 4� ], a common thread among disparate approaches emphasizes learning by doing rather than managing * Brady J. Mattsson brady.mattsson@boku.ac.at systems as if they were static or so unpredictably dynamic that they preclude the possibility to learn. Theoretical expositions and frameworks for adaptive natural resource management Institute of Silviculture, University of Natural Resources and Life Sciences, Peter-Jordan-Straße 82, 1190 Vienna, Austria (henceforth, adaptive management) are quite common, whereas documented examples showing successful implementation of Present address: Institute of Wildlife Biology and Game Management, University of Natural Resources and Life Sciences, adaptive management and learning through multiple cycles of Gregor-Mendel Straße, 1180 Vienna, Austria decision-making have remained uncommon [3� , 5]. Proposed University of Freiburg, 79106 Freiburg, Germany factors limiting implementation of adaptive management based �� 112 Curr Forestry Rep (2018) 4:111–124 on literature reviews are many and include level of organiza- complex system, as component indices can be examined indi- tional and financial support for the approach [6, 7], appropriate vidually and in aggregate [19]. Investigating a total index application of the approach [3� ], concise and vetted problem score can account for simultaneous yet minor shifts in multi- definition [2], stakeholder involvement in tailoring adaptive ple subcomponents that would otherwise go unnoticed if ex- management plans to local situations [5, 8], acceptance of the amined individually. On the other hand, the quality of a sys- approach by practitioners [9], communication between scien- tem as a whole may remain unchanged while the quality of tists and managers [3� , 4� ], and balance between planning and subcomponents shift in contrasting ways that cancel each oth- real-world action [10]. This leads to the question: Which factors er out. Such component dynamics could be crucial for improv- are most limiting for successful implementation of adaptive ing understanding and decision-making in natural resource management and how can they be overcome? management. Therefore, examining both total scores and sub- Viewing the adaptive management process through the lens of component scores of a multimetric index is important to ob- quantitative decision support (QDS) offers a way to examine and tain a comprehensive understanding of a multipart system. decompose this question. We define QDS as an approach to Developing and using a QDS to inform natural resource man- inform management choices by specifying ultimate management agement is a highly complex and multidimensional endeavor objectives, identifying at least two possible management strate- [13 , 21, 22 ], which warrants such a multimetric index ap- gies, and comparing the strategies using a predictive modeling proach to examining quality of the components and entirety of procedure that includes at least one quantitative variable to rep- the QDS development and application. resent the objectives. This definition complements definitions of Forests offer a particularly interesting study system for de- decision support systems in the field of forest management (e.g., veloping indicators of QDS quality for adaptive management [11, 12]), which have emphasized all-in-one computer-based sys- and formal learning. Forests are managed at multiple spatial tems that usually require quantitative inputs. QDS need not be scales (e.g., stand, forest, and landscape level; [23]) and the entirely computer-based and would therefore be suitable for ap- majority is publically owned, but forests are becoming in- plication in regions of the world where computer resources are creasingly owned by private individuals [24]. Management limited. We propose that QDS offers a transparent means to of forests is challenged due to many sources of uncertainty inform adaptive forest management (AFM) and thereby improve and long turnover rates and time lags of woody productivity learning and achievement of management objectives relative to [25, 26]. Forest management is also a diverse subject that we approaches that do not use QDS. The parameter values for the define as any decision-making process that affects conditions predictive modeling may be all theoretical, all empirical, or some in a forest. The conditions could be abiotic (e.g., temperature, combination of these as long as they are quantified. To increase precipitation, fire), biotic (e.g., tree growth, animal population the chances that QDS will inform real-world decision making, demographics), or a combination of these [27]. Forest man- stakeholders and end users should be involved throughout the agement actions can also be diverse (e.g., timber harvest, tree development process [13 ]. planting, protection against disturbance by recreationists, pro- QDS canbeusedtoinformimplementationof forestman- viding artificial nest cavities). Correspondingly, there are agement, but published examples of cases are rare [6, 14, 15]. many types of QDS, which range from more academically One possible explanation for this rarity is the lack of documen- driven dynamic optimization tools (e.g., [28]) to graphical tation rather than a low rate of true uptake of QDS for AFM, interfaces that managers can readily learn and use on their and certainly better documentation of use vs. nonuse is needed. own (e.g., Heureka; [29]). For reviews of decision support The reasons for the lack of uptake of QDS overlap with the systems available to inform forest management, see forest reasons for the lack of implementation of adaptive manage- DSS Community of Practice [30] and Segura et al. [31]. ment. Lack of sufficient involvement by stakeholders and end Existing reviews have pointed out prospects for QDS and users is a commonly cited reason for the rare uptake of QDS in formal learning to be useful to inform forest management in natural resource management (e.g., [13, 16]). Developing an the face of uncertainties about climate change [23, 32]. indicator of quality of adaptive management among tasks of Although the topic of learning through actual implementation QDS development and implementation would provide a means of adaptive management has been reviewed [5], a comparative to learn about drivers of successful implementation, but metrics assessment of the quality of decision support to inform to construct such an indicator are currently lacking. adaptive management is needed for a better understanding Recently, multimetric indicators have been developed to of the mechanisms limiting the application of adaptive quantify impacts of management activities on ecosystem func- management. tions and services [17–19], for assessing participatory devel- Our overarching objective is to identify the key gaps in the opment of decision support systems [14] and for evaluating development and application of QDS that allow for the imple- collaborative planning approaches to support natural resource mentation of AFM to improve learning and achievement of management [20]. Using multimetric indices allows for com- management objectives in a changing world. Toward this end, prehensively and consistently characterizing elements in a we develop a multimetric index of quality of applications of �� �� �� �� �� Curr Forestry Rep (2018) 4:111–124 113 QDS to inform AFM. We then conduct a structured literature optimization, < 60 focused on decision support, and < 50 fo- review to evaluate them according to the selected attributes cused on Bayesian updating. and the multimetrics index. Next, we utilize the index and In the second round of searching, we developed a final set component metrics to examine temporal and geographic var- of search keywords (Table 1), which were chosen based on iation in the use of particular methods and approaches, to our main objective, the first round of searches, and delibera- enhance understanding about any changes or patterns in the tion among coauthors. Our goal in this final search