fwb12683-sup-0001-AppendixS1

advertisement
Appendix S1: Herpetofaunal Compositional Differences in Four Natural Clusters in
JDSP
Four natural clusters of herpetofaunal community composition were apparent at
45% similarity (Fig. S1). Of these, one contained eight of the nine fishless wetlands; and
the remaining three consisted of the wetlands where fish were present – except for the
fishless Site 16, which was in cluster 3. Cluster 1 (Fig. S1) was made up of fishless
wetlands, and of its four contributors that made up 90% of community similarity, H.
femoralis and H. gratiosa were found nearly exclusively in these wetlands (Fig. S2). The
other two, R. grylio and R. sphenocephala, were present in nearly every wetland in JDSP,
but especially abundant here.
Cluster 2 consisted entirely of wetlands with fish, including large (>4 cm) fish.
Herpetofaunal community assemblage consisted primarily by the aforementioned two
Ranids, which were present in all wetlands in this cluster. Cluster 3 consisted of wetlands
that were fishless, wetlands that had small (<4cm) fish present, and wetlands that had
large fish present. It separated from clusters 1 and 2 by having fewer R. grylio captures
and R. sphenocephala captures intermediate between the two. Additionally, K.
steindachneri was found abundantly in this cluster and though its abundance was
relatively low when compared to total herpetofaunal captures, this cluster had sufficiently
low captures of other common herpetofaunal to drive up its influence on community
similarity.
Finally, cluster 4 was largely depauperate of herpetofaunal life, and coincidentally
had the highest catch rates of large fish. These wetlands had very few captures, and even
fewer species present: 100% of the community similarity was due to Rana grylio and the
snake Nerodia floridana. Despite both having elevated CPUEs of fish, and the presence
of extremely large predatory fish which are absent in other clusters, these sites are
different in nearly every other metric (Table 1, Table S1): they are spatially separated by
more than 4 km, occur in two distinct habitat types and elevations (scrub and pine
flatwoods), and one is an open water pond while the other is an emergent-vegetation
filled marsh.
- - 4- - - 2- - - - - - - - -3 - - - - - - -1 - - - - - - -
Site
Fig. S1. Cluster dendrogram of log transformed herpetofaunal
community assemblages. Sites that are clustered together below
the dotted line are at least 44% similar in community composition.
At this level of similarity there are four clusters.
Table S1: SIMPER analysis of species contributing to similarity
within each natural cluster at 44% similarity.
Community Composition By Cluster
Cluster
Avg.
Species
Avg. Abundance
Avg. Sim.
Contribution
Sim.
1
2
3
4
44.32
R. grylio
1.6
19.85
44.79%
44.32
H. femoralis
.68
8.62
19.45%
44.32
R. sphenocephala
.85
7.82
17.65%
44.32
H. gratiosa
.55
4.93
11.13%
63.79
R. grylio
.87
51.00
79.96%
63.79
R. sphenocephala
.16
6.66
10.43%
46.71
R. grylio
.18
18.48
39.56%
46.71
R. sphenocephala
.31
13.83
29.61%
46.71
K. standecheri
.18
12.56
26.88%
46.06
R. grylio
.07
31.54
68.47%
46.06
N. floridana
.04
14.52
31.53%
b.)
a.)
2
4
3
c.)
d.)
Fig. S2. NMDS ordination with individual species abundances in wetland herpetofaunal
communities in JDSP. a.) pinewoods treefrogs (Hyla femoralis) – with cluster numbers
labeled, b.) pig frog (Rana grylio), c.) Florida mud turtle (Kinosternon steindachneri),
and d.) southern leopard frog (Rana sphenocephala), bubble size represents relative
abundance of given species in a wetland, and wetlands with no captures of such species
are not represented.
Download