What can a flock tell us about HPAI?从一群水鸟中，我们能看出什么 HPAI 风险信息？

What makes some waterbirds more likely hosts?为什么有些水鸟更容易成为 AIV 宿主？

We started by comparing 143 waterbird species found in Europe. We combined historical AIV records with information on diet, body size, reproduction and movement-related traits. Because closely related species share many characteristics, our statistical comparisons also accounted for evolutionary relationships [1].我们首先比较了欧洲 143 种水鸟。我们将历史 AIV （禽流感）监测记录与鸟类的食性、体型、繁殖和移动相关性状结合起来分析。由于近缘物种往往共享许多生物特征，我们的统计分析也纳入了系统发育关系，而不是把近缘物种当作完全独立的样本 [1]。

Of the 143 species, 82 had been recorded as AIV hosts by 2021. Larger body size, earlier maturity and wing shape associated with greater dispersal ability were linked to host status. Greater dispersal ability was also associated with more recorded virus subtypes, together with larger clutch size and longer lifespan [1].截至 2021 年，143 种水鸟中有 82 种曾被记录为 AIV 宿主。体型更大、性成熟更早，以及与较强扩散能力相关的翅形（hand-wing index），都与宿主状态有关。更强的扩散能力还与记录到更多病毒亚型相关；窝卵数较大和寿命较长的物种，也往往记录到更多 AIV 亚型 [1]。

This species-level analysis covered avian influenza viruses broadly, not only HPAI H5 viruses, and host status alone does not tell us how efficiently a species transmits a virus. Still, these patterns show that ecological and life-history traits can help explain why infection records are unevenly distributed across waterbird species [1,4].需要注意的是，这一物种层面的分析针对的是 AIV 整体，并不只局限于 HPAI H5；“被记录为宿主”本身，也不能直接说明一个物种传播病毒的效率有多高。但这些结果表明，水鸟之间的感染记录并不是随机分布的，生态和生活史性状可以解释其中的一部分差异 [1,4]。

The company birds keep matters鸟和谁生活在一起，也很重要

Birds do not live in isolation. To investigate whole communities, we linked winter waterbird counts from more than 7,000 sites across 26 European countries with reported HPAI outbreaks and environmental information, including weather, waterbodies and poultry density. We focused on November-March outbreaks to match the bird survey season [1-3].鸟类并不是彼此孤立地生活。为了从单个物种走向整个群落，我们将欧洲 26 个国家 7,000 多个地点的冬季水鸟计数，与报告的 HPAI 疫情及天气、水体、家禽密度等环境信息对应起来。分析集中在 11 月至次年 3 月，以与水鸟调查季节保持一致 [1-3]。

Wild-bird outbreaks were more often reported where waterbirds were abundant, and where key species such as Mallard, Eurasian Wigeon, Tufted Duck and Mute Swan were more abundant. But numbers were only part of the story. Communities containing more distantly related species were consistently less likely to have HPAI outbreaks [2,3].我们发现，在水鸟总量更高、以及绿头鸭、赤颈鸭、凤头潜鸭和疣鼻天鹅等关键物种更丰富的地点，更常监测到野鸟 HPAI 疫情。但“数量”只解释了故事的一部分。系统发育多样性更高，也就是群落中的物种在演化上彼此更远的地点，HPAI 疫情的概率反而持续较低 [2,3]。

The trait-based study added another layer. Greater differences in diet were associated with lower HPAI occurrence. One possible explanation is that species with similar diets or habitat use encounter one another more often, while close relatives can also share biological characteristics that make cross-species infection easier [1-3].性状研究又补上了一层证据：食性差异更大的群落，HPAI 发生率更低。一种可能的解释是，食性或栖息地利用更相似的物种更容易在同一地点接触；而亲缘关系较近的物种，也可能共享一些有利于跨物种感染的生物学特征 [1-3]。

This is not simply “more species means less disease”. A wetland can contain many species that are nevertheless closely related and share similar ecological traits, or fewer species that are more distantly related and functionally different but in high abundance. Thus, which birds live together matters, not just how many [2,3].因此，我们的发现并不简单的是“物种越多，疾病越少”。一个湿地可以有很多物种，但这些物种彼此可能很近缘、生态性状也很相似；反过来，一个物种数较少的群落，也可能由亲缘关系更远、功能差异更大的物种组成，同时个体总量依然很高。真正重要的不只是“有多少种鸟”，还包括“哪些鸟生活在一起” [2,3]。