was to comprehensiveness and application of QDS to inform AFM. identify approximately 200 papers, which would be a feasible The geographic analysis enables us to identify broadscale number of papers for the coauthors to review. We also had a gaps in the application of QDS and AFM. We then discuss goal of selecting at least 30 case studies, and we believed this the evolution of the interface between QDS and the applica- would be possible from a set of 200 papers identified via the tion of AFM, along with complementary cultures of decision keyword search. When combining the final set of keywords support and forest management that could learn from each with an OR operator, 188 papers were identified. Of these other in terms of best practices. We use regression models to matching papers, nine were classified as review papers in examine a priori hypotheses, and we use an exploratory clus- the Web of Science database and were removed from our list. ter analysis to generate ideas for future research on adaptive We then read the title and abstract from each of the 179 management that complements the vast existing literature. candidate papers identified from the final set of search key- words to determine if it would be selected for inclusion in the analysis. In particular, we determined whether each candidate Methods paper included a case study that developed and described QDS in a forest management context (for definitions of We hypothesized that the quality of case studies using quan- QDS and forest management, see BIntroduction^). In some titative decision support to inform adaptive forest manage- cases, we read the candidate paper if we could not determine ment has varied across time (temporal hypothesis), among its relevance with confidence based on reading the title and global geographies (geographic hypothesis), and among and abstract alone. To examine the recent trends, we included pa- within tasks of QDS development and application (task hy- pers published between 2005 and 2015 in the final analysis. pothesis) or has been constant across time, space, and tasks (null hypothesis). To examine these hypotheses, we identified Attributes and Scoring Selected Case Studies relevant case studies in the literature and developed a method for scoring their quality. To score the quality of QDS to inform AFM in each selected case study, we identified 23 attributes comprising three main Case Study Selection tasks of adaptive management, including conceptual setup Table 1 Final search criteria used to identify relevant papers for the To ensure relevant case studies were included in the analysis literature review on quantitative decision support to inform adaptive and that a suitable number of papers were selected, we used forest management. Asterisks (*) indicate wildcard characters, quotes two rounds of literature search within the Web of Science (B) surround exact phrases, and numbers at the end of each row indicate database on 10 August 2017. In designing the selection pro- the number of papers matching that set of criteria in the Web of Science database on 16 December 2017. Unless otherwise noted with italics, cess, we followed the principles of systematic literature re- search terms were applied to the titles, abstracts, and keywords within view [33, 34], with some simplifications to keep the process the database feasible. In the first round of searching, our goal was to iden- tify a series of keywords that reflected our main objective and Search criteria Number of papers assess the number of matching papers for multiple combina- (forest* or woodland*) AND tions of keywords. Toward this end, we conducted a series of (Badaptive management^ AND learn*) OR 94 exploratory keyword searches with increasing specificity (Badaptive management^ AND Bdecision 25 (Online Resource 1). For example, adaptive forest* manage- support^)OR ment yielded 1429 matches, whereas refining these with AND (bayes* updat* AND manage*) OR 17 Bdecision support^ yielded 54 papers. We aimed for a close fit (cited paper with title or title is BUsing 13 to our main objective while keeping the search as broad as Bayesian belief networks in adaptive possible to allow for any application of adaptive forest man- management^ )OR ((Bdynamic programming^ OR Bdynamic 58 agement, not limited to any particular definitions of adaptive optimization^) AND stochastic) management [3� , 4� ]. During the first round of searches, we found that there were over 1400 papers that appeared to focus Search term was used to limit all sets of search criteria, to maintain a on adaptive forest management, and of these, < 300 focused focus on forest management on modeling, < 200 focused on learning or dynamic Ref. [15 ] �� 114 Curr Forestry Rep (2018) 4:111–124 (n = 5), modeling (n = 10), and application (n = 8; Table 2). Recognizing the importance of balanced emphasis among We further distinguished these into subtasks, including defini- attributes of tasks and the tasks themselves, we assumed equal tion of objectives and management options (setup); specifying weight among attributes within tasks and equal weight among uncertainty, prediction, and optimization (modeling); and tasks when computing scores for tasks and total scores. In stakeholder involvement along with practice and learning (ap- particular, we summed attribute scores within each task, and plication). We recognize that some subtasks can be done si- each sum was standardized on a scale from 0 to 10. Likewise, multaneously (e.g., stakeholder involvement and defining ob- we summed task scores and computed a total score that was jectives), but each subtask on its own represents an important then standardized on the 0 to 10 scale. aspect of adaptive management. We are aware of the many definitions of adaptive management (e.g., [7, 10, 42–44]), which to a large degree represent the inherent flexibility in Statistical Analysis how the approach is applied in a given context [10] but also owing to fundamental differences in approaches [4� ]. We have We used linear regression to examine our hypotheses about chosen the tiered system of tasks and subtasks to account for spatiotemporal variation and differences among tasks and sub- these diverse approaches while facilitating our analysis and to tasks regarding quality of QDS to inform adaptive forest man- help ensure that our framework could be extended to other agement, after confirming that the model assumptions were domains beyond forest management with varying levels of met (i.e., normally distributed residuals and homoscedastici- complexity. The three main tasks we identified are largely ty). Each model included the index of quality as a response consistent across approaches, whereas the subtasks within variable, along with one of the following predictor variables: modeling are more specific to a decision-theoretic school year (continuous), continent, task, subtask, or attribute. We of thought [4� ]. In addition to attributes for scoring, we used an alpha level of 0.05 for determining statistical signifi- also classified each selected case study according to the cance and Tukey’s honest significant difference test to conduct location and climatic zone (Table 2). These classifications pairwise contrasts of means. We also conducted k-means clus- were not used for the scoring, but rather to examine the ter analysis to explore similarities within groups of case stud- geographic hypothesis by comparing scores among loca- ies across scores for subtasks of adaptive management. To tions and climatic zones. determine the number of clusters for this analysis, we exam- For each attribute describing a task of adaptive manage- ined the frequency of optimal numbers of clusters across 30 ment, we assigned a score from 0 to 10 representing low to independent methods, setting the maximum possible number high quality for that attribute (Table 2). Assignments of qual- of clusters to 10 to ensure feasible interpretation [45].We used ity were subjective, and we used our