Impact of exploratory offshore drilling on benthic communities in the Minerva gas field, Port Campbell, Australia

lickforkabsorbingOil and Offshore

Nov 8, 2013 (3 years and 9 months ago)

122 views

Impact of exploratory offshore drilling on
benthic communities in the Minerva gas
field,Port Campbell,Australia
D.R.Currie
*
,Leanne R.Isaacs
Centre for Environmental Management,Central Queensland University,P.O.Box 1319,Gladstone,
Qld.4680,Australia
Received 29 October 2002;received in revised form 26 April 2004;accepted 4 May 2004
Abstract
Changes to benthic infauna caused by exploratory gas drilling operations in the Minerva
field were examined experimentally using a BACI (before,after,control,impact) design.
Analysis of 720.1 m
2
Smith–McIntyre grab samples obtained from one pre-drilling and
three post-drilling periods yielded a diverse fauna consisting of 196 invertebrate species
and 5035 individuals.Changes to benthic community structure were assessed using ANOVA
and nonmetric multidimensional scaling (MDS).The abundances of two common species
(Apseudes sp.1 and Prionospio coorilla) decreased significantly at the well-head site immedi-
ately after drilling.The size of these reductions in abundance ranged between 71% and 88%,
and persisted for less than 4 months after drilling.A third common species (Katlysia sp.1)
increased in abundance 200 m east of the well-head following drilling.Most species occurred
at densities too low to be analysed individually and so were pooled at higher taxonomic levels.
Changes in the abundance of species aggregated by phylum varied,but significant declines in
the most abundant phyla (Crustaceans and Polychaetes) of 45–73% were observed at all sites
within a 100 m radius of the well-head following drilling.In most cases these changes became
undetectable four months after drilling following species recruitments.MDS ordinations
confirm that drilling related changes to benthic community structure are most pronounced at
stations located closest to the well-head.Additionally,the ordinations indicate that modified
communities persist at the well-head for more than 11 months following exploratory drilling.
 2004 Elsevier Ltd.All rights reserved.
Keywords:Benthos;Infauna;Environmental impact;Drilling muds;Oil and gas platforms;Australia
*
Corresponding author.
E-mail address:d.currie@cqu.edu.au (D.R.Currie).
0141-1136/$ - see front matter  2004 Elsevier Ltd.All rights reserved.
doi:10.1016/j.marenvres.2004.05.001
www.elsevier.com/locate/marenvrev
Marine Environmental Research 59 (2005) 217–233
MARINE
ENVIRONMENTAL
RESEARCH
1.Introduction
The Minerva gas field,which is situated approximately 12 km off the southern
Victorian coast,is regarded as one of Australia’s most significant gas reserves.Es-
timated to contain about 300 billion cubic feet of gas,the field has the capacity to
supply Victoria’s total gas market for approximately two years.Since its discovery,
key stakeholder groups have been investigating the feasibility of a Minerva devel-
opment with the aim of producing first gas by 2004.As part of an environmental
impact assessment,a monitoring program was initiated to determine the extent and
persistence of the effects of test drilling operations on benthic infauna within the
Minerva field.
Benthic communities are widely used in the monitoring of effects of marine im-
pacts as the organisms are mostly sessile and integrate effects of pollutants over time
(Gray,Clarke,Warwick,& Hobbs,1990).Most infaunal communities comprise a
large number of species and because of varying sensitivities of species it should be
possible to identify subtle effects of contaminants reflected in changes in community
structure.Oil companies are usually required to monitor the effects of their activities
on marine life,but there are few published accounts detailing the effects of drilling on
benthos in Australian waters.
This paper examines drilling-related changes to the abundance and diversity of
infaunal animals at Minerva using a BACI (before,after,control,impact) experi-
mental design (Stewart-Oaten,Murdoch,& Parker,1986).This design involves si-
multaneous sampling of one ‘control’ site,and five ‘impact’ sites,on a number of
occasions both before and after experimental drilling.
Changes to infaunal communities following drilling were assessed using Bray–
Curtis dissimilarity measures and multidimensional scaling (MDS).Changes to the
abundance of the more common species,and to groupings of species based on higher
order taxonomic classifications,were further examined using analysis of variance
(ANOVA),and by documenting the percentage change experienced by different
species and species groupings.
2.Materials and methods
2.1.Study design
The Minerva-2A well site is situated in 60 m of water approximately 12 km off-
shore from Port Campbell on the southern Victorian coast (38 43
0
04.3
00
S,142
57
0
19.9
00
E).Sampling sites were arranged as radii and spaced logarithmically on two
axes;a principal axis oriented east–west (in parallel with predicted current move-
ments),and a shorter secondary axis extending north from the well-head at site cp
(Fig.1).While a total of 23 sites were sampled during the course of this study,most
sites (17) were only sampled using qualitative techniques (video and dredge sam-
pling).High costs associated with sorting and identifying infaunal organisms limited
quantitative benthic sampling to six sites (i.e.,cp,well-head;k,t and b,100 m west,
218 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
north and east of well-head,respectively,d,200 m east of well head;and e,400 m
east of well-head).
The distribution and abundance of infauna at each site was determined from
replicate 0.1 m
2
Smith–McIntyre grab samples.A total of three samples were taken
from each of the six sites on one sampling period before and three after the exper-
imental drilling (i.e.,a total of 72 grabs).Drilling was conducted over a three week
period (September–October) using water-based muds,and samples were taken 1
week before,and 2 weeks,4 months and 11 months after drilling.The position of
each sample was fixed using a differential global positioning system (DGPS) con-
sisting of a Trimble 4000 reference and differential locator,and a UHF radio data
link.This system provided an accuracy better than 10 m in 93% of fixes.Samples
were drained,weighed and a one litre sediment sub-sample retained for physical/
chemical analysis.All animals retained on a 1.0 mm sieve were preserved in a 10%
formaldehyde solution and later sorted to the lowest practical taxonomic level
(generally species) under a dissecting microscope,before being counted.
2.2.Analysis of changes to community structure
Differences between the six sites in each sampling period were examined using
Bray–Curtis (B–C) dissimilarity measures (Bray & Curtis,1957).This measure was
N

