Even some ardent¹ conservationists acknowledge that the diversity of life on Earth cannot be fully² sustained as human populations expand use more resources nudge the climate and move weedlike pests and predators³ from place to place.

Given that some losses are inevitable⁴ the debate among many experts has shifted to an uncomfortable subject what level of loss is acceptable. The discussion is taking place at both the local and global levels How small can a fragment of an ecosystem⁵ be and still function in all its richness and thus be considered preserved﹖ And as global biodiversity diminishes is it a valid⁶ fallback strategy to bank organisms and genes⁷ in zoos DNA⁹ banks or the like or does this simply justify¹⁰ more habitat destruction﹖ Is nature on ice a sufficient substitute for the real thing﹖ Some conservation groups have strenuously¹¹ avoided or even attacked such calculations and strategies. They say there is no safe diminution¹² of habitat as long as human understanding of ecology is as sketchy¹³ as it is a fallback strategy is unthinkable. Furthermore banking¹⁴ nature in a deep freeze or database of gene⁸ sequences cannot capture context. For instance even if a vanished bird was someday reconstituted from its genes would it warble with the same fluency¹⁵ as its ancestors﹖ On the other side of the debate those considering what the smallest viable¹⁶ habitats are or how to expand archives as an insurance policy say that recent trends have proved that old conservation strategies are no longer sufficient. A few decades ago the issue seemed fairly uncomplicated identify biological “hot spots” or species of concern and establish as many reserves as possible. But the picture has grown murky¹⁷.

Twentyfour years ago Dr. Thomas E. Lovejoy and other biologists began a remarkable¹⁸ experiment on the fasteroding fringe of rain forest near the Brazilian city of Manaus. They established 11 forest tracts¹⁹ ranging from 2.5 to 250 acres each surrounded by an isolating²⁰ sea of pasture similar to what is advancing around most other tropical forests. Among the many findings an analysis published last week on birds in the lower layers of greenery found that it would take a fragment measuring at least 2 500 acres—10 times as large as the biggest one in the experiment—to prevent a decline of 50 percent in those bird varieties in just 15 years or so.

In the understated language of science the new study in The Proceedings²¹ of the National Academy of Sciences concludes “This is unfortunate when one considers that for some speciesrich areas of the planet a large proportion of remaining forest is in fragments smaller than 2 500 acres.”

In the face of this and other evidence a growing group of conservation biologists say try everything at the same time. “Clearly the most effective way to protect biodiversity is to protect natural areas ” said Dr. Peter H. Raven²² the director of the Missouri Botanical Garden “and to find those organisms most endangered in nature and somehow protect them in typeculture collections botanical gardens zoos seed banks or whatever.” But most important he said is to find ways to limit human pressures on the world’s last wild places by slowing population growth and using resources more efficiently²³. One pioneer of genetic²⁴ deconstruction Dr. J. Craig Venter agrees with Dr. Raven. Dr. Venter has moved from sequencing the DNA of humans and other species to assaying genes in entire ecosystems²⁵ most recently the waters of the Sargasso Sea. In five 50gallon samples gathered in February he said his team had found 1 million distinct genes quite a haul compared with the 26 000 or so of a human being. And that is the tiniest scratch in the surface he added. His is one effort among many. Britain has a Millennium²⁶ Seed Bank a growing archive of all the country’s plants. The San Diego Zoo has its parallel Frozen Zoo an archive of thousands of DNA samples and cell lines from a host of species. Nonetheless given the overwhelming complexity²⁷ of nature Dr. Venter added “we’re better off trying to preserve the diversity of what we have rather than trying to regenerate²⁸ it in the future.”