collective judgment and program R for all statistical analysis [46]. logic in designing the scoring criteria. For example, stakehold- er input on the objective function was assigned 5 points if stakeholders (or literature produced by stakeholders) were consulted on the structure or parameterization. If stakeholders Results were consulted on both structure and parameterization, then 10 points were assigned. We reasoned that structure and pa- Based on a structured literature review, we selected 34 case rameters of the objective function are equally important and studies, each of which used QDS to inform adaptive forest additive in contributing to this criterion. management and was published between 2005 and 2015 The scores were readily assigned based on simple attributes (Online Resources 2 and 3). Across these case studies, there except the quality of recommendations, which was scored were 25 unique lead authors and 1–5 case studies published according to the level of conciseness of the text and graphics per year. Two of the 34 case studies reported that QDS was representing the suggested course of action given the QDS used to inform real-world AFM [15�� , 47�� ]. The majority of inputs. For example, a case study received a higher score if case studies (n = 21, 62%) modeled a single group of ecosys- it presented a clear graph showing how predicted outcomes (in tem services (ESs), and more case studies modeled two or terms of the objectives) change among management options. more groups of ESs (n = 8, 24%) than those that modeled none We also assigned a higher score for quality of recommenda- (n = 6, 18%). Provisioning services were modeled by the ma- tion if there was a clear one to two sentence summary of how jority of case studies (n = 24, 71%), followed by regulating the results could be used to inform decision-making. Such (n = 14, 41%) and cultural services (n = 1, 3%). The most concise descriptions would be interpretable by less technical frequently modeled group within provisioning services was stakeholders and decision-makers, and the quality of the rec- Bbiomass^ (n = 23), and the most frequently modeled group ommendation is an important indicator of successful knowl- within regulating services was Blifecycle, habitat, genetic^ edge transfer between science in practice in the context of (n = 9; Table 3). A cultural service (i.e., recreation access) QDS and AFM. was only modeled in one case study [55]. Curr Forestry Rep (2018) 4:111–124 115 Table 2 Attribute descriptions and scoring schemes for a multimetric the attributes (numbered) within subtasks (lettered) and tasks (Roman index applied to case studies using quantitative decision support (QDS) to numerals) of adaptive management are provided, and wherever none of inform adaptive forest management. Descriptions of points assigned to the conditions were met the score was set to 0 Attribute Scoring Description and justification Location of landscape/region Attributes used to examine geographic patterns in the multimetric index Continent(s) – Continents where the case study was applied; does not imply that the case study is relevant to entire continents Country(ies) – Countries where the case study was applied; does not imply that the case study is relevant to entire countries Climatic zone(s) – Climatic zones where the case study was applied; does not imply that the case study is relevant to entire climatic zones (I) Conceptual setup (A) Objectives 1. Group(s) of ecosystem services 1 ES = 2 points; 2 ES = 5 points; Classification based on CICES [35] version 4.3. Useful for (ES) modeled 3+ ES = 10 points characterizing diversity of objectives, acknowledging calls for multifunctional forest management [36], and attaining sustainable development goals [37] 2. Duration of temporal horizon 10–19 = 1 point; 20–39 = 2 Evaluating applicability of model, considering long for modeled objectives (years) points; … 100+=10 points turnover rates of woody productivity [25, 26] 3. Spatial extent of modeled objectives Specified = 10 points Important to specify the area to which the model has been developed when applying it to real-world forest management [27] (B) Management options 4. Number of time intervals during which 1 point per for each interval 2–10; Evaluating the applicability of the model, considering actions can change for a given 11+ intervals = 10 points the iterativeness and adaptability of forest management unit management practices [27] 5. Specified size of smallest area within Specified = 10 points Evaluating the applicability of the model, recognizing which actions were modeled that forest managers require spatial specificity to implement recommendations [27] (II) Modeling (C) Uncertainty 6. Modeled climate scenario(s) 1 scenario = 2 points; 2 scenarios = 5 Important consideration recognizing strong influences points; 3+ scenarios = 10 points of climatic variables on forest dynamics and interactions with forest management [38] 7. Multiple relationships considered Yes = 10 points Important to consider given the high level of uncertainty when modeling effects of climate about future climate [39]. See also justification for variables BModeled climate scenario(s)^ 8. Multiple relationships considered when Yes = 10 points Important to consider given uncertainties about modeling effects of management options long-term outcomes of forest management strategies [27, 38] 9. Multiple relationships considered 1 uncertainty = 2 points; 2 Important to consider given uncertainties about when modeling effects of factors other uncertainties = 5 points; 3+ influence of other factors having strong influence than climate or management effects uncertainties = 10 points on forest management, e.g., timber prices [27]. (D) Predicting consequences of management options 10. Specified probability distributions Yes = 10 points Useful for explicitly incorporating uncertainty in a (stochasticity)aspartofpredictive predictive model and can be updated using Bayes’ modeling theorem [15�� ] 11. Displayed model as a box-and-arrow Yes = 10 points Visualizations allow forest managers and other relevant diagram stakeholders to understand and provide input to the model structure and parameters [15�� ]. 12. Model parameterized Yes = 10 points Populating model with quantitative information allows for transparent forest management recommendations [27]. 116 Curr Forestry Rep (2018) 4:111–124 Table 2 (continued) Attribute Scoring Description and justification (E) Optimization 13. Type of optimization Nondynamic optimization = 5 points; Optimization allows for management recommendations dynamic optimization = 10 points that are explicit for predefined states of the system, e.g., forest condition [27]. Dynamic optimization is a special case that accounts for alternative future states and management decisions. 14. Multiattribute value function Yes = 10 points Allows for optimization accounting for multiple management objectives [40] 15. Considered all possible Yes = 10 points Useful for providing spatially explicit recommendations combinations of options among needed to inform stand- or forest-level forest multiple spatial units management [27] (III) Application (F) Use of stakeholder input 16. Objectives Yes = 10 points It is important that modeled objectives are relevant to stakeholder wishes and concerns [27]. 17. Management options Yes = 10 points It is important that the modeled management options are relevant to forest managers [27]. 18. Predictive model structure and/or Structure = 5 points; parameterization Incorporating stakeholder input on model structure parameterization = 5 points; both = 10 points and parameters increases uptake and use of QDS [13]. 