The diversity of species inhabiting freshwater relative to marine habitats is striking, given that freshwater habitats encompass <1% of Earth's water. The most commonly proposed explanation for this pattern is that freshwater habitats are more fragmented than marine habitats, allowing more opportunities for allopatric speciation and thus increased diversification rates in freshwater. However, speciation may be generally faster in sympatry than in allopatry, as illustrated by lacustrine radiations such as African cichlids. Differences between rivers and lakes may be important to consider when comparing diversification broadly among freshwater and marine lineages. Here I compared diversification rates of teleost fishes in marine, riverine and lacustrine habitats. I found that lakes had faster speciation and net diversification rates than other aquatic habitats. However, most freshwater diversity arose in rivers. Surprisingly, riverine and marine habitats had similar rates of net diversification. Biogeographic models suggest that lacustrine habitats are evolutionary unstable, explaining the dearth of lacustrine species in spite of their rapid diversification. Collectively, these results suggest that diversification rate differences are unlikely to explain the freshwater paradox. Instead, this pattern may be attributable to the comparable amount of time spent in riverine and marine habitats over the 200-million-year history of teleosts.

中文翻译：

---

比较湖泊，河流和海洋中的多样化率

鉴于淡水栖息地占地球水的1％以下，因此相对于海洋栖息地而言，居住在淡水中的物种的多样性令人震惊。关于这种模式的最普遍的解释是，淡水生境比海洋生境更加零散，从而为异源物种形成提供了更多机会，从而增加了淡水的多样化率。但是，如非洲湖丽鱼科鱼等的放射线所表明的那样，共生形态中的物种形成通常要比同异相的物种形成更快。在广泛比较淡水和海洋世系之间的多样性时，可能需要考虑江湖之间的差异。在这里，我比较了海洋，河流和湖泊栖息地中硬骨鱼类的多样化率。我发现湖泊比其他水生生境具有更快的物种形成和净多样化速度。但是，大多数淡水多样性出现在河流中。令人惊讶的是，河流和海洋生境的净多样化率相似。生物地理学模型表明，湖泊栖息地是进化不稳定的，这解释了尽管湖泊物种迅速多样化，却缺乏它们。总的来说，这些结果表明多样化率差异不可能解释淡水悖论。相反，这种模式可能是由于在2亿年硬骨鱼类的历史中，在河流和海洋栖息地所花费的时间相当。生物地理学模型表明，湖泊栖息地是进化不稳定的，这解释了尽管湖泊物种迅速多样化，却缺乏它们。总的来说，这些结果表明多样化率差异不可能解释淡水悖论。相反，这种模式可能是由于在2亿年硬骨鱼类的历史中，在河流和海洋栖息地所花费的时间相当。生物地理学模型表明，湖泊栖息地是进化不稳定的，这解释了尽管湖泊物种迅速多样化，却缺乏它们。总的来说，这些结果表明多样化率差异不可能解释淡水悖论。相反，这种模式可能是由于在2亿年硬骨鱼类的历史中，在河流和海洋栖息地所花费的时间相当。