Microbiota Influences Fitness and Timing of Reproduction in the Fruit Fly Drosophila melanogaster

Abstract

Associated microorganisms ("microbiota") play a central role in determining many animals' survival and reproduction characteristics. The impact of these microbial influences on an animal's fitness, or population growth, in a given environment has not been defined as clearly. We focused on microbiota-dependent host fitness by measuring life span and fecundity in Drosophila melanogaster fruit flies reared individually with 14 different bacterial species. Consistent with previous observations, the different bacteria significantly influenced the timing of fly life span and fecundity. Using Leslie matrices, we show that fly fitness was lowest when the microbes caused the flies to invest in life span over fecundity. Computational permutations showed that the positive fitness effect of investing in reproduction was reversed if fly survival over time was low, indicating that the observed fitness influences of the microbes could be context dependent. Finally, we showed that fly fitness is not influenced by bacterial genes that shape fly life span or fly triglyceride content, a trait that is related to fly survival and reproduction. Also, metagenome-wide association did not identify any microbial genes that were associated with variation in fly fitness. Therefore, the bacterial genetic basis for influencing fly fitness remains unknown. We conclude that bacteria influence a fly's reproductive timing more than total reproductive output and that (e.g., environmental) conditions that influence fly survival likely determine which bacteria benefit fly fitness. IMPORTANCE The ability of associated microorganisms ("microbiota") to influence animal life history traits has been recognized and investigated, especially in the past 2 decades. For many microbial communities, there is not always a clear definition of whether the microbiota or its members are beneficial, pathogenic, or relatively neutral to their hosts' fitness. In this study, we report the influence of individual members of the microbiota on Drosophila melanogaster fitness using Leslie matrices that combine the microbial influences on fly survival and reproduction into a single fitness measure. Our results are consistent with a previous report that, in the laboratory, acetic acid bacteria are more beneficial to the flies than many strains of lactic acid bacteria. We add to the previous finding by showing that this benefit depends on fly survival rate. Together, our work helps to show how the microbiota of a fly influences its laboratory fitness and how these effects may translate to a wild setting.

Keywords: Drosophila melanogaster; Leslie matrix; acetic acid bacteria; fitness; lactic acid bacteria; microbiota.

FIG 1

Natural microbiota isolates influence D. melanogaster fecundity, life span, and fitness. (A) For each of the microbial treatments, the average number of offspring produced per D. melanogaster female each day, measured in three time-staggered experiments each with twice weekly intervals across the life cycle. (B) The average total estimated offspring produced per female. Significant differences between treatments were defined by a Kruskal-Wallis test. (C) Fly life span when reared with the same microorganisms as in the previous panels, and significant differences were determined by a log rank test. (D) The effect of fly survival during the first reproductive period on fitness lambda, calculated as the first eigenvalue from a Leslie matrix. Colors are distributed according to the high-level taxonomic classification of the microbial treatment as follows: red, AAB; blue, LAB; purple, Bacillus subtilis; green, Enterobacteriales; magenta, 5-species; black, axenic. Significant differences between treatments were determined by a Kruskal-Wallis test. In all panels, significant differences between treatments are shown by compact letter displays.

FIG 2

Changes in D. melanogaster survival and fecundity differentially affect fly fitness. The relative contributions of microbiota influences on D. melanogaster fitness were determined by permuting fly survival (A, C) and fecundity (B, D) to 0.5 (A, B) or 0.00001 (C, D) of observed values. Each point represents a single vial, and values were permuted on a per-vial basis. Points are shaded according to the high-level taxonomic classification of the microbial treatment the vial received as follows: red, AAB; blue, LAB; magenta, B. subtilis; cyan, Enterobacteriales; purple, 5-species; black, axenic. The correlation between observed and permuted fitness values was calculated. P, P value; rho, Spearman’s rho. Data from vials containing flies reared with P. putida, which had characteristics of a fly pathogen, were omitted from these analyses.

FIG 3

Bacterial mutants that influence Drosophila life history traits do not influence fly fitness. Lifespan (A), daily fecundity (B), total fecundity (C), and fitness (D) of flies reared with bacterial mutants or control strains. Significant differences were determined by a log-rank test (A) or a Kruskal-Wallis (Acetobacter strains) or Wilcoxon (E. coli strains) test (C, D). Compact letter displays show significant differences between treatments. Strains are shown in red (A. pasteurianus 3p3 bearing plasmids pCM62 [empty vector], pCM62-SDR [S-oxidoreductase], pCM62-GDH [glucose dehydrogenase], and pCM62-GnDH [gluconate dehydrogenase]) or green (E. coli BW25113 [WT, wild type] and E. coli BW25113 ΔmetH786::kan [metH]).

FIG 4

Leslie matrix construction. Sample construction of a Leslie matrix from raw data is shown for one vial (‘11c-13’, or replicate vial 3 [of 3] for A. tropicalis DmCS_006 on experimental day 1 [of 3]).