Influence of management practice on the microbiota of a critically endangered species: a longitudinal study of kākāpō chick faeces and associated nest litter

Abstract

Background: The critically endangered kākāpō is a flightless, nocturnal parrot endemic to Aotearoa New Zealand. Recent efforts to describe the gastrointestinal microbial community of this threatened herbivore revealed a low-diversity microbiota that is often dominated by Escherichia-Shigella bacteria. Given the importance of associated microbial communities to animal health, and increasing appreciation of their potential relevance to threatened species conservation, we sought to better understand the development of this unusual gut microbiota profile. To this end, we conducted a longitudinal analysis of faecal material collected from kākāpō chicks during the 2019 breeding season, in addition to associated nest litter material.

Results: Using an experimental approach rarely seen in studies of threatened species microbiota, we evaluated the impact of a regular conservation practice on the developing kākāpō microbiota, namely the removal of faecal material from nests. Artificially removing chick faeces from nests had negligible impact on bacterial community diversity for either chicks or nests (p > 0.05). However, the gut microbiota did change significantly over time as chick age increased (p < 0.01), with an increasing relative abundance of Escherichia-Shigella coli over the study period and similar observations for the associated nest litter microbiota (p < 0.01). Supplementary feeding substantially altered gut bacterial diversity of kākāpō chicks (p < 0.01), characterised by a significant increase in Lactobacillus bacteria.

Conclusions: Overall, chick age and hand rearing conditions had the most marked impact on faecal bacterial communities. Similarly, the surrounding nest litter microbiota changed significantly over time since a kākāpō chick was first placed in the nest, though we found no evidence that removal of faecal material influenced the bacterial communities of either litter or faecal samples. Taken together, these observations will inform ongoing conservation and management of this most enigmatic of bird species.

Keywords: Avian; Bird; Conservation; Experimental; Microbiome; Microbiota; Threatened.

Fig. 1

Bray–Curtis dissimilarity distances based on 16S rRNA gene sequences for faecal samples visualised via principal coordinate analysis (PCoA) ordination. Each dot of the PCoA represents the microbiota of a single kākāpō chick faecal sample. Samples are shaped by location and coloured by A whether samples were collected while the chick was in the hand rearing facility versus in a nest, B relative abundance of Escherichia-Shigella coli, and C chick age at sample collection. Panel D depicts the most influential ASV vectors plotted using the vegan::envfit function

Fig. 2

Bray–Curtis dissimilarity distances of 16S rRNA gene sequences for litter samples visualised via principal coordinate analysis (PCoA) ordination. Each dot of the PCoA represents the microbiota of a single nest litter sample. Samples are shaped by island location and coloured by A island location, B relative abundance of Escherichia-Shigella coli, and C number of days since the nest sampled first housed a chick. Panel D depicts the most influential ASV vectors plotted using the vegan::envfit function

Fig. 3

A Observed richness and B Inverse Simpson alpha-diversity indices for kākāpō chick faecal samples grouped by location. Significant Dunn’s test pairwise comparisons with Benjamini–Hochberg adjustment between location groups in the box-plots are denoted by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001). Boxes represent the median (within-box horizontal line), 25^th (lower hinge) and 75^th (upper hinge) percentiles. Whiskers extend to the smallest and largest values within 1.5 times interquartile range above the 25^th and 75^th percentiles, respectively. Data beyond the end of the whiskers are outliers and plotted individually. C Species-level taxonomic distribution of bacteria by the relative abundance of 16S rRNA gene sequences within location groups (at time of sampling) for individual kākāpō chick faecal samples (samples for a given chick may include those collected both in captivity and in the nest). Individual samples are ordered within location groups chronologically and alphabetically. Taxa with mean relative 16S rRNA gene sequence abundance < 0.3% are grouped as ‘Other species’

Fig. 4

A 16S rRNA gene sequence-based taxonomic distribution of bacteria within age groups and individual kākāpō chick faecal samples at the species level. Bars on the aggregated graph (left) are in the same order as the sample-level profile age groups from left to right. Individual samples are ordered within age groups chronologically and alphabetically. Taxa with mean relative 16S rRNA gene sequence abundance < 0.3% are grouped as ‘Other species’. Underlined samples are those collected from chicks being hand reared at the time of collection. B Observed richness and C Inverse Simpson (plotted on a log10 scale) alpha-diversity indices for kākāpō chick faecal samples grouped by chick age. The y-axes for corresponding bar- and box-plots are identical. Significant Dunn’s test pairwise comparisons with Benjamini–Hochberg adjustment between age groups in the box plot are denoted by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001). Both box- and bar- plots show a decrease in bacterial community diversity with increasing chick age. Box-plot details are as described for Fig. 3

Fig. 5

A 16S rRNA gene sequence-based taxonomic distribution of bacteria by days since the nest sampled first housed a chick and individual nest litter faecal samples at the species level. Bars on the aggregated graph (left) are in the same order as the sample-level profile groups from left to right. Individual samples are ordered within groups chronologically and alphabetically. Taxa with mean relative 16S rRNA gene sequence abundance < 0.5% are grouped as ‘Other species’. B Observed richness and C Inverse Simpson (plotted on a log10 scale) alpha-diversity indices for litter samples grouped by days since the first chick was introduced to the nest. The y-axes for corresponding bar- and box-plots are identical. Significant Dunn’s test pairwise comparisons with Benjamini–Hochberg adjustment between groups in the box-plots are denoted by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001). Box-plot details are as described for Fig. 3