19. Objective function structure Structure = 5 points; parameterization It is important that the functional form of the objective and/or parameterization = 5 points; both = 10 points function represents stakeholder wishes and concerns [27]. (G) Practice (and learning) 20. Management recommendations 0 points = no recommendations; 1–10 Concise management recommendations are important to based on output from QDS points depending on the conciseness: ensure that forest managers can use the output from QDS. least = 1 point, most = 10 points 21. Number of real-world 1 = 5; 2+ = 10 Evaluating the applicability of the model, considering management cycles to which the the iterativeness of forest management practices [27] QDS-based recommendations were applied 22. Demonstrated theoretical updating Yes = 10 points Important to show how new information can be used of model parameters for adaptive to update the model and ultimately informs iterative management forest management decisions and learning [41] 23. Reported real-world updating and Yes = 10 points Represents actual implementation of QDS in on-ground adapting of management actions adaptive management Table 3 Numbers of case studies where ecosystem services (ESs) were modeled to inform adaptive forest management 2005–2015. ES classification is based on CICES [35] version 4.3 Section Division Group Number of case studies Example services Provisioning Materials Biomass 20 Providing timber, pulpwood [48] Provisioning Nutrition Biomass 3 Providing understory mushrooms [49] Provisioning Nutrition Water 1 Providing drinking water [50] Regulation Maintenance of physical, chemical, Lifecycle, habitat, genetic 9 Persistence of birds and mammals [51] biological conditions Regulation Maintenance of physical, chemical, Atmosphere and climate 3 Regulating CO [52] biological conditions Regulation Maintenance of physical, chemical, Pests and diseases 1 Regulating gypsy moth biological conditions (Lymantria dispar) population [53] Regulation Maintenance of physical, chemical, By natural chemical and 1 Regulating fire [54] abiotic conditions physical processes Cultural Human interactions with the Physical and experiential 1 Recreation access [55] environment Curr Forestry Rep (2018) 4:111–124 117 Temporal Trends found that five of six case studies with an overall score ≥ 5 modeled multiple ecosystem services and used stochastic dy- Basedontheattributes of the selected case studies, the overall namic programming to evaluate management options. score of AFM declined by 35% from 4.5 ± 0.7 to 2.9 ± 0.7 between 2005 and 2015 (Fig. 1). The score of the modeling Geographic Patterns task also declined during this period by a similar amount from 4.8 ± 0.7 to 3.3 ± 0.8. We detected no other statistically signif- Case studies were either globally relevant (n = 6) or focused icant temporal trends in scores for the remaining tasks nor sub- on forest management in one of the following four continents: tasks of adaptive management, although the parameter for the Asia (n = 2), Australia (n = 7), Europe (n = 9), or North year effect on score was negative for all subtasks. Notably, the America (n = 10). Place-specific case studies each covered number of case studies published per year (range 1 to 5) neither one (n = 22), two (n = 2) or three (n = 1) of seven climatic increased nor decreased significantly. Scores for prediction and zones, including temperate (n = 20) followed by cold (n =5), management options had the greatest interannual variability (0 arid-steppe (n = 2), and tropical-rainforest zones (n =1; to 9), compared to the other subtasks (range of ca. 5). We also Online Resource 4). With the exception of two papers focused on China and Costa Rica, respectively, place-specific case studies focused on forest management in developed countries (Online Resource 4). All place-specific case studies focused on forest management within a single country, except for one that focused on forests in parts of the USA and Canada [55]. The overall score of adaptive management did not vary significantly among the three continents having ≥ 7casestud- ies (i.e., Australia, Europe, and North America). When exam- ining the raw average scores, none of the average scores for a given geography exceeded those of all other geographies across all subtasks (Fig. 2). Subtask scores for Asia were less than or similar to those of the other continents, except for the management options task. Australia had similar or greater scores for each subtask, except for optimization. Compared to other geographies, case studies in Australia and North America had higher scores for prediction, stakeholder in- volvement, and practice. Scores were most similar for these two continents. Globally relevant case studies had a higher optimization score compared to place-specific case studies. Variation within and among Tasks When comparing scores among the three tasks, the application task (stakeholder involvement and practice) score (1.8 ± 0.7) was nearly 25% less than those of the modeling or setup tasks (4.0 ± 0.7 or 4.7 ± 0.7, respectively). When comparing sub- tasks, the scores for objectives, options, and predictions exceeded those for uncertainty, stakeholder, and implement (Fig. 3). Except for uncertainty, which had a lower score, the subtask scores were balanced between modeling (predictions and optimization) and the conceptual setup tasks (objectives and options). Average scores for individual attributes ranged from 0.3 to 7.4, and we detected significant pairwise contrasts between attributes within each subtask except for uncertainty and Fig. 1 Changes in scores of overall quality of quantitative decision stakeholder (Fig. 4). Within the objectives subtask, the aver- support (a)and qualityofprediction (b) to inform adaptive forest age score for temporal horizon exceeded that of ES classes management in 34 case studies published 2005–2015. The solid line is (attributes 3 vs. 1). Considering the options subtask, the aver- the best fit to the data based on a linear regression, and dashed lines represent the 95% confidence limits age score for management intervals was greater than that of 118 Curr Forestry Rep (2018) 4:111–124 Fig. 2 Scores for 34 case studies (published 2005–2015) using quantitative decision support to inform adaptive management by subtask among focal continents (or global focus) stand level (attributes 5 vs. 4). In the predictions subtask, the scoring attributes. Within the practice subtask, the average average score for model visualize (attribute 11) was lower score for management recommendation (attribute 23) than both stochastic and parameterize (attributes 10 and 12), exceeded that of the remaining attributes (20 through 22), and there was no significant difference between these higher which included applied QDS, applied model update, and dem- onstrated model update. Cluster Analysis The majority of indices indicated that two clusters are optimal when considering the variation in scores among subtasks (Online Resource 5). Average overall score did not differ sig- nificantly between the two clusters, but differences were ob- served when comparing the mean scores for subtasks in the exploratory analysis (Fig. 5). Compared to cluster 2, cluster 1 had higher scores for stakeholder (6.0 ± 1.1 vs. 0.5 ± 0.8) and prediction (9.2 ± 1.4 vs. 4.5 ± 1.1) but a lower score for opti- mize (2.6 ± 1.4 vs. 5.1 ± 1.1). Discussion We developed a multimetric index of quality of applications of quantitative decision support to inform adaptive management and applied this through an intensive structured literature review of case studies on adaptive forest management. The index pro- Fig. 3 Scores for quality of case studies using quantitative decision vides not only a total score for a given case study but also scores support to inform adaptive forest management by subtask in each of the for each of the main tasks, subtasks, and attributes of developing 34 case studies published 2005–2015. Dots are averages, whiskers are and applying QDS to inform AFM. This way the index measures 95% confidence intervals, and differing letters between means indicate a significant difference the multiple dimensions of success of specific applications and Curr Forestry Rep (2018) 4:111–124 119 Optimize Stakeholder (SH) Practice Objectives Options