Sampling technique
Video only
v

Quantitative benthic grab
u

Video + Sediment core
t

s
j

r

q p

o n

m k

l cp a b c d e f g h
i
Distance from well head (site cp) in metres Sites
50 a,j,s
100 b,k,t
150 c,l
200 d,m,u
400 e,n,v
800
f,o
1600 g,p
3200 h,q
6400 i,r
Fig.1.Schematic diagram showing the distribution of sampling sites around the Minerva-2A well head
(cp) situated 12 kmoff Port Campbell on the southern Victorian coast (38 43
0
04.3
00
S,142 57
0
19.9
00
E) in 60
m of water.The principal axis of this cross-hair design is orientated east-west in parallel with prevailing
currents.Smith–McIntyre grab samples were taken at all sites filled either black or white,while the re-
maining sites were sampled by video.Only those benthic samples taken at six sites (b,cp,d,e,k,t) and
four sampling periods ()1 week;+2 weeks;+4 months;+11 months) are examined in this paper.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 219
chosen because it is not affected by joint absences,it gives more weighting to
abundant than rare species,and it has consistently performed well in preserving
‘ecological distance’ in a variety of simulations on different types of data (Faith,
Minchin,& Belbin,1987).
On each sampling period the number of individuals of each species was calculated
from the total number of individuals found on each site i.e.,data from the three
replicate grabs on each site were pooled.Before calculating the B–C dissimilarity
measures a double square root transformation was applied to the number of indi-
viduals of each species.This transformation was used to prevent abundant species
from influencing the B–C dissimilarity excessively.
Bray–Curtis dissimilarity measures calculated for all 24 site ·period (6 sites 4
periods) combinations,resulted in a triangular matrix of dissimilarities which was
used to map the site ·period inter-relationships in two dimensions using multidi-
mensional scaling (MDS).MDS plots a measure of similarity between objects into
two or more dimensional space so that distances between objects correspond closely
to their input similarities.While the computational algorithm for MDS is complex
the graphical representation is easily communicated (Clarke,1993) and ecologically
meaningful patterns become more apparent (Gamito & Raffaelli,1992).PATN
(Belbin,1990) was employed for ordinations in this study.Monotonic regression was
used after a ‘cut-value’ less than the lowest dissimilarity measure was specified.The
final configurations presented were the best solutions (i.e.,exhibited the lowest
‘stress’ values,or least distortion) from 100 random starts.
To test the robustness of the species level MDS,further MDS ordinations were
prepared using three separate aggregations of the species data based on their hier-
archical taxonomic classification (i.e.,‘Family’,‘Class’ and ‘Phylum’).In all cases a
double square root transformation was applied to the raw abundances before dis-
similarity measures were calculated.
All foraminifera (three species) were excluded from the MDS analysis,and sub-
sequent ANOVA’s as live and dead individuals were difficult to distinguish.
2.3.Analysis of changes in species richness and abundance
Changes in total abundance and richness of benthos at each site were examined
using nested ANOVA.The statistical model applied was
X
ijk
¼ l þa
i
þs
kðiÞ
þb
j
þðabÞ
ij
þe
ijk
;
where l is the overall mean,a
i
is the effect of period (i ¼before or after),s
kðiÞ
is times
within period (k ¼ 1;2;...;t
B
for i ¼before),b
j
is the effect of location (j ¼control or
impact),ðabÞ
ij
is the interaction between period and location,and e
ijk
is the residual
error.
