对生物多样性保护的更多关注是明确和紧急的5。世界正在进入其第六次巨大灭绝事件6，这是一个归因于单一物种的第一个：同性恋。生物多样性损失是全球关注的关注点，也是多个国际承诺的重点，包括最近在《联合国生物多样性公约》的《昆明 - 蒙特利尔全球生物多样性框架》中承诺的那些。然而，物种灭绝继续累积7。

　　保护工作可以阻止物种灭绝8,9，但资金仍然不足10,11。此外，如果相对于风险12的努力针对的努力很差，那么节省的物种将比其他可能性更少。可以通过增加资源和更有效的可用资源分配来改善生物多样性的前景。更有效的分配要求我们确定降低物种灭绝风险的保护干预措施，以及已实施的干预措施以及已实施的干预措施。

　　直到最近，很少关注评估保护中有效的方法2,13。对保护区（PA）的有效性评估是一个值得注意的例外，多项研究发现，管理良好的PAS减轻生物多样性损失14,15。同样，广泛的数据证明了对岛生物群岛的侵入性物种的好处16。但是，还有许多其他类型的保护干预措施17，其中许多尚未进行评估和大规模评估。

　　国际自然保护联盟（IUCN）通过其红色威胁物种清单1,18进行物种级别的评估。评估人员还汇总了针对每个物种实施的保护干预措施的信息。研究使用这些数据来确定与鸟类灭绝风险降低相关的干预措施，哺乳动物8,19,20。在这里，我们通过手动审查其“到位的保护行动”文本的手动审查，并结合其他四个有关特定干预措施的数据库，对跨地区和分类群体的保护行动模式进行全球评估：Life21（https：//mol.org）；保护区域的世界数据库22（WDPA）；物种+（https://speciesplus.net）;以及岛屿入侵物种消除的数据库23（DIES）。我们考虑在分类学家族中所有5,963种陆地威胁物种（被列为脆弱，濒危或濒危濒危的物种），这些家族已经过全面评估红色清单（补充表1）。

　　我们最初专注于生物多样性的三个最大威胁：栖息地丧失（包括栖息地退化），国际贸易和入侵物种的过度开发4。每个人都有明确匹配的保护干预措施：栖息地保护（通过PAS），贸易控制和入侵物种控制。尽管我们认识到国内过度开发的潜在严重影响24，但我们仅关注国际贸易，鉴于对这种威胁及其匹配的干预措施的数据有更大的可用性。

　　关于栖息地的丧失，我们认为只有在符合重叠的特定物种代表阈值的阈值的阈值（方法）（方法）的物种保护评分所定义的情况下，才能实现有意义的栖息地保护。物种保护评分有助于物种保护指数，这是联合国全球生物多样性框架的指标，该框架涉及物种代表21,25。即使在PAS内实现了足够的代表，我们还指出，物种仍然可能受到栖息地损失的威胁，在这种情况下，PA的管理不善或受到降级，缩小尺寸或Degazettement26的降级。我们还承认除了PA之外的干预措施的价值，包括其他有效的基于区域的保护措施27，空间规划28和栖息地恢复。但是，我们缺乏相当全面的数据，无法在此处包含它们。

　　对于受到上面概述的生物多样性损失的三个主要驱动因素中的任何一个威胁的物种，我们从两个关键问题开始：（1）哪些比例的物种接受了适当的保护干预措施？（2）物种的分类法，生物地理或灭绝风险会影响进行适当干预的可能性吗？最后，我们考虑了Red List17记录的所有原位保护干预措施（包括重新引入，国际立法和教育）的所有类别，以识别未经证实的保护干预措施的威胁物种。对于缺乏保存干预措施的物种，我们探讨了明显缺乏保护注意力是否与红色列表中物种状态的变化在质量上相关。

　　在全球范围内，我们发现对世界威胁陆地物种的保存干预措施有实质性的缺点。大多数受到国际贸易过度开发的风险的威胁物种记录在国际贸易控制下（76％）。但是，在受到栖息地丧失威胁的人中，​​只有9％的人在PAS中有足够的栖息地来满足目标阈值。如其他地方所述，在当前的PA网络中，小型物种的代表特别差26,29,30。在主要受入侵物种威胁的物种中，只有15％的物种被记录为接受入侵物种控制（图1）。如果我们放宽栖息地保护的标准至少仅包括一个PA，我们发现涵盖了75％受栖息地丧失威胁的物种（补充图2和3）。某些分类偏见也存在于物种的威胁被适当解决的程度上。For example, 0.76% and 0% of threatened flowering plants in the class Magnoliopsida (Fig. 1) are documented as receiving meaningful habitat protection or invasive species control, respectively (Fig. 1), whereas 14%, 63% and 46% of threatened birds are documented as receiving the appropriate interventions to tackle habitat loss, international trade and invasive species, respectively.

　　缺乏适当类型的保护干预措施的物种的分布显示出很大的地理变异性（P< 0.001) (Fig. 2 and Extended Data Figs. 1–4). Several regions contain high numbers of amphibians requiring invasive species control (P < 0.001), including Madagascar, Central America and Australia (Fig. 2c). The majority of cases (74%) in which control of invasive species is needed but lacking pertain to a lack of control of chytrid fungus, Batrachochytrium dendrobatidis, for which there is not yet an effective treatment31.

