LTREB: On the Consequences of Removing Ground-dwelling Mammals for Tropical Forest Diversity: Towards a New Conceptual Framework

0212054CarsonLTREB我们正在进行一项长期、大规模的实验，旨在解决一个主要假设，即哺乳动物是草食动物、种子捕食者和散布者，是热带树木多样性和丰富度的关键决定因素。这项实验本质上是对詹森-康奈尔假设的社区层面的测试，该假设预测，地面哺乳动物的消失最终将导致热带树木多样性的下降。在这种情况下，哺乳动物优先捕食丰富的树种的幼树，并最终控制这些树种。研究系统是巴拿马中部巴罗科罗拉多岛(BCI)和邻近大陆(吉甘特半岛)的成熟期热带潮湿森林。自1994年以来，我们监测了8个有围栏的大型围栏(1355平方米)和邻近的无围栏的控制区中排除地面哺乳动物的序列。在BCI上设置了4对处理和对照小区，在类似成熟林的吉甘特上设置了4对处理和对照小区。在开始任何处理之前，我们对所有地块中所有50厘米高的木本植物进行了测量、鉴定和标记。此外，我们在每个大样地的28个较小的小区中对新招募的苗木进行了量化。总体而言，我们现在已经跟踪了60,000多株树苗和树苗的命运。到目前为止的实验结果表明，至少在短期内(~10年)，排除哺乳动物可以显著增加苗层木本植物物种的多样性。因此，我们的结果与詹森-康奈尔假说在社区层面的预测形成鲜明对比。我们提出了一个新的概念模型，该模型表明，短期内(2-25年)哺乳动物从热带森林中消失的后果，无论是通过实验排除还是更广泛的狩猎，将首先导致木本植物物种多样性的增加，然后可能减少(见图1)。这一结果将首先出现在苗层(1-50厘米高)，然后以波的形式通过较大的尺寸级。然而，到目前为止的数据，即使在7年后，仍然是模棱两可的。围栏内记录清楚的多样性的短期增加可能是1998年厄尔尼诺南方涛动(ENSO)事件的短暂影响。具体地说，ENSO事件造成了严重和延长的旱季，增加了幼苗的死亡。由于围栏内的幼苗密度较高，这些地点可能已经缓解了干旱的影响。我们认为，当务之急是继续这项实验并监测下一阶段，因为苗层进入了小树苗级别。这项工作的保护意义是显而易见的；对许多热带森林保护区、公园和碎片中的哺乳动物动物的过度开发可能会对森林多样性产生长期的负面影响，这可能在几十年内不会变得明显。只有通过长期的实验，我们才能明确地区分短期动态和长期的森林轨迹。

0212054CarsonLTREBWe have underway a long-term, large-scale experiment that is addressing one of the primaryhypotheses proposed to explain the maintenance of tree species diversity in tropical forests, namely,that mammals acting as herbivores, seed predators, and dispersers, are critical determinants of tropicaltree diversity and abundance. This experiment is essentially a community-level test of the Janzen-ConnellHypothesis that predicts that the removal of ground-dwelling mammals will ultimately cause adecline in tropical tree diversity. Under this scenario, mammals preferentially prey on the juveniles oftree species that are abundant and ultimately keep these tree species in check.The study system is mature-phase tropical moist forest in central Panama on Barro ColoradoIsland (BCI) and the adjacent mainland (Gigante Peninsula). Since1994, we have monitored theconsequences of excluding ground-dwelling mammals in 8 large (1355m2) fenced exclosures andadjacent unfenced control plots. Four pairs of treatment and control plots were established on BCI and4 pairs were located on Gigante in similar mature forest. Prior to initiating any treatments, wecensused, measured, identified, and tagged all woody plants 50 cm in height in all plots. Inaddition, we quantified newly recruited seedlings in 28 smaller subplots in each large plot. In total,we have now followed the fate of more than 60,000 individual seedlings and saplings. Overall, theexclusion of mammals led to a substantial and sustained increase in seedling density.The results of this experiment to date suggest that excluding mammals may significantlyincrease woody species diversity of the seedling layer at least in the short-term (~10 yrs). Thus, ourresults contrast sharply with the community-level predictions of the Janzen-Connell hypothesis. Wepropose a new conceptual model that suggests that the consequences of the short term removal ofmammals (2-25 yrs) from tropical forests, whether by experimental exclusion or more broadly viaover hunting, will cause first an increase in woody species diversity and thereafter potentially adecrease (see Fig. 1). This result will first appear in the seedling layer (1-50 cm in height) and thenmove as a wave through larger size classes. The data to date, however, even after 7-years, remainequivocal. The short term increase in diversity clearly documented inside exclosure plots may be atransient effect of the 1998 El Nino Southern Oscillation (ENSO) event. Specifically, the ENSO eventcaused a severe and extended dry season that increased seedling mortality. Because there was a higherseedling density inside the exclosures, these sites may have been buffered from the drought. Webelieve that it is imperative that we continue this experiment and monitor the next phase as the seedlinglayer moves into the small sapling size class. The conservation implications of this work are clear; theoverexploitation of the mammalian fauna in many tropical forest reserves, parks, and fragments mayhave long-term negative consequences for forest diversity that may not become apparent for decades.It is only through long-term experiments that we can unequivocally separate out short-term dynamicsfrom long-term forest trajectories.