A. Sandionigi , Saverio Vicario, E. M. Prosdocimi, A. Galimberti, A. Ferri, A. Bruno, Bachir Balech, V. Mezzasalma, M. Casiraghi
Biological diversity was early identified in ecological studies as an important emergent property of natural systems, linked to the ability to persist through time. Unfortunately the difficulties in identifying common metrics for diversity contributed to produce a vast literature with unclear conclusion. Within an information theoretical framework it is possible to define diversity as the exponential of Shannon entropy and thereby obtain a self consistent measure of overall within and between group diversity. This framework was recently enlarged to take into account that biological entities (lineage, guild, species, or group of individuals) are differently diverse by using tree shaped similarity structure, allowing to define phylogenetic diversity within the frame of information theory (Chao A, Chiu CH, Jost 2010; Marcon and Herault 2014: Sandionigi A et al 2014).
21 pairs of larva bees and its parassite were sampled across 7 beehives in northern Italy. The microbiomes were extracted from each biological sample and V3 SSU region was amplified and sequenced with 454 FLX technologies. The reads were de-noised using ACACIA and the chimera detector of USEARCH. We compared the phylogenetic diversity approach as implemented in our tool PhyLOH with the distance approach (Jaccard, Chord, wUnifrac, Unifrac) implemented in QIIME taking in account or not phylogenetic information and abundance.
We tested phyloH to characterize microbiome in the honeybee (Apis mellifera, Insecta: Hymenoptera) and its parasitic mite varroa (Varroa destructor, Arachnida: Parasitiformes). The rationale is that the comparative analysis of honeybee and varroa microbiomes could open new perspectives concerning the role of the parasites on honeybee colonies health. Our results showed a dramatic change of the honeybee microbiome when varroa occurs, suggesting that this parasite is able to influence host microbiome. Among the different approaches used, only the entropy method, in conjunction with phylogenetic constraint as implemented in phyloH, was able to discriminate varroa microbiome from that of parasitized honeybees.
The phylogenetic diversity approach seems more sensitive than distance method and allow a more fine grain data exploration by inspecting the branch contribution to beta diversity. The method need to be enlarged to take in account phylogenetic uncertainty.