In these nested ANOVAs site ·time interactions were tested against the mean
square for the site ·period(time) term as outlined in Underwood (1991,Table 1C).
This test is equivalent to the t-test recommended by Stewart-Oaten et al.(1986).
Homogeneity of variance was examined using Cochran’s test and heterogeneity re-
moved by log
10
(N +1) transformation.To provide an objective test of the persistence
220 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
of drilling impacts,ANOVAs were performed with 1,2,and 3,post-drilling sampling
dates.In cases were just one post-drilling date is included,F ratios were calculated
using the mean square of the site ·time interaction term.
Site e,located 400 m east of the well-head was used as the ‘control’ site for all
analyses.This selection was determined prior to the commencement of drilling,and
reflected a high degree of community similarity between sites.Moreover,site e was
geographically most removed from the well-head site and was considered,for hy-
drographical reasons,outside the influence of drilling impacts.
Changes in the abundances of individual species were also examined using the
same ANOVA model detailed above.Only three species occurred in sufficient
numbers to allow comparative tests of changes in abundance.To examine changes in
those species too rare to analyse individually,species were grouped into five phyla;
Crustacea,Echinodermata,Mollusca,Polychaeta and Others.
Estimates of the percentage change in abundances of individual species and the
five phyla groupings were calculated using the pre-drilling sampling and the three
post-drilling samplings.Percentage change was estimated from the difference be-
tween ratios of the sum of individuals on sites before and after drilling (Currie &
Parry,1996).i.e.,
% Difference ¼ ½ðN
da
N
ca
=N
ca
Þ ðN
db
N
cb
=N
cb
Þ 100;
where N is the mean number of individuals,and the subscripts are as follows:c is the
control site (e),d is the drill site (b,cp,d,t,or k),a is the after drilling,and b is the
before drilling.
3.Results
3.1.Physical changes to seabed
The seafloor within the 13 kmstudy area was broadly uniformin water depth (65–
71 m),and largely composed of coarse,well-sorted sediments (Table 1).Video in-
spections showed that bedforms were dominated by low-relief (<20 cm),parallel
sand-ripples,spaced at approximately 2 m intervals.This bottom relief is believed to
be due to the action of ground-swell waves,that may be generated over thousands of
kilometres of open ocean to the southwest of the study area.The highly dynamic and
unconsolidated nature of sediments at this location undoubtedly limits the attach-
ment and growth of sedentary organisms,and probably accounts for the fact that no
epifaunal organisms were observed during video inspections.
Because of inclement weather,it was not possible to conduct video inspections
during the period prior to drilling,and all video observations were made two weeks
after the cessation of drilling operations.While this sampling regime precluded an
objective visual assessment of change resultant from drilling operations,it did allow
an insight into the geographical extent of the drilling impact.In consort with grab
sampling,video assessments confirmed that the physical influence of the exploratory
drilling was initially restricted to approximately 100 m distance from the well-head
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 221
(Table 1).Drill-cuttings remained present in grab samples taken from the well head
site 4 months after the completion of drilling.However,no drill-cuttings were ob-
served at any of the grab sampling station 11 months after drilling,most probably