Uncertainty Predictions Fig. 4 Scores for quality of attributes within 7 subtasks (bold x-axis published 2005–2015. Dots are averages, whiskers are 95% confidence labels) and 3 tasks (gray boxes) of quantitative decision support to intervals, and differing letters between means within a subtask indicate a inform adaptive forest management in each of the 34 case studies significant pairwise difference can reveal important contrasts and synergies among these tasks to the poorly developed QDS). Our multimetric index is designed as they contribute to quality of the application as a whole. For to capture such contrasts within a given case study. We therefore example, AFM could be implemented based on a poorly devel- argue it is important to consider task-specific scores in addition to oped QDS, and therefore, the effort could be considered success- total scores to have a more complete picture of QDS for ful at implementation but having a low potential for learning (due informing adaptive management. Fig. 5 Scores for quality of quantitative decision support to inform adaptive forest management by subtask in each of the 34 case studies published 2005–2015. Radar graph depicts average scores for each cluster identified by cluster analysis, average scores across all case studies, and scores for individual case studies with the highest and lowest scores across all attributes 120 Curr Forestry Rep (2018) 4:111–124 Temporal Trends Variation Within and Among Tasks We discovered that over the past decade, overall quality of Ecosystem services relevant to management objectives QDS to inform AFM declined by 33% from 4.5 to 3.0 of 10, being modeled in the selected case studies lacked diver- which supports the temporal hypothesis. Quality of modeling sity, as the strong majority of case studies have focused (especially optimization and prediction) associated with QDS on biomass as a provisioning service. Case studies fo- for AFM has also declined during this period, while metrics cusing on regulating and especially cultural services are related to uncertainty have remained low (< 5 of 10). Declines needed to demonstrate the potential of QDS to inform in quality of modeling within QDS for AFM have not been AFM addressing a broader suite of ESs, which is rele- balanced with improvements in other tasks, for which there vant for the United Nations Sustainable Development are at least three possible explanations. First, the decline may Goals [37]. be related to reduced per capita funding for developing QDS We confirmed the expectation that there has been a for AFM. This hypothesis stems from observations of stable to substantial lack of stakeholder engagement during QDS declining funding available for environmental research in gen- development tasks and on-ground implementation of eral within the USA [56, 57], along with an increasing number adaptive forest management. Aligning with the reviews of publications in the field of adaptive management [4� ]. As a of adaptive natural resource management [3� , 5], report- result of this decoupling, environmental research institu- ed on-ground application of QDS to inform AFM was tions and researchers may have shrunken the time and re- rare (2 out of the 34 case studies). We were therefore sources they devote toward application-oriented modeling unable to examine relationships between successful im- activities. Another hypothesis is that the per capita funding plementation and other tasks such as stakeholder in- for research supporting adaptive management has volvement. The question of which factors are most lim- remained stable, but researchers in this area are developing iting for successful implementation of adaptive forest models that are increasingly complex but less comprehen- management remains an important area of future work, sive when it comes to developing QDS to inform AFM. A and examining patterns in scores within our multimetric third hypothesis is that there may be a growing publication index helps to generate hypotheses about the limiting bias toward publishing novel tools and approaches, which factors. For example, the spatial scale of management has pushed researchers to publish early iterations of QDS options considered by the reviewed case studies was rather than invest in demonstrating actual implementation. often broader than stand level, which could limit the Whatever the mechanism, our findings indicate that contem- real-world application of QDS. Furthermore, many case porary QDS to inform AFM is less intensive and more qualita- studies lacked a visualization of the structure of the tive than their predecessors. We argue that this comes at a cost modeling framework(s) embedded within the QDS. of less transparent decision-making for natural resource man- Successful uptake of QDS by forest managers likely agement and undermines the ability of researchers to contribute depends on their understanding of it, which would be a critical task for adaptive management. Despite a large interest supported by clear diagrams illustrating the logic and in adaptive management, as evidenced by many review papers key parameters. on the subject (10 published 2006–2014; cited in the first par- Our finding that stakeholder involvement and practice agraph of the BIntroduction^), we hypothesize that research occurs in a minority of decision support applications is funding priorities have drifted away from comprehensive supported by previous literature reviews (e.g., [5]). Of the modeling needed to inform adaptive management. Such a shift 108 case studies on adaptive comanagement reviewed by could have been spurred by a belief that rare implementation of Plummer et al. [8], only 17 were focused on forest man- adaptive management is the fault of the approach itself rather agement. They found that stakeholder participation was the than how QDS is developed to inform it [10]. As such, the most commonly cited factor contributing to the potential underlying mechanism for this hypothesized decline in per success of adaptive comanagement, whereas this was one capita funding may be that the number of such researchers in of the less commonly cited factors contributing to actual this field has increased faster than total funding for such efforts. success. They also reported that conflict of interest among We are unaware of any peer-reviewed study that has examined participating stakeholders was the most commonly cited trends in funding and researchers in the field of environmental factor leading to failures of adaptive comanagement. science (including QDS and AFM), but there is some When asked about suggested guidelines on developing de- supporting evidence from gray literature on recent trends in cision support systems to inform forest management, funding for environmental science in the USA [56, 57]. We stakeholder involvement was one of the most frequently are unaware of any database that tracks funding nor numbers voted suggestions according to experts as part of another of researchers in the realm of QDS or AFM, although this study [58]. Although caution must be taken when applying would be a useful tool for future research. adaptive management in contexts with stakeholder Curr Forestry Rep (2018) 4:111–124 121 conflicts, resolution of conflicts was the most commonly resources may be limiting the merging of these schools of cited outcome of these processes. They also found that adaptive management (sensu [1]) to enable both high- learning was the most common emergent property of adap- quality optimization and high-quality application to over- tive comanagement. These findings highlight the impor- come the current ceiling of moderate-quality QDS to in- tance of future research that compares and evaluates ap- form AFM. Additional resources could allow for a new proaches to stakeholder involvement in adaptive manage- school of adaptive management that mitigates these ment with respect to outcomes for learning and manage- tradeoffs in a more balanced approach that improves