Assessing the influence of deer browsing on forest ecosystem services with new stand-level optimization and agent-based models

鹿吃草可能会严重影响森林的发展，造成严重的环境后果。虽然现有的研究已经深入分析了鹿的浏览的组成和森林的生长的生物化学物理的影响，鹿浏览的生态系统服务（ES）的供应的影响的综合评估仍然缺乏。在这里，我们提出了三个相互关联的研究领域，以填补这一空白。1)我们建议创新的林分水平的多标准优化下的不确定性，以获得理想的林分作为基准，将最佳地满足利益相关者的喜好。建议的创新，以提高这种模式，包括治疗ES发生在遥远的未来和评估的健壮性ES供应。这第一个研究领域将考虑木材供应，碳固存，地下水补给，地下水质量（例如氮污染），经济回报和生物多样性作为决策标准（枯木作为指标）。2)基于最佳组合的基准林分，我们建议模拟实验，以阐明鹿浏览的影响。我们建议在森林景观模型中实施基准森林，并将其暴露给鹿浏览。为了模拟鹿的影响，我们将概念化一个基于代理的模型。在这样的模型中，小树和鹿是自主代理，多代理研究和仿真框架将支持建立这些代理之间的交互。对当前项目“生物野生”和现有文献的观察将提供信息，以建立幼树对浏览的反应和鹿对森林特征的反应的规则。3)假设激烈的鹿浏览将形成约束的最佳组成和管理的森林立场，我们建议推导影子价格的鹿对树木的影响。这些将包含在我们的基准森林优化中的所有ES的成本。我们希望，这一研究概念将允许一个全面的环境影响评估沉重的鹿浏览树木。

Deer browsing may strongly affect the development of forests with severe environmental consequences. While the existing research has intensely analyzed the bio-chemo-physical impact of deer browsing on the composition and growth of forest stands, a comprehensive assessment of the influence of deer browsing on the provisioning of ecosystem services (ES) is still lacking. Here we propose three interrelated research fields to fill this gap. 1) We suggest innovative stand-level multiple criteria optimization under uncertainty to derive ideal forest stands to be used as benchmarks that will optimally meet stakeholders’ preferences. The proposed innovations for enhancing such models include the treatment of ES occurring in the far future and the assessment of the robustness of ES provisioning. This first research field will consider timber provisioning, carbon sequestration, groundwater recharge, groundwater quality (for the example of nitrogen contamination), economic return and biodiversity as decision criteria (with dead wood as an indicator). 2) Based on the optimally composed benchmark forest stands we suggest simulation experiments to elucidate the impact of deer browsing. We propose implementing the benchmark forests in a forest landscape model and exposing them to deer browsing. For simulating the impact of deer, we will conceptualize an agent-based model. In such a model, young trees and deer are autonomous agents, where a framework for multi-agent research and simulation will support establishing interactions among such agents. Observations of the current project “BioWild” and existing literature will provide information to set up rules for the young trees’ response to browsing and the deer’s response to forest characteristics. 3) Assuming that intense deer browsing will form constraints for the optimal composition and management of a forest stand we suggest deriving shadow prices for the deer’s influence on the trees. These will encapsulate the costs for all ES included in the optimization of our benchmark forests. We expect that this research concept will allow a comprehensive assessment of the environmental impact of heavy deer browsing on trees.