How does global land-use change reshape ecological assemblages over time?

由于人类的影响，全球陆地生物多样性正在下降，其中土地利用变化迄今为止是最重要的。当人们改变土地的使用方式时，许多原本存在的物种在该地区减少或消失，而其他以前没有的物种则在该地区建立起来。虽然有些物种会立即受到影响，但其他物种可能只是在土地利用变化的后果波及整个生态系统后才会做出反应。这种延迟或拖延的反应，我们称之为“生物滞后”，迄今为止在大规模模型中基本上被忽略了。以前许多研究的另一个缺点是，它们关注的是物种的数量，而不是它们的作用。由于变化中的“赢家”在生态上可能与“输家”不同，土地利用变化影响了群落的功能，以及它包含的物种数量。随着人们试图在对生态系统产品的短期需求和对可持续性的长期需求之间取得平衡，了解土地利用变化如何以及多快地影响当地群落，对于支持未来几十年更好的土地利用决策至关重要。要评估土地利用变化对当地生态群落的全球影响，最明显的方法是监测几十年来，在一组具有代表性的大型地点上，土地利用和群落是如何变化的。这些地点必须具有代表性，以避免有偏差的结果，而且需要很长的时间尺度，因为这些反应可能会持续多年。因为没有这样一组站点，所以需要不那么直接的方法。我们计划搜集生态学文献，比较土地利用变化前后的群落。我们可以纠正偏差，因为我们对土地利用的不同变化有多普遍进行了估计；我们将对土地利用变化后的反应如何随时间变化进行建模，这样我们就可以进行长期和短期的研究。有数百个合适的研究，我们将要求研究人员与我们分享他们的数据；然后我们会在项目结束时将它们提供给每个人。我们将结合土地利用变化前后的物种丰度数据及其生态作用信息，揭示土地利用变化如何以及多快地影响各种物种的相对丰度以及群落的生态结构和功能。例如，自然栖息地向农业的转变是否倾向于小物种而不是大物种，如果是这样，速度有多快？在人类主导的土地使用中，新陈代谢是否更快？这些分析将需要对几个生态上重要和高度多样化的节肢动物群体进行新的特征数据汇编；为了制作这些作品，我们将利用自然历史博物馆的专业知识、藏品和图书馆。在早期nerc资助的一个项目（predic.com: www.predicts.org.uk）中，我们已经汇编了500多个数据集——由300多个不同的研究人员提供——这些数据集比较了土地利用不同的其他匹配地点。“预测”数据库已经从88个国家的18000多个站点收集了超过200万条记录。该数据库包含的物种数量是正式描述的物种数量的1%以上。我们对这一空前庞大且具有代表性的数据集的分析表明，土地利用变化对平均地方多样性产生了显著的全球影响。然而，由于“预测”的数据集是空间比较而不是时间比较，它们不太适合分析群落如何对土地利用变化作出反应的动态。更根本的是，“预测”关于空间比较足以替代时间数据的假设现在需要检验。这一建议将提供必要的测试，并产生最全面的关于土地利用变化如何随着时间的推移重塑生态组合的图景。

Terrestrial biodiversity is declining globally because of human impacts, of which land-use change has so far been the most important. When people change how land is used, many of the species originally present decline or disappear from the area, while others previously absent become established. Although some species are affected immediately, others might only respond later as the consequences of the land-use change ripple through the ecosystem. Such delayed or protracted responses, which we term 'biotic lag', have largely been ignored in large-scale models so far. Another shortcoming of much previous work is that it has focused on numbers of species, rather than what they do. Because 'winners' from the change are likely to be ecologically different from 'losers', the land-use change impacts how the assemblage functions, as well as how many species it contains. Understanding how - and how quickly - land-use change affects local assemblages is crucial for supporting better land-use decisions in the decades to come, as people try to strike the balance between short-term needs for products from ecosystems and the longer-term need for sustainability.The most obvious way to assess the global effects of land-use change on local ecological communities would be to have monitored how land use and the community have changed over a large, representative set of sites over many decades. The sites have to be representative to avoid a biased result, and the long time scale is needed because the responses can unfold over many years. Because there is no such set of sites, less direct approaches are needed. We are planning to scour the ecological literature for comparisons of communities before and after land-use change. We can correct for bias because we have estimates of how common different changes in land use have been; and we will model how responses change over time after a land-use change so that we can use longer-term and shorter-term studies alike. There are many hundreds of suitable studies, and we will ask the researchers who produced them to share their data with us; we will then make them available to everyone at the end of the project.We will combine data on species' abundances before and after the land-use change with information about their ecological roles, to reveal how - and how quickly - changing land use affects the relative abundances of the various species and the ecological structure and function of the community. Does conversion of natural habitats to agriculture tend to favour smaller species over large ones, for instance, and if so how quickly? Is metabolism faster in more human-dominated land uses? These analyses will require new compilations of trait data for several ecologically important and highly diverse arthropod groups; to produce these, we will make use of the expertise, collections and library of the Natural History Museum.In an earlier NERC-funded project (PREDICTS: www.predicts.org.uk), we have already compiled over 500 data sets - provided by over 300 different researchers - that compared otherwise-matched sites where land use differed. The PREDICTS database has amassed over 2,000,000 records, from over 18,000 sites in 88 countries. The database contains more than 1% as many species as have been formally described. Our analyses of this unprecedentedly large and representative data set indicates that land-use change has had a marked global impact on average local diversity. However, because PREDICTS' data sets are spatial rather than temporal comparisons, they are not well-suited to analysing the dynamics of how assemblages respond to land-use change. More fundamentally, PREDICTS' assumption that spatial comparisons are an adequate substitute for temporal data now needs testing. This proposal will deliver the necessary tests, as well as producing the most comprehensive picture of how land-use change reshapes ecological assemblages through time.