qual- ment objectives. ity of all subtasks. Our scoring method provides a trans- On a more positive note, the quality of considered man- parent way of monitoring existing and emerging schools of agement options, predictive modeling, and optimization in adaptive management. QDS has been high in many cases. There has also been a general balance between the quality of the conceptual set- Bayesian Decision Networks up task and modeling task of QDS to support AFM. Although examples of implementation were lacking, sev- We argue that Bayesian decision networks (i.e., Bayesian belief eral (n = 5) of the selected case studies demonstrated theo- networks containing at least one decision node and one utility retical updating of model parameters for adaptive manage- node) [15] are useful yet underused for QDS. A Bayesian de- ment. Such illustrations can facilitate the use of QDS to cision network (BDN) is a parameterized influence diagram (i.e., inform actual implementation of AFM. visual representation of a predictive model that links manage- ment objectives to decision alternatives and possibly factors be- Geographic Variation yond direct control of managers), allowing for explicit represen- tation of uncertainty in the relationships between factors that are Our ability to evaluate the geographic hypothesis was limited, as usually represented by conditional probability tables [15 ]. The none of the reviewed case studies were in developing regions. expanded use of BDNs could form the basis of an emerging This finding is consistent with a literature review on adaptive school of thought that would bring together the more comanagement, which included case studies that incorporated application-focused resilience school with the more learning by multiple actors collaborating on implementing adap- optimization-focused decision-theoretic school of adaptive man- tive management in an environmental context [8]. In contrast agement. We postulate that harmonizing these two schools of with our sample that only included one case study from Asia, thought could help to slow or reverse the declining quality of published case studies on adaptive comanagement across all QDS to inform AFM. Another advantage is that BDNs can be ecosystem types (published between years 2000 and 2010) were designed in a way that is intuitive for nontechnical stakeholders relatively common in Asia, some were in Africa and South to visualize and understand how their inputs are used to form the America, and a small minority was in Australia. Further work structure and content of QDS that informs decision-making is needed to understand and address geographic gaps of pub- ([59]; B.J. Mattsson, unpublished data). lished case studies using QDS to inform AFM. Despite the known advantages of this tool, only one of the Based on our analysis, applications of QDS to inform AFM 32 case studies explicitly used a BDN as part of QDS [15�� ]. in Australia and North America were more successful with As with any computer-based decision support method, using a practice compared to those in other continents. By contrast, BDN carries the risk of a management team getting bogged case studies from Asia were strong in defining management down in planning and computing rather than on-ground action options. This could provide a basis for intercontinental work- and learning. BDNs do, however, provide a user-friendly and ing groups to exchange knowledge and support improved ap- customizable platform for planning and learning in the context plications of QDS to inform implementation of AFM. of adaptive management that can be applied to many possible contexts within and beyond forests [15�� , 60, 61]. There are Cluster Analysis many applications that users can use to construct simple BDNs free of charge (e.g., Netica, Norsys Software Corp, Our exploratory cluster analysis revealed two clusters of Vancouver, British Columbia, Canada; http://www.norsys. case studies. One cluster performed relatively well on op- com). While having the ability to serve on its own as QDS, timization, but it performed relatively poorly on prediction a BDN can also be embedded in more complex decision and stakeholder involvement. This cluster fits well under support systems that include a geographic information the decision-analytic school of adaptive management, with system or that offer multicriteria decision analysis tools (e.g., a greater emphasis on modeling compared to application Ecosystem Management Decision Support; http://emds. and practice. Despite their differences in quality of sub- mountain-viewgroup.com/explore). A BDN can also serve tasks, the two clusters were similar in terms of overall as a starting point for dynamic algorithms to inform iterative decision-making [6, 62]. Increasing interest and demand index of quality. An emergent hypothesis, then, is that �� �� 122 Curr Forestry Rep (2018) 4:111–124 Open Access This article is distributed under the terms of the Creative should lead to a community of practice that will allow for Commons Attribution 4.0 International License (http:// the development of open-source BDNs that are freely creativecommons.org/licenses/by/4.0/), which permits unrestricted use, available for more widespread and advanced use (see for distribution, and reproduction in any medium, provided you give example [38]). Four case studies used a Bayesian belief appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. network (i.e., BDN without decision or utility nodes) [6, 47��, 59, 63], and 10 presented an influence diagram to represent a stochastic optimization (e.g., [64]). Extending References these existing applications to specifying a BDN would re- quire a rather small additional time investment, with a dis- Papers of particular interest, published recently, have been tinct advantage of providing a more transparent means of highlighted as: linking QDS to real-world decision-making. � Of importance �� Of major importance Conclusion 1.�� Williams BK, Brown ED. Adaptive management: from more talk to real action. Environ Manag. 2014;53(2):465–79. https://doi.org/10. Our study represents the first comprehensive assessment of the 1007/s00267-013-0205-7. The authors demonstrate a quality of QDS to inform adaptive management, and our frame- framework for adaptive management and point out needs of work is extensible such that it can be applied to decision con- balancing technical specificity with stakeholder involvement texts outside of forest management. We have also demonstrated while addressing challenges of sustainable development and maintaining ecosystem services. for the first time how cluster analysis and radar charts can be 2. Greig LA, Marmorek DR, Murray C, Robinson DCE. Insight into useful for exploring patterns according to the seven subtasks of enabling adaptive management. Ecol Soc 2013;18(3). https://doi. QDS development and application for adaptive management. org/10.5751/es-05686-180324. We found these techniques to be useful for generating new 3.� Westgate MJ, Likens GE, Lindenmayer DB. Adaptive management hypotheses about how the multidimensional quality of QDS of biological systems: a review. Biol Conserv. 2013;158(Supplement C):128–39. https://doi.org/10.1016/j. for AFM has evolved and might continue to vary over time biocon.2012.08.016. Based on a structured review of literature and space. This approach would be useful to inform future through 2011 on adaptive management for biodiversity and research agendas to ensure that key knowledge gaps are being ecosystem integrity, the authors provide suggestions for filled, for example addressing how QDS can better integrate improvement such as better collaboration between scientists and managers and better communicating risks of non- stakeholder input and be implemented on the ground as part adaptive approaches. of AFM. An expanded literature review would enable sufficient 4.