Page 5 - DeidunSchembri2006

Mo 1ile fauna of infra lilfora/ fringe of Ma/lese beaches 359

The majority of species collected were omnivores/scavengers (9 species - 25.7%
of ali species collected), followed by micro-carnivores (8 species - 22.9%) and
deposit feeders/detritivores (6 species - 17.1%) (Fig. 2). There were evident sea-
sonal changes in feeding types: macro-phytophagous species were almost exclu-
sively collected duri11g winter and spring for both years a11d micro-phytophagous
species were only collected during summer and autumn for both years. In terms
of abunda11ce, micro-carnivores and om11ivores/scave11gers were the most repre-
se11ted duri11g practically ali seaso11s.

Fig. 3 shows a hierarchical clustering plot for ali the beach collections made
over the entire two year period (i.e. ali seasons together). Samples separated

o

60

100 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
~xxxxxx~~~~~~~~~~~~~x~x~~~~~~~~~
---N-NN--NN-N-N-~_NNN--NN-- - NNNN
Samples

Fig. 3 - Dendrogram resuJting from hierarchical clustering of abundance data based on the
Bray-Curtis similarity measure for ali Maltese beach collections (both years, all sea-
sons). W=White Tower Bay; G=Golden Bay; R=Ramla; X=Xatt 1-Ahmar; Au=Au-
tumn; Wi= Winter' Sp = Spring; Su= Summer.

Dendrogramma risultante dal clustering gerarchico dei dati di abbondanza basati sulla misura della
similarità di Bray -Curtis per tutti i campionamenti condotti sulle spiagge ma/tesi( due anni, tutte
le stagioni). W= White Tower Bay; G=Golden Bay; R=Ramla; X=Xatt 1-Ahmar; Au=autunno;
Wi= inverno; Sp =primavera; Su= estate.

largely 011 spatial distributio11 rather than 011 seaso11al differe11ces, with collectio11s
from Xatt 1-Ahmar, and to a lesser extent from Ramla, exhibiting least similar-
ity with the other beaches. The equivale11t NMDS plot (Fig. 4) gave two broad
groups: Xatt 1-Ahmar winter and autumn samples, and all the others. Of the
abiotic parameters measured, none was useful in explaining the observed pat-
tern, with the possible exception of grain size for comparisons betwee11 Maltese
beaches only (Fig. 4).

For the Maltese beaches, BIO-ENV gave very poor correlatio11s (p = 0.063 and

0.418 for first and second year samples, respectively) between exposure, median
grain size, sediment orga11ic co11tent and POM content, and the observed biotic
patterns. Similarly, there were no correlations between the same abiotic variables,
excepting POM content, and biotic patterns for spring samples only (p = -0.044).