because of sediment reworking during natural hydrodynamic processes.
3.2.General changes to benthic community structure
In common with many ecological communities,the benthic infauna at the Mi-
nerva sites was composed of a small number of abundant species and a large number
of rare species.A total of 196 invertebrate species,and 5035 individuals were col-
lected at the six Minerva sites during this study.Most species collected were crus-
taceans 124 (63%),followed by polychaetes 30 (15%),molluscs 15 (8%),echinoderms
9 (5%),and members of other phyla 18 (9%).The bivalve mollusc Katlysia sp.1 was
the most abundant species and contributed 18%of the animals collected.Collectively
the ten most abundant species contributed 61% of the animals collected.By contrast
128 species were represented in fewer than 10% (7 of 72) of grab samples,and 47
species occurred in only one grab.
Table 1
Summary table detailing differences in depth,sediment structure and the presence of drill cutting at 23
sampling stations located around the Minerva-2A well head (cp)
Site Distance from
well head (m)
Depth (m) Sampling
method
Sediment
structure
Presence of
drill cuttings
cp 0 67 Grab Gravel Yes
a 50 66 Video n/a Yes
j 50 67 Video n/a Yes
s 50 68 Video n/a Yes
b 100 68 Grab V.coarse sand Yes
k 100 68 Grab Coarse sand Yes
t 100 67 Grab Coarse sand No
c 150 67 Video n/a No
l 150 68 Video n/a No
d 200 70 Grab Coarse sand No
m 200 70 Video n/a No
u 200 68 Video+core Coarse sand No
e 400 68 Grab Coarse sand No
n 400 67 Video+core Coarse sand No
v 400 67 Video+core Coarse sand No
f 800 67 Video+core Coarse sand No
o 800 68 Video+core Coarse sand No
g 1600 68 Video+core V.coarse sand No
p 1600 71 Video+core Coarse sand No
h 3200 67 Video+core Coarse sand No
q 3200 71 Video+core Coarse sand No
i 6400 65 Video+core Medium sand No
r 6400 71 Video+core Coarse sand No
All sampling was conducted two weeks after the cessation of exploratory drilling.
222 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
Before the experimental drilling the numbers of species on the control (e) and
well-head sites (cp) were very similar (Fig.2,49 vs.47),but following drilling there
was a significant 36% decrease in the number of species at the well-head (ANOVA
including one post-drilling sampling period,p < 0:05;Table 2).This difference
persisted for less than four months,and at the end of the study numbers of species on
the well-head site slightly exceeded that of the control site (54 vs.56).The collective
species richness at all other sites did not change significantly following drilling.
Temporal and spatial patterns in total species abundance closely mirrored those
described for species richness.Two weeks after drilling,total abundances at the well-
head (cp) were significantly reduced by approximately 66% (ANOVA including one
post-drilling sampling,p < 0:05;Table 3).This decline in total abundance persisted
less than four months after drilling.In contrast,changes in total abundance at all
other sites were non significant (see Fig.3).
3.3.Multidimensional scaling (MDS)
The four MDS ordinations in Fig.4 map the spatial and temporal variation in
benthic community structure of the six Minerva sites during the study period.The
stress coefficients (<0.23),indicate that the ordinations are not unduly distorted
(Clarke,1993),and are a fair representation of the input dissimilarities in two di-
mensions.
The species level MDS ordination (Fig.4) shows a pronounced ecological gra-
dient across the six Minerva sites before drilling with sites/samples cp1,e1,and t1
forming a close grouping,sites/samples k1 and d1 plotting equidistant from this
No. species/0.3m
2

