High-resolution paleoecological studies elucidate the effects of biotic and abiotic factors on the diversity and distribution of insect herbivory in deep time. Previous analyses of fossil leaf assemblages from intervals of significant environmental change have shown that insect damage frequency and diversity are strongly influenced by temperature and carbon dioxide levels. Similar studies that span intervals of environmental quiescence are a necessary complement and can help disentangle the relative influences of biological versus abiotic factors. The 21.73 ± 0.03 million year old (early Miocene) Mush Valley lagerstätten, Ethiopia, provides one such example. Insect damage censuses were conducted at six stratigraphic levels that span ~50,000 years and whose plant ecology has been studied in detail. A total of 2200 leaves (47 morphotypes) were examined and 35 damage types (DTs) observed. We documented variations in insect damage frequency, diversity, and composition among plant host taxa and across stratigraphic levels and tested for correlations with abiotic and biotic factors. Interspecific variations in damage frequency and diversity among host plants at Mush with >20 leaves in the damage census could not be explained solely by leaf mass per area, presence/absence of trichomes, symbiosis with nitrogen-fixing bacteria, or host plant abundance. The strongest predictor of insect herbivore damage on the bulk floras (all leaves) was the prevalence of legumes, whose symbiotic relationships with nitrogen-fixing bacteria likely made them more nutrient-rich. This pattern was also observed across the climatically-variable Paleocene-Eocene boundary in the Bighorn Basin, USA, suggesting that the soil microbiome, plant nutrient composition, and climate work synergistically in regulating insect herbivore distribution, population sizes, and feeding.

高分辨率古生态研究阐明了生物和非生物因素对昆虫食草动物多样性和分布的影响。先前对来自重大环境变化间隔的化石叶子组合的分析表明，昆虫损害频率和多样性受到温度和二氧化碳水平的强烈影响。跨越环境静止时间间隔的类似研究是必要的补充，可以帮助理清生物因素与非生物因素的相对影响。埃塞俄比亚 21.73 ± 03 万年前（中新世早期）的 Mush Valley 拉格斯泰滕就是这样一个例子。昆虫损害普查是在六个地层级别进行的，跨度约为 50,000 年，并对其植物生态进行了详细研究。总共检查了 2200 片叶子（47 种形态类型）并观察到 ​​35 种损伤类型（DT）。我们记录了植物宿主分类群之间和跨地层水平的昆虫损害频率、多样性和组成的变化，并测试了与非生物和生物因素的相关性。在损害普查中，> 20 片叶子的 Mush 寄主植物之间的损害频率和多样性的种间差异不能仅用单位面积的叶质量、毛状体的存在/不存在、与固氮细菌的共生或寄主植物丰度来解释。昆虫食草动物对大量植物群（所有叶子）造成损害的最强预测因素是豆类的盛行，豆类与固氮细菌的共生关系可能使它们的营养更加丰富。