连一些积极的自然资源保护论者都承认，随着人口的膨胀、消耗更多的自然资源、引起气候的变化，以及造成大量害虫和捕食动物的迁移等，地球上的生物多样性肯定不会完全地持续下去。

许多专家的争论焦点已经转到了一个令人不安的话题，假使一些损失是不可避免的，多大程度的损失是可以容忍的呢？关于这个问题的辩论在局部和全球范围两个层面上同时展开：生态系统的一小部分可以小到何种程度仍能维持其完整、丰富的功能，从而可以认为是受到了保护呢？在全球生物多样性减少的过程中，把生物有机体和基因保存在诸如动物园、基因库之类的地方是一种有效的保全策略吗？或者这样做仅仅为更多的(动、植物)栖息地的破坏提供了借口？冷藏的自然能够充分地代替真正的自然吗？一些自然资源保护组织一直极力避开甚至反对这样的推论和策略。他们说，只要人类对生态系统的认识还是一知半解的，那就不存在对栖息地的安全缩小；因此也就谈不上什么保险策略了。更何况，把自然生态深冻起来或者将其存入基因序列数据库并不能保存与其相关的背景。比如，就算一只灭绝的鸟儿某一天被人们从它的基因中重新组合出来，它的啁啾声能像其先辈们一样婉转动听吗？另一方面，寻求最小可行栖地的人们或试图扩大现有档案库作为一种保全策略的人们则说，最近的趋势已经表明，旧的保护策略不再够用了。几十年之前，问题似乎还不是那么复杂：只要确认出那些受到威胁的生态地区或者令人担忧的物种，然后建立尽可能多的保护区就是了。然而，目前这个状况已经变得模糊起来。

24年以前，托马斯?E?洛夫乔伊博士和其他一些生物学家在巴西马瑙斯市附近遭受快速侵蚀的热带雨林地区边缘开始了一项备受瞩目的实验。他们建立了11块森林试验区，面积大小从2.5英亩到250英亩不等，每一块都被一片分割开来的广袤草场包围起来，这些草场与正在向大多数其他热带雨林周边推进的草场相类似。在大量的调查结果中，上周出版的一篇关于生活在绿地较低层的鸟类的分析报告发现，至少需要一块2500英亩的森林区域——相当于实验中划出来的最大的一块试验区的10倍——才能防止那些鸟类的品种在仅仅15年左右的时间里减少50%。

美国国家科学院学报》上新刊登的一篇专题研究论文用毫不夸张的科学语言总结道：“当你考虑到在地球上一些物种资源丰富的地区，留存下来的森林中一大部分是小于2500英亩的分散小块时，这是多么不幸。”

面对这些情况和其他证据，愈来愈多的自然资源保护生物学家说：应该同时尝试所有可行的办法。美国 密苏里州植物园园长彼得?H?雷文博士说：“显然，最有效的保护生物多样性的方法是保护自然栖息地，同时，还要找出那些自然界中最为濒危的物种，用某种方法把它们保护起来，比如，把它们放入物种培育采集库、植物园、动物园、种子银行等诸如此类的地方。”他说，但最为重要的是通过减缓人口增长和更有效地利用资源找到减少人类对世界最后原始生态地区的压力的方法。遗传解构学的先驱之一，J?克雷格?文特尔博士同意这一看法。文特尔博士从对人类和其他物种的DNA基因排序的研究转到了对整个生态系统的基因分析，最近开始了对马尾藻海海水的研究。他说，他的小组在2月份(指2003年2月份——译者)收集的5份50加仑的样本中发现了上百万种不同的基因类型，这与人类个体具有的约26000种基因相比实在是太多了。他补充说，这不过才触及到皮毛而已。他所做的只是许多努力中的一部分。英国有一个“千禧年种子银行”。它不断扩大，收藏了该国所有植物。(美国)圣地亚哥动物园有一座与其相应的“冷藏动物园”，其中保存了许多物种的数千个DNA样本和细胞株的资料。即使如此，考虑到自然界极为复杂，文特尔博士补充说：“如果我们尽力保护好现存生物的多样性，而不是试图在将来去重新创造它，我们才更明智。”