Do the same patterns hold across epidemics?这些规律能否跨越不同疫情？

In our 2024 study, a model developed from the 2016/17 epidemic successfully distinguished HPAI outbreak sites from comparison sites in the 2020/21 epidemic. Our 2025 study extended that test that models trained on the 2016/17 and 2020/21 H5N8-dominated epidemics also performed well when applied to H5N1 records from 2021/22 [2,3].在我们 2024 年的研究中，用 2016/17 疫情建立的模型，能够较好地区分 2020/21 疫情中的 HPAI 空间分布，而2025 年的研究进一步扩展了这一检验：用 2016/17 和 2020/21 两轮以 H5N8 为主的疫情训练的模型，在应用到 2021/22 的 H5N1 记录时仍然表现良好 [2,3]。

Across four epidemic periods from 2005/06 to 2021/22, associations with waterbird abundance and phylogenetic diversity were broadly consistent in wild birds. The models identify spatial conditions associated with reported outbreaks. The same consistency, however, did not appear in poultry. Waterbird community variables were weaker and changed between epidemics, while agro-climatic and poultry-related factors played larger roles, likely reflecting differences in dominant hosts and transmission pathways within farmed birds. This matters because predictors of HPAI in wild birds should not be assumed to transfer directly to poultry [2,3].从 2005/06 到 2021/22，跨越四轮疫情，水鸟丰度和系统发育多样性与野鸟 HPAI 的关联总体上保持一致。但是，我们并没有在家禽中看到同样的模型稳定性。水鸟群落变量的作用更弱，而且在不同疫情之间变化较大；相比之下，农业气候和家禽相关因素的作用更突出。这可能反映了养殖鸟类中主导宿主和传播路径与野鸟系统并不相同。因此，用于解释野鸟 HPAI 的风险因子，不应直接套用到家禽疫情上 [2,3]。

From bird counts to better surveillance从水鸟计数走向更好的监测

Together, these studies suggest that routine bird surveys could do more than track populations. Combined with virus surveillance and environmental data, they can provide ecological information for identifying places where closer HPAI monitoring may be most useful and for building early-warning maps [2,3].把这些研究放在一起看，常规水鸟调查的价值可能不只是追踪种群变化。与病毒监测和环境数据结合后，这些调查可以提供重要的生态信息，帮助识别哪些地点更值得加强 HPAI （或者其他疾病？）监测，并用于构建早期预警风险地图 [2,3]。

Complementary work led by Johanna Harvey (Assistant Professor from the University of Rhode Island), which we co-authored, adds a conservation perspective. A species’ ability to contribute to transmission is not the same as its vulnerability to severe disease or mortality. Considering both dimensions and recognizing the large gaps in species-specific infection data, can help separate species that may be important for transmission from those that are priorities for conservation [4].我们共同参与的、由 Johanna Harvey 博士（美国罗德岛大学助理教授）领衔的另一项研究，则补充了一个保护生物学视角：一个物种在传播中的作用，并不等于它遭受重症或死亡的脆弱性。把“传播能力”和“感染后脆弱性”这两个维度放在一起看，同时正视不同物种感染数据仍存在大量缺口，可以帮助我们区分哪些物种可能对传播更重要，哪些物种更需要优先保护 [4]。

There are limits to what we can infer. These studies are based largely on observational associations, but passive outbreak surveillance misses infections, and some later analyses relied on older bird counts. The European winter patterns also need testing in other regions and seasons [1-3].这些研究也有明确的局限。结论主要来自观察性关联,被动疫情监测会漏掉部分感染记录,而后期分析中部分水鸟数据也来自较早年份。此外，这些欧洲冬季的规律，还需要在其他地区和季节中继续检验 [1-3]。

The broader message is that a flock is more than a number. Its species composition, the traits those species carry, and their evolutionary and ecological relationships contain information about HPAI risk. Bringing those data into surveillance gives us a more biologically informed way to decide where and in which species to look more closely [1-4].更广义地说，一个鸟群远不只是一个数量。它由哪些物种组成，这些物种具有什么性状，以及它们之间的演化和生态关系，都包含着 HPAI 风险的信息。把这些生态学信息纳入监测，可以让我们更有依据地判断，应该在哪些地点、哪些物种上投入更多关注 [1-4]。