� McFadden JE, Hiller TL, Tyre AJ. Evaluating the efficacy of adap- sample sizes to formally evaluate these hypotheses. This further tive management approaches: is there a formula for success? J work is important to quantify relationships between scores for Environ Manag. 2011;92(5):1354–9. https://doi.org/10.1016/j. jenvman.2010.10.038. Based on a structured literature review actual implementation and tasks related to QDS development, (2000–2009), the authors examine temporal trends in which will provide insight into the usefulness of the index to frequency of publications and degrees of success originating examine drivers of successful implementation of AFM. To min- from two adaptive management schools of thought. imize bias, scoring of attributes should be conducted indepen- 5. Fabricius C, Cundill G. Learning in adaptive management: insights from published practice. Ecol Soc. 2014;19(1). https://doi.org/10. dently by multiple experts and then discussed before assigning 5751/es-06263-190129. a final score. Another area of future research is a controlled 6. Rumpff L, Duncan DH, Vesk PA, Keith DA, Wintle BA. State-and- experiment where adaptive management teams are assigned transition modelling for adaptive management of native woodlands. restrictions as to the levels of emphasis on particular subtasks Biol Conserv. 2011;144(4):1224–36. https://doi.org/10.1016/j. (e.g., stakeholder involvement and predictive modeling). We biocon.2010.10.026. 7. Williams BK. Adaptive management of natural resources—frame- encourage any research that critically examines what works work and issues. J Environ Manag. 2011;92(5):1346–53. https:// and what does not work in the field of QDS and adaptive doi.org/10.1016/j.jenvman.2010.10.041. management, so that we can more quickly learn and achieve 8. Plummer R, Crona B, Armitage DR, Olsson P, Tengö M, Yudina O. conservation and natural resource management goals. Adaptive comanagement: a systematic review and analysis. Ecol Soc. 2012;17(3):11. https://doi.org/10.5751/ES-04952-170311(3):11. Funding Information Open access funding provided by University of 9. Gregory R, Ohlson D, Arvai J. Deconstructing adaptive manage- Natural Resources and Life Sciences Vienna (BOKU). ment: criteria for applications to environmental management. Ecol Appl. 2006;16(6):2411–25. https://doi.org/10.1890/1051- 0761(2006)016[2411:damcfa]2.0.co;2. Compliance with Ethical Standards 10. Allen CR, Fontaine JJ, Pope KL, Garmestani AS. Adaptive man- agement for a turbulent future. J Environ Manag. 2011;92(5):1339– Human and Animal Rights and Informed Consent This article does not 45. https://doi.org/10.1016/j.jenvman.2010.11.019. contain any studies with human or animal subjects performed by any of 11. Borges JG, Nordström E-M, Garcia Gonzalo J, Hujala T, the authors. Trasobares A. Computer-based tools for supporting forest Curr Forestry Rep (2018) 4:111–124 123 management. 2014. http://pub.epsilon.slu.se/id/eprint/11417. 23. Yousefpour R, Jacobsen JB, Thorsen BJ, Meilby H, Hanewinkel M, Oehler K. A review of decision-making approaches to handle un- Accessed 2 April 2018. certainty and risk in adaptive forest management under climate 12. Muys B, Hynynen J, Palahi M, Lexer MJ, Fabrika M, Pretzsch H, et change. Ann For Sci. 2012;69(1):1–15. https://doi.org/10.1007/ al. Simulation tools for decision support to adaptive forest manage- s13595-011-0153-4. ment in Europe. For Sys. 2010;19:86–9. https://doi.org/10.5424/fs/ 24. Whiteman A, Wickramasinghe A, Pina L. Global trends in forest 201019S-9310. ownership, public income and expenditure on forestry and forestry 13.�� McIntosh BS, Ascough JC, Twery M, Chew J, Elmahdi A, Haase employment. For Ecol Manag. 2015;352:99–108. https://doi.org/ D, et al. Environmental decision support systems (EDSS) develop- 10.1016/j.foreco.2015.04.011. ment—challenges and best practices. Environ Model Softw. 25. Radke N, Yousefpour R, von Detten R, Reifenberg S, Hanewinkel 2011;26(12):1389–402. https://doi.org/10.1016/j.envsoft.2011.09. M. Adopting robust decision-making to forest management under 009. As authors of the paper, 27 developers of environmental climate change. Ann For Sci. 2017;74(2):43. https://doi.org/10. decision support systems (EDSSs) identified ways of overcom- 1007/s13595-017-0641-2. ing challenges of EDSS adoption and uptake such as the use of a 26. Pasalodos-Tato M, Mäkinen A, Garcia-Gonzalo J, Borges JG, champion to promote the EDSS within target user groups, en- Lämås T, Eriksson LO. Review. Assessing uncertainty and risk in sure EDSS elements are easy and inexpensive to use and up- forest planning and decision support systems: review of classical date, build capacity within target groups to ensure they can methods and introduction of new approaches. For Sys. 2013;22(2): efficiently use the EDSS, and develop metrics of success of 282–303. https://doi.org/10.5424/fs/2013222-03063. EDSS applications. 27. Bettinger P, Boston K, Siry JP, Grebner DL. Forest management 14. Valls-Donderis P, Ray D, Peace A, Stewart A, Lawrence A, Galiana and planning. New York: Academic; 2009. F. Participatory development of decision support systems: which 28. Moore CT, Conroy MJ. Optimal regeneration planning for old- features of the process lead to improved uptake and better out- growth forest: addressing scientific uncertainty in endangered spe- comes? Scand J For Res. 2014;29:71–83. https://doi.org/10.1080/ cies recovery through adaptive management. For Sci. 2006;52(2): 02827581.2013.837950. 155–72. 15.�� Nyberg JB, Marcot BG, Sulyma R. Using Bayesian belief networks 29. Wikström P, Edenius L, Elfving B, Eriksson LO, Lämås T, in adaptive management. Can J For Res. 2006;36(12):3104–16. Sonesson J, et al. The heureka forestry decision support system: https://doi.org/10.1139/x06-108. Through general arguments and an overview. Math Comput For Nat Res Sci. 2011;3(2):87. a case study, this paper provides a clear demonstration of how 30. forestDSS. List of forest decision support systems. 2017. http:// influence diagrams and BDNs can be useful to inform formal www.forestdss.org/wiki/index.php?title=Category:DSS. Accessed adaptive natural resource management in collaborations 2 April 2018. between scientists and stakeholders. 31. Segura M, Ray D, Maroto C. Decision support systems for forest 16. Lawrence A, Stewart A. Sustainable forestry decisions: on the in- management: a comparative analysis and assessment. Comput terface between technology and participation. Math Comput For Electron Agric. 2014;101:55–67. https://doi.org/10.1016/j. Nat Res Sci. 2011;3(1):42. compag.2013.12.005. 17. Blattert C, Lemm R, Thees O, Lexer MJ, Hanewinkel M. 32. Keenan RJ. Climate change impacts and adaptation in forest man- Management of ecosystem services in mountain forests: review agement: A review. Annals of Forest Science. 2015;72(2):145–67. of indicators and value functions for model based multi-criteria https://doi.org/10.1007/s13595-014-0446-5. decision analysis. Ecol Indic. 2017;79:391–409. https://doi.org/ 33. Collaboration for Environmental Evidence. Guidelines for system- 10.1016/j.ecolind.2017.04.025. atic review and evidence synthesis in environmental management. 18. Irauschek F, Rammer W, Lexer MJ. Evaluating multifunctionality Version 4.2. 2013. http://www.environmentalevidence.org/wp- and adaptive capacity of mountain forest management alternatives content/uploads/2014/06/Review-guidelines-version-4.2-final.pdf. under climate change in the eastern alps. Eur J For Res. Accessed 2 April 2018. 2017;136(5–6):1051–69. https://doi.org/10.1007/s10342-017- 34. Chai J, Liu JNK, Ngai EWT. Application of decision-making tech- 1051-6. niques in supplier selection: a systematic review of literature. 19. Mina M, Bugmann H, Cordonnier T, Irauschek F, Klopcic M, Expert Syst Appl. 2013;40(10):3872–85. https://doi.org/10.1016/j. Pardos M, et al. Future ecosystem services from european mountain eswa.2012.12.040. forests under climate change. J Appl Ecol. 2017;54(2):389–401. 35. Common International Classification of Ecosystem Services https://doi.org/10.1111/1365-2664.12772. (CICES). Cices v4.3 spreadsheet. 