　　With respect to habitat loss, amphibians lack meaningful habitat protection in Central America (Fig. 2a) and mammals are notably lacking in habitat protection in Indonesia (Fig. 2g), as are birds in South America, Central America and Indonesia (Fig. 2d). Most species threatened by international trade are documented as receiving some international trade control (Fig. 1b), although relatively high numbers of exploited birds in Indonesia seemingly lack such protection (Fig. 2e and Extended Data Fig. 1).

　　Finally, we assessed how many of the threatened species in our database are not documented as receiving meaningful habitat protection or any of the other categories of conservation intervention, expanded to include additional measures such as reintroduction17. Overall, we find that 58% (3,467 out of 5,963) of threatened terrestrial species lack meaningful habitat protection or any other documented conservation interventions (Fig. 3). We emphasize that this percentage is derived on the basis of Map of Life’s Species Protection Score (https://mol.org/indicators), which uses a species-specific representation threshold to determine whether a species has meaningful representation in PAs21. When we relax that threshold to whether a species occurs in any PA to any extent (as determined by the Red List), the proportion falls to 19% (1,105 out of 5,963; Supplementary Figs.  4 and 5). Taxa such as amphibians fare worse than others (deviance = 1,533, d.f. = 7, P < 0.001), with 84% lacking meaningful habitat protection or any other documented conservation interventions, compared with 44% for threatened birds. Taxonomic biases probably result from increased attention to charismatic and easily studied groups12,32,33, which also translates to more frequent Red List assessments. Across some (but not all) taxa, species at greatest risk of extinction are more likely to have documented attention (deviance = 19, d.f. = 1, P < 0.001). This corroborates Luther et al.19, who concluded that species at greater risk of extinction receive more conservation attention.

　　Although many threatened species are beneficiaries of documented conservation attention, whether those interventions work remains a critical question. From 2006 to 2020, 279 species were uplisted to a higher threat category and 41 species were downlisted to a lower threat category owing to a genuine increase and decrease in extinction risk, respectively (Extended Data Fig. 8). Of the downlisted species, only 15% (6 out of 41) lacked any documented conservation interventions. Alarmingly, 67% (187 out of 279) of uplisted species had received at least some documented conservation attention, suggesting that the measures used were insufficient to reverse declines. Focusing on species facing only one major threat (habitat loss, international trade or invasive species), a higher proportion were uplisted when the appropriate intervention was apparently lacking, although this was not consistently the case (Fig. 4). Thus, although conservation interventions are qualitatively associated with improvements in species’ Red List status, corroborating that conservation can succeed8,9, the mere existence of ‘something rather than nothing’ is not sufficient to reverse declines. Previous studies have documented large variation in how well conservation interventions are implemented, such as variation in PA management34. Additionally, ecological time lags occur in response to both positive and negative change35, and there are time lags inherent to the Red List process itself. Species must have met the criteria for a lower threat category for at least five years before the status change is implemented36. Few taxa were reassessed in the period 2006–2020, and birds were reassessed more frequently (approximately every 4 years) than any other group.

　　Given the geographic patterns in conservation attention, it is possible that certain groups of species are disproportionately neglected. We found a weak trend whereby ‘evolutionary distinctiveness’—a measure of species’ relative contribution to phylogenetic diversity37—was lower in species lacking meaningful habitat protection or any other documented conservation interventions, compared with species with at least one documented intervention (Extended Data Fig. 9; deviance = 4, d.f. = 1, P = 0.05). The total number of endemic threatened species with documented conservation interventions was positively correlated with the number that lacked such interventions (Extended Data Fig. 6). This may suggest that apparent neglect does not result from lack of will, but rather from insufficient capacity to act or to report for countries with greater numbers of threatened endemic species. However, the relationship was not straightforward (Extended Data Fig. 5), as the probability of apparent neglect for any given threatened endemic species was not statistically associated with the gross domestic product (GDP) of the country (deviance = 0, d.f. = 1, P > 0.05) or total number of endemic threatened species (deviance = 0, d.f. = 1, P >0.05;扩展数据图7）。

　　分开观察到的地理模式（图3）以辨别因果关系，需要对保护资源进行更细微的评估，这些评估又是国际援助和干预措施的充分性34。同样重要的是要注意，即使有详细的评估指南以及评估者报告中的广泛培训，地理和分类偏见，也存在38。这不会破坏我们的结果，因为至关重要的是科学家可以确定需要改进的文档。但是，这确实使解释复杂化，因为我们目前无法区分文档的需求和真正缺乏保护干预措施。

　　展望未来，旨在跟踪物种恢复的物种的绿色地位39，再加上扩大的保护证据数据库3，应该使研究人员能够更好地探索红色列表上某些物种的积极趋势。此外，到2030年，有50多个国家的承诺将扩大全球PA网络至少包含30％的陆地和海洋生态系统，这为保护那些受到栖息地损失威胁的物种提供了机会，这些物种目前在PAS40中代表性不足。针对PAS扩展的重要地点包括关键的生物多样性区域，其中大多数是根据他们支持的威胁物种的种群确定的41。有效确保足够物种代表的方法也将是补充基于区域保护的现有方法的关键42。在传统的PA之外，其他有效的基于地区的保护措​​施将在保护受威胁物种27,43中发挥重要作用，尤其是在土著社区拥有和管理的土地上。但是，改善全球保护实践的关键第一步是记录我们已经在做什么。在全球范围内，记录的物种中存在分类学和地理偏见，这些物种受到保护的关注，有限的物种被降低到较低类别的威胁。保护可以成功，但是如果没有更多和更好的针对性投资，我们可能会冒着投降世界威胁物种大规模灭绝的风险。