Figure 1.
Coverage comparison of 16S and CO1 barcoding markers on prey groups.
Number (right, log scale) and proportion (left) of sequences belonging to various types of identified prey taxa from 16 puffin adult, 22 puffin chick, and 28 herring samples. All samples produced at least one sequence for each gene. Only 20 (8 adult, 9 chick, 3 herring) of the 66 samples produced over 50 sequences per marker (Table 3).
Figure 2.
Summary of taxa identified from all sequenced puffin adult and chick fecal samples and from all sequenced herring stomach contents.
Overlapping regions represent taxa occurring in multiple sample types (eg: centre region contains taxa found in all three predator diet samples). Non-animal and non-informative taxa are listed in grey text in square brackets.
Figure 3.
A graphical representation of the Machias Seal Island Atlantic puffin (Fratercula arctica) and Atlantic herring (Clupea harengus) food chain.
Increasing line thickness relates to higher frequency of occurrence in adult (blue), chick (green), and herring (purple) diet. Data are all derived from samples within common sampling period (13 June - 29 July).
Figure 4.
Frequency of occurrence (FOO) of top 16S taxa (>10%) in diet of puffin adults and chicks.
Blue bars represent adult (n=17) data, green bars represent chick (n=41) data.
Figure 5.
Non-metric multidimensional scaling (MDS) plots depicting differences in prey assemblage of adult and chick puffin diet.
Analysis was limited to 16S samples with >50 sequences per fecal sample and taxa which could be considered targeted prey for puffins. Each symbol represents a fecal sample, and the distance between the different symbols represents the difference in taxa composition between fecal samples. There is much overlap between adult and chick fecal samples, reflecting diet overlap. Vectors are the correlations >0.35 between a prey taxon and the MDS axes, where vector length and direction reflect taxon frequency; the big circle indicates the maximum vector length.
Figure 6.
Frequency of occurrence of chick prey through field observations and DNA sequencing.
Coloured bars represent DNA-identified taxa which are assigned to one of 7 field observation prey categories (black bars).
Figure 7.
Non-metric multidimensional scaling (MDS) plots depicting differences in prey assemblage of chick diet through DNA and field observations.
Grey circles represent a single feeding delivery to a puffin chick. Stars represent individual chick fecal samples. Vectors are the correlations between prey types (herring, sandlance, polychaete, sculpin, hake, butterfish, and krill) and the MDS axes where vector length and direction reflects taxon frequency. See caption on Figure 5 for more explanations on MDS plots and vector overlays.
Figure 8.
Species accumulation curve for DNA-based chick diet.
Only samples with >50 sequences from kingdom Animalia were included. Crosses with standard deviation error bars represent Chao2 S extrapolator for prediction of the true total number of taxa present in samples.