2016. https://cices.eu/ 20. Vacik H, Kurttila M, Hujala T, Khadka C, Haara A, Pykäläinen J, et resources/. Accessed 2 April 2018. al. Evaluating collaborative planning methods supporting 36. Cubbage F, Harou P, Sills E. Policy instruments to enhance multi- programme-based planning in natural resource management. J functional forest management. Forest Policy Econ. 2007;9(7):833– Environ Manag. 2014;144:304–15. https://doi.org/10.1016/j. 51. https://doi.org/10.1016/j.forpol.2006.03.010. jenvman.2014.05.029. 37. UN General Assembly. Transforming our world: the 2030 agenda 21. Acosta M, Corral S. Multicriteria decision analysis and participato- for sustainable development. Report no. A/res/70/1. 2015. https:// ry decision support systems in forest management. Forests. sustainabledevelopment.un.org/post2015/transformingourworld/ 2017;8(4) https://doi.org/10.3390/f8040116. publication. Accessed 2 April 2018. 38. Seidl R, Lexer MJ. Forest management under climatic and social 22.�� Vacik H, Lexer MJ. Past, current and future drivers for the devel- opment of decision support systems in forest management. Scand J uncertainty: trade-offs between reducing climate change impacts For Res. 2014;29:2–19. https://doi.org/10.1080/02827581.2013. and fostering adaptive capacity. J Environ Manag. 2013;114:461–9. Based on a review of development of decision support 39. Collins M, Knutti R, Arblaster J, Dufresne J-L, Fichefet T, sy Friedlingstein P et al. Long-term climate change: projections, com- stems (DSSs) to inform forest management, the authors sug- mitments and irreversibility. 2013. gest that evolving toolkits of decision-support approaches rath- 40. Huang IB, Keisler J, Linkov I. Multi-criteria decision analysis in er than stand-alone DSSs can help balance wishes for scientific environmental sciences: ten years of applications and trends. Sci rigor with diverse demands by forest managers and policy- Total Environ. 2011;409(19):3578–94. https://doi.o makers for concise and user-friendly tools that fit within their rg/10.1016/j. normal work routine. scitotenv.2011.06.022. 124 Curr Forestry Rep (2018) 4:111–124 41. Yousefpour R, Jacobsen JB, Meilby H, Thorsen BJ. Knowledge 53. Bogich T, Shea K. A state-dependent model for the optimal man- agement of an invasive metapopulation. Ecol Appl. 2008;18(3): update in adaptive management of forest resources under climate change: a Bayesian simulation approach. Ann For Sci. 2014;71(2): 748–61. https://doi.org/10.1890/07-0642.1. 301–12. https://doi.org/10.1007/s13595-013-0320-x. 54. Chades I, Bouteiller B. Solving multiagent markov decision pro- 42. Walters CJ. Adaptive management of renewable resources. New cesses: a forest management example. Modsim 2005: International York: Macmillan; 1986. Congress on Modelling and Simulation: Advances and 43. Lee KN. Chapter 1. Appraising adaptive management. Biological Applications for Management and Decision Making. 2005:1594– diversity: balancing interests through adaptive collaborative man- agement. Boca Raton: CRC; 2001. p. 1–26. 55. Tomberlin D. Endangered seabird habitat management as a partially 44. Rist L, Campbell BM, Frost P. Adaptive management: where are we observable markov decision process. Mar Resour Econ. now? Environ Conserv. 2012;40(1):5–18. https://doi.org/10.1017/ 2010;25(1):93–104. S0376892912000240. 56. Jahnke A. Who picks up the tab for science? 2017. http://www.bu. 45. Charrad M, Ghazzali N, Boiteau V, Niknafs A. Nbclust: an R pack- edu/research/articles/funding-for-scientific-research/. Accessed 2 age for determining the relevant number of clusters in a data set. J April 2018. Stat Softw. 2014;61:1–36. 57. Neubauer SJ. Foundation funding trends for environmental & con- 46. R Core Team. R: a language and environment for statistical com- servation efforts 2015. https://www.calandtrusts.org/wp-content/ puting. Vienna: R Foundation for Statistical Computing. 2017. uploads/2015/03/Foundation-Funding-Trends-for-Environmental- https://www.R-project.org/. Accessed 2 April 2018. and-Conservation-Efforts.pdf. Accessed 2 April 2018. 47.�� Howes AL, Maron M, McAlpine CA. Bayesian networks and adap- 58. Marques A, Ficko A, Kangas A, Rosset C, Ferriti F, Rasinmaki J, et tive management of wildlife habitat. Conserv Biol. 2010;24(4): al. Empirical guidelines for forest management decision support 974–83. https://doi.org/10.1111/j.1523-1739.2010.01451.x. The systems based on the past experiences of the expert’s community. authors provide a case study demonstrating how an influence Forest Systems. 2013;22:320–39. diagram and Bayesian belief network are constructed based on 59. Marcot BG, Hohenlohe PA, Morey S, Holmes R, Molina R, Turley inputs from land managers and then updated based on new MC, Huff MH, Laurence JA. Characterizing species at risk II: using data to inform adaptive management of feral grazing and Bayesian belief networks as decision support tools to determine prescribed burning regimes to maintain avian diversity within species conservation categories under the northwest forest plan. woodlands of subtropical eastern Australia. Ecol Soc. 2006;11(2). 48. Yousefpour R, Temperli C, Bugmann H, Elkin C, Hanewinkel M, 60. Dalyander PS, Meyers M, Mattsson B, Steyer G, Godsey E, Meilby H, et al. Updating beliefs and combining evidence in adap- McDonald J, et al. Use of structured decision-making to explicitly tive forest management under climate change: a case study of incorporate environmental process understanding in management Norway spruce (Picea abies l. Karst) in the Black Forest, of coastal restoration projects: case study on barrier islands of the Germany. J Environ Manag. 2013;122:56–64. https://doi.org/10. Northern Gulf of Mexico. J Environ Manag. 2016;183:497–509. 1016/j.jenvman.2013.03.004. https://doi.org/10.1016/j.jenvman.2016.08.078. 49. Ferreira L, Constantino MF, Borges JG, Garcia-Gonzalo J. A sto- 61. Thorne KM, Mattsson BJ, Takekawa J, Cummings J, Crouse D, chastic dynamic programming approach to optimize short-rotation Block G, et al. Collaborative decision-analytic framework to max- coppice systems management scheduling: an application to euca- imize resilience of tidal marshes to climate change. Ecol Soc. lypt plantations under wildfire risk in Portugal. For Sci. 2012;58(4): 2015;20(1):30. https://doi.org/10.5751/es-07018-200130. 353–65. https://doi.org/10.5849/forsci.10-084. 62. Golovin D, Krause A, Gardner B, Converse SJ, Morey S. Dynamic 50. Spring DA, Kennedy JOS, Mac Nally R. Optimal management of a resource allocation in conservation planning. Proceedings of the forested catchment providing timber and carbon sequestration ben- twenty-fifth AAAI conference on artificial intelligence. 2011. efits: climate change effects. Glob Environ Chang. 2005;15(3): http://www.aaai.org/ocs/index.php/AAAI/AAAI11/paper/ 281–92. https://doi.org/10.1016/j.gloenvcha.2005.04.002. download/3617/4079/. Accessed 2 April 2018. 51. Sabbadin R, Spring D, Bergonnier E. A reinforcement-learning 63. Rumpff L, Duncan D, Vesk P, Wintle B. Development of a state- application to biodiversity conservation in costa-rican forest. and-transition model to guide investment in woodland vegetation Modsim 2007: International Congress on Modelling and condition. Landscape Logic: Integrating Science for Landscape Simulation: Land, Water and Environmental Management: Management. 2012:157–72. Integrated Systems for Sustainability. 2007:2189–95. 52. Haim D, Plantinga AJ, Thomann E. The optimal time path for 64. Shahi S, Pulkki R. A simulation-based optimization approach to carbon abatement and carbon sequestration under uncertainty: the integrated inventory management of a sawlog supply chain with case of stochastic targeted stock. Resour Energy Econ. 2014;36(1): demand uncertainty. Can J For Res. 2015;45(10):1313–26. https:// 151–65. https://doi.org/10.1016/j.reseneeco.2013.11.006. doi.org/10.1139/cjfr-2014-0373.

Journal

Current Forestry Reports – Springer Journals

Published: Jul 18, 2018

There are no references for this article.