Basic HTML Version
International Journal of Marine Science 2013, Vol.3, No.41, 333-343
http://ijms.sophiapublisher.com
335
NH
4
and SiO
4
) were filtered and immediately stored
on ice before being frozen for later analysis.
Each water sample was then placed in 200 ml
containers and preserved with Acidified Lugols
solution. One hundred ml of the Lugols fixed sample
was then settled out in the laboratory for 48 hours
using a 100 mL graduated cylinder. After 48 hours, 90
ml from the top layer of water in the settling tube was
removed and the remaining 10 ml was placed into a
vial, which was capped and kept in the dark prior to
counting.
Examination of the phytoplankton samples used the
Lund cell method, which involved counting 3
traverses for each sample. An upright light microscope
with white light achromatic objective of 400×
magnification was used to identify and count the
phytoplankton cells. The cell abundances for each
species (expressed as number of cells per mL) and
thus the cell abundances for each of the major groups
and of the total sample were determined using the
formula as described by Hotzel and Croome (1999).
2.3 Data analysis
A series of one-way Analysis of Variances (ANOVAs)
was used to determine whether the total cell
abundance and abundance of the Cyanophyceae,
Bacillariophyceae, Chlorophyceae and the overall
total number of species differed significantly among
treatments, i.e. control, control mesocosms and impact
mesocosms (SPSS, Inc). Thus, ANOVA’s were run
separately for each time period, i.e. 30 min, 3 h, and 1,
2, 3, 4 and 5 (day) after disturbance, with treatment
being the main factor. If ANOVA detected a
significant difference, the
a posteriori
LSD test was
used to determine where any significance differences
revealed.
Non-metric multi-dimensional scaling (MDS) ordinations
were employed on log-transformed (n + 1) data to
determine whether the species composition of the
phytoplankton differed among times and treatments
(PRIMER5, Clarke and Gorley, 2001). Thus, the
log-transformed abundance data for each phytoplankton
species were calculated for the three treatments in
each time interval. These data were then subjected to
the Bray-Curtis measure and the resultant similarity
matrices ordinated. The stress levels on all plots
were < 0.2 and thus provide an accurate representation
of the relationships between the data (Clarke and
Warwick, 1994).
3 Results
3.1 Effects on water quality and nutrient concentrations
During the 5 day mesocosm experiment in Neranie
Bay, there was no rainfall, salinity was constant at 3 ppt
and water temperature was 24-26
℃
. Due to a
non-functioning oxygen probe during the early part of
the study, Dissolved Oxygen (DO) measurements
were only available from Day 2 onwards; they showed
mean DO levels of >8.8 mg/L in the control area, >9
mg/L in the control mesocosms and a marked increase
in DO from 6.8 mg/L on Day 2 to 10.1 mg/L on Day 5
in the impact mesocosms. The lack of DO data during
the first 2 days of this mesocosm experiment was
addressed by Sampaklis (2003), who tested the effects
of mixing known amounts of gyttja with known
volumes of lake water; when the volume of gyttja in
mixed samples was 20%, DO levels dropped to about
6 mg/L but recovered to 8 mg/L within 3 days. Other
water quality data in which effects were noted
(turbidity and pH) are plotted in Figure 3.
Initial turbidity readings in the study area (prior to any
gyttja disturbance) were low (< 2 NTU) (Figure 3a).
In the case of the control sites (no mesocosms), the
turbidity never exceeded 2 NTU throughout the
experiment. The installation of the control mesocosms
(CMs) into the sediments resulted in a short-term
minor increase in turbidity to 4-6 NTU, followed by a
decline to
∼
3 NTU within one day. Thirty minutes
after gyttja in the IMs was mixed with overlying water,
turbidity levels were 18-24 NTU; they declined to
∼
6.5 NTU after 1 day and stayed at or below this level
for the remainder of the experiment. A one-way
ANOVA showed a significant difference in turbidity
among treatments at all times following disturbance,
with turbidity in IMs consistently higher than both
control sites and the CMs up until Day 4 (Table 1).
Prior to disturbance, pH levels in the study area were ~
8.5 and remained between 8 and 9 throughout in the
control mesocosms (Figure 3b). The impact of mixing
gyttja into overlying water was immediate, with pH
levels dropping from 8.5 to 6.5, but steadily increasing
thereafter to a near normal pH levels of ~ 8 on Day 5.
The one-way ANOVA result showed a significant
difference in pH between treatments for all periods of