While many molecular studies have documented arbuscular mycorrhizal fungi (AMF) communities in temperate ecosystems, very few studies exist in which molecular techniques have been used to study tropical AMF communities. Understanding the composition of AMF communities in tropical areas gains special relevance as crop productivity in typically low fertility tropical soils can be improved with the use of AMF. We used a hierarchical sampling approach in which we sampled soil from cocoa (Theobroma cacao L.) plantations nested in localities, and in which localities were nested within each of three regions of Côte d’Ivoire. This sampling strategy, combined with 18S rRNA gene sequencing and a dedicated de novo OTU-picking model, allowed us to study AMF community composition and how it is influenced at different geographical scales and across environmental gradients. Several factors, including pH, influenced overall AMF alpha diversity and differential abundance of specific taxa and families of the Glomeromycotina. Assemblages and diversity metrics at the local scale did not reliably predict those at regional scales. The amount of variation explained by soil, climate, and geography variables left a large proportion of the variance to be explained by other processes, likely happening at smaller scales than the ones considered in this study. Gaining a better understanding of processes involved in shaping tropical AMF community composition and AMF establishment are much needed and could allow for the development of sustainable, productive tropical agroecosystems.

尽管许多分子研究已记录了温带生态系统中的丛枝菌根真菌（AMF）群落，但很少有分子研究被用于研究热带AMF群落的研究。了解热带地区的AMF群落组成具有特殊的意义，因为使用AMF可以提高通常在低肥力热带土壤中的作物生产力。我们使用了一种分层采样方法，在该方法中，我们从可可粉中提取了土壤（Theobroma cacaoL.）人工林筑巢在当地，科特迪瓦三个区域中的每个地方都筑有巢穴。这种采样策略，结合18S rRNA基因测序和专用的从头OTU筛选模型，使我们能够研究AMF群落组成及其在不同地理尺度和环境梯度上的影响。包括pH在内的几个因素影响了整个AMFα多样性以及特定分类群和Glomeromycotina家族的差异丰度。当地规模的集合体和多样性指标不能可靠地预测区域规模的集合体和多样性指标。由土壤，气候和地理变量解释的变化量留下了很大一部分方差，而其他过程则可以解释，可能发生的规模小于本研究中考虑的规模。