Predicting the scale of landscape context effects: How big is a landscape?

背景：可以说，保护中最重要的空间尺度问题是确定景观环境对特定种群或社区的影响距离。我们知道没有池塘、森林或保护区是岛屿。生态变量，如人口丰度或物种丰富度受到景观背景的强烈影响，例如周围道路密度、森林覆盖或城市化。为了采取有效的保护行动，我们需要知道这些影响发生的距离，即景观的“影响规模”。预测景观生态效应规模也是景观生态学研究的重要内容。如果测量景观变量的程度不正确，景观背景的影响可能无法被发现，并且/或者景观效应的相对重要性将被错误地估计。然而，许多研究人员武断地选择一个单一的空间尺度进行分析，而另一些人则根据生物的运动生物学来衡量景观，他们合理地假设，效应的尺度取决于生物与其环境相互作用的空间程度。到目前为止，对这一假设的检验结果好坏参半，这表明还有其他因素在起作用。例如，人口统计参数（例如，繁殖率）、物种运动行为细节和物种相互作用如何影响效应尺度？根据景观预测因子（道路密度、森林覆盖等），尺度如何变化？它是否取决于测量的响应，例如，丰度，物种丰富度？目前还没有全面的理论来预测景观文脉对人口和社区反应的影响程度。目的：在接下来的5年里，我的目标是建立并验证一套假设，以预测景观背景对人口和社区反应的影响规模。方法：我提出了8个假设，表明效应规模与(I)物种特征（扩散、繁殖率、资源利用、天敌特征）、（ii）响应变量（丰度、发生率、物种丰富度）和（iii）景观变量类型（通过资源可用性，如栖息地数量，或通过死亡率，如道路密度）之间的关系。我们将使用基于个体的空间模拟模型来评估这些假设，并估计影响效应尺度的因素的相对重要性。然后，我们将首先通过(i)估算景观预测因子对多个个体物种和群落响应的影响规模，使用多尺度实证研究，然后（ii）进行跨物种（或物种组）比较，以检验观察到的影响规模是否与假设和模拟预测的一样。由于我的目标是建立一个通用的、有效的理论来预测效应的规模，我们将对几个不同的物种群体进行实证测试，包括：森林小型哺乳动物、池塘和森林生活的两栖动物、树皮和木材蛀虫、森林地面植物和阿尔瓦植物。意义：我们知道选择“正确”的空间尺度对于景观和保护生态非常重要，但我们通常是“随心而为”，因为我们不知道正确的尺度先验。拟议的研究将建立有效的理论，使景观和保护生态学家能够选择正确的空间尺度进行分析和应用。这将(i)增加景观生态学研究的推理强度，（ii）大大提高景观尺度保护行动的有效性，确保它们在正确的空间范围内实施，以最大限度地成功实现保护目标。

BACKGROUND: Arguably the most important spatial scaling problem in conservation is identification of the distance within which the landscape context affects a given population or community. We know that no pond, forest patch, or protected area is an island. Ecological variables such as population abundance or species richness are strongly influenced by the landscape context, e.g. the surrounding road density, forest cover, or urbanization. For effective conservation action we need to know the distance within which these effects occur, i.e. the ‘scale of effect’ of the landscape. Predicting the scale of effect is also critical to landscape ecological research. If landscape variables are measured at the wrong extent, effects of landscape context can go undetected, and/or relative importance of landscape effects will be incorrectly estimated. Nevertheless, many researchers arbitrarily select a single spatial scale for analysis and others scale the landscape to the movement biology of the organism(s), on the reasonable assumption that the scale of effect depends on the spatial extent over which the organism interacts with its environment. Tests of this assumption so far have been mixed, indicating that other factors come into play. For example, how is the scale of effect influenced by demographic parameters (e.g., reproductive rate), by details of a species’ movement behaviour, and by species interactions? How does the scale change, depending on the landscape predictor (road density, forest cover etc.)? And, does it depend on the response measured, e.g., abundance, species richness? There is currently no comprehensive theory for predicting the scale of effect of the landscape context on population and community responses. OBJECTIVE: Over the next 5 years I aim to build and validate a set of hypotheses for predicting the scale of effect of the landscape context on population and community responses. APPROACH: I propose eight hypotheses, suggesting relationships between the scale of effect and (i) species traits (dispersal, reproduction rate, resource use, traits of natural enemies), (ii) response variables (abundance, occurrence, species richness), and (iii) type of landscape variable (acting through resource availability e.g. habitat amount, or through mortality e.g. road density). We will use individual-based spatial simulation modelling to evaluate the hypotheses and to estimate relative importance of factors influencing the scale of effect. We will then test the hypotheses by first (i) estimating the scale of effect of landscape predictors on multiple individual species and community responses, using multi-scale empirical studies, and then (ii) conducting cross-species (or species group) comparisons to test whether the observed scale of effect changes as predicted from the hypotheses and simulations. Since my goal is to develop a general, validated theory for predicting the scale of effect, we will conduct the empirical tests for several different species groups, including: forest small mammals, pond and forest dwelling amphibians, bark and wood boring beetles, forest floor plants, and alvar plants. SIGNIFICANCE: We know that selecting the ‘right’ spatial scale is hugely important in landscape and conservation ecology, but we generally ‘wing it’ because we do not know the right scale a priori. The proposed research will build validated theory that will allow landscape and conservation ecologists to select the right spatial scale for analyses and application. This will (i) increase inferential strength in landscape ecology studies and (ii) greatly increase effectiveness of landscape-scale conservation actions, ensuring that they are implemented at the correct spatial extent for best success in effecting conservation goals.