Geoscience Reference
In-Depth Information
Table 14.1
Number of macroinvertebrate samples from
the bank and channel habitats in each of the electrical
conductivity (EC) categories.
Thus we hypothesized that rising salinity levels
should result in a decrease in delta
+
for riverine
macroinvertebrates and that the response of this
index could differ from that of species richness,
which Kefford
et al
. (2011) have recently examined
for this region. In addition, we wanted to test
the idea that decreases in delta
Number of
samples
(bank)
Number of
samples
(channel)
EC category
(mS cm
−1
)
Category
are accompanied
by reductions in trophic diversity, as suggested by
Warwick and Clarke (1998) and Marchant (2007).
+
<
0.050
1
390
367
0.050-0.099
2
429
368
0.10-0.19
3
296
223
0.20-0.29
4
166
72
0.30-0.49
5
187
105
Methods
0.50-0.99
6
317
171
1.0-1.49
7
196
109
Two species-level datasets were compiled from lotic
macroinvertebrate samples taken in Victoria and
South Australia by the Environment Protection
Authority (EPA) Victoria and the South Australian
EPA, one dataset for each of the two habitats
at a site: the bank (a low velocity region near
the edge of the channel) and the main channel
(often a riffle). More samples were available for
the bank (2966 samples from 941 sites) than the
channel (1829 samples from 607 sites). Only data
on insect species were extracted; these constitute
approximately 93% of the species recorded at
Victorian sites (Marchant
et al
., 2006). These sites
spanned a range of potential disturbances, e.g. poor
water quality, variations in land use, reductions in
discharge. However, most samples (approximately
70%) were from reference sites, which exemplified
the least disturbed conditions available. Samples
were assigned to different predefined (Kefford
et al
.,
2006) salinity categories (Table 14.1). As many sites
were amalgamated into each salinity group other
disturbance gradients were probably obscured or at
least minimized.
Before delta
1.5-1.9
8
194
81
2.0-2.9
9
211
99
3.0-3.9
10
122
56
4.0-4.9
11
94
45
5.0-6.9
12
140
55
7.0-9.9
13
104
31
10-14.9
14
71
28
15-30
15
28
10
>
30
16
21
9
recognized as an environmental hazard (Kefford
et al
., 2006). Since the early 1990s monitoring
of macroinvertebrate communities in rivers in
Victoria and South Australia has resulted in
the accumulation of a large dataset of species
distribution at many sites. An obvious difficulty in
using such data to determine the effects of salinity
on biological diversity is that sampling effort varies
across the dataset. When sites are assigned to
salinity categories in order to reveal major trends
it becomes clear that different salinity ranges have
been sampled with very different numbers of sites
and thus quite different amounts of effort (Table
14.1). Generally, the higher salinity categories
have many fewer samples because such sites are
relatively uncommon among river habitats, even
in south-eastern Australia. Taxonomic distinctness
is well suited as a measure of diversity in this
situation because variations in sampling effort do
not affect its value. It thus may provide a reliable
measure of diversity when only a species list is
available for each site.
can be calculated a master list of
all species potentially present must be compiled. In
this case the master lists were simply the combined
list of species for each habitat. Five taxonomic levels
were present: species, genus, family, order and
class. Presence/absence species data for samples in
each electrical conductivity (EC) category (Table
14.1) were pooled and an average delta
+
value
was then calculated using the TAXDTEST routine
in PRIMER v.6 (Clarke and Gorley, 2006), with a
constant path length between different taxonomic
+
