Cascading extinctions due to loss of indirect ecological interactions

目前物种灭绝的高比率得到了很好的宣传，很明显，这在很大程度上是由于人类活动，如过度开发、栖息地破坏和温室气体排放导致的全球环境变化。物种的丧失不仅仅是一个道德和美学问题。已经多次证明，生态系统服务(粮食生产、吸收碳排放、虫害防治、防洪等)人类社会所依赖的生物多样性与生态系统中的物种多样性呈正相关。因此，保护生物多样性是各国政府的主要优先事项。这种保护的有效性取决于我们对导致物种减少的过程的理解程度。有时这是显而易见的：鳕鱼资源的崩溃显然是由于过度捕捞。然而，这种直接影响往往伴随着其他物种的二次灭绝，原因并不总是显而易见，这可能会导致一连串的进一步灭绝和生态系统崩溃。预测这些级联是具有挑战性的，需要详细了解生态系统中物种的相互关联性如何影响人类对一个物种的影响向与其没有直接联系的其他物种传播。这对于营养水平较高的物种(食肉动物)尤其重要，因为它们最容易灭绝。长期以来，有一种观点认为，专门捕食不同猎物的食肉动物物种通过限制猎物数量，从而防止一个猎物物种与另一个猎物物种竞争，从而对彼此产生积极影响。这样做的结果是，如果一个食肉动物从群落中消失，它的猎物就会失去控制，随后可能会在竞争中胜过其他猎物物种，导致其他肉食动物失去食物，面临灭绝。这可能是灭绝级联发生的一个重要机制，然而，很难获得这种效应的实验证据。我们在实验室用简单的昆虫群落进行了初步实验，证明了一种食肉动物物种的消失确实会导致其他物种在短短几代人内灭绝。现在的挑战是将这一模型系统扩大到更现实的规模。我们建议在基于野外的中观系统中进行实验-大约2米的立方体围栏，在其中我们可以控制生态群落的确切组成。我们将在这些地区聚集昆虫群落，对目标物种实施特定的收获制度，并跟踪这对其他物种的间接影响，特别是是否发生预测的灭绝级联。在这种情况下，我们可以操纵重要的变量，如猎物物种之间的竞争强度，并测试它们对结果的影响。我们还可以验证这样一种预测，即物种丰富、复杂的群落将对物种灭绝的级联反应更具抵抗力，这意味着随着生物多样性的丧失，进一步的损失引发灭绝级联的可能性也会增加。

The current high rates of species extinctions are well publicised and it is clear that much of this is due to human activities such as overexploitation, habitat destruction and global environmental change due to greenhouse gas emissions. The loss of species is not just a moral and aesthetic issue. It has been demonstrated repeatedly that the ecosystem services (food production, soaking up of carbon emissions, pest control, flood control etc.) that human societies rely on are positively related to species diversity in ecosystems. The protection of biodiversity is therefore a major priority for governments. The effectiveness of this protection depends on how well we understand the processes that lead to species declines. Sometimes this is obvious: The collapse of cod stocks is clearly due to overfishing. Often however, such direct impacts are followed by secondary extinctions of other species, for not always obvious reasons, with the danger that this leads to a cascade of further extinctions and ecosystem collapse. Predicting these cascades is challenging and requires a detailed understanding of how the interconnectedness of species in ecosystems affects the transmission of human impacts on one species to other species that are not directly linked to it. This is particularly important for species at higher trophic levels (carnivores) which are most vulnerable to extinction. The idea has long existed that species of carnivore that specialise on different prey have positive effects on each other by limiting their prey populations and thereby preventing one prey species from outcompeting the other. A consequence of this is that if a carnivore is lost from the community, its prey is released from control and may subsequently out-compete the other prey species leaving the other carnivore without food and facing extinction. This is potentially an important mechanism by which extinction cascades occur, however, it is difficult to obtain experimental evidence for such effects. We have done preliminary experiments with simple insect communities in the laboratory which have demonstrated that the removal of one carnivore species does indeed lead to the extinction of others in just a few generations. The challenge now is to scale this model system up to a more realistic scale. We propose to carry out experiments in field-based mesocosms - roughly 2 meter cubed enclosures in which we can control the exact composition of the ecological community. We will assemble communities of insects in these and impose specific harvesting regimes on target species and follow the indirect impact this has on the other species and, in particular, whether the predicted extinction cascades occur. Within this setting we can manipulate important variables, such as the strength of competition among prey species, and test their impact on the outcome. We can also test the prediction that more species-rich and complex communities will be more resistant to extinction cascades, which would mean that as biodiversity is lost, the chance that further losses trigger extinction cascades increases.