0
10
20
30
40
50
60
70
b1 b2 b3 b4 cp1 cp2 cp3 cp4 d1 d2 d3 d4 e1 e2 e3 e4 k1 k2 k3 k4 t1 t2 t3 t4

Sampling sites / periods
Fig.2.Change in total number of species found in 3·0.1 m
2
replicate grab samples taken at each of the
six sampling stations (b,cp,d,e,k,t) during four sampling periods (1,)1 week;2,+2 weeks;3,+4 months;
4,+11 months).Note all foraminifera excluded from counts.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 223
No. individuals/0.3m
2
PHYLUM
Annelida
Crustacea
Echinodermata
Mollusca
Others
0
50
100
150
200
250
300
350
400
450
b1 b2 b3 b4 cp1 cp2 cp3 cp4 d1 d2 d3 d4 e1 e2 e3 e4 k1 k2 k3 k4 t1 t2 t3 t4

Sampling sites / periods
Fig.3.Change in total abundance of five phyla found in 3 ·0.1 m
2
replicate grab samples taken at each of
the six sampling stations (b,cp,d,e,k,t) during four sampling periods (1,)1 week;2,+2 weeks;3,+4
months;4,+11 months).Note all foraminifera excluded from counts.
Table 2
Probability levels for ANOVAs for changes in species numbers including one sampling period before
drilling and three sampling periods after drilling (+2 weeks,+4 months and +11 months)
Site
Time after drilling
+2 weeks +4 months +11 months
b 0.1032 0.0509 0.3452
cp 0.0210 0.0900 0.2294
d 0.0521 0.9707 0.8548
k 0.7065 0.3655 0.4940
t 0.6269 0.8310 0.7963
Site e is used as a control in all analyses.For the +4 and +11 months analyses,site ·time interactions
are tested against site ·period (time) terms.All species frequencies were transformed using log
10
(N +1).
Table 3
Probability levels for ANOVAs for changes in total animal abundances including one sampling period
before drilling and three sampling periods after drilling (+2 weeks,+4 months and +11 months)
Site
Time after drilling
+2 weeks +4 months +11 months
b 0.6353 0.5152 0.7890
cp 0.0126 0.3481 0.3994
d 0.4899 0.5295 0.6553
k 0.0685 0.4364 0.3041
t 0.1697 0.3489 0.1715
Site e is used as a control in all analyses.For the +4 and +11 months analyses,site ·time interactions
are tested against site ·period (time) terms.All abundance data were transformed using log
10
(N +1).
224 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
grouping,and site/sample b1 plotting furthest away towards the top of the ordina-
tion.This gradient is broadly retained throughout the one year monitoring program
with samples from sites b and d,always plotting at one edge of the ordination,and
all but two samples from sites cp,e,k,and t plotting at the other.Most of this
observed inter-site dissimilarity is due to elevated numbers of the bivalve mollusc
Katlysia sp.1 at sites b and d,and an absence of the tanaid crustacean Apseudes sp.2
from most samples taken at sites cp,e,k and t.The gradient parallels observed
physical variations between sites,with sediments at sites b and d comprising course
shell/sand aggregate,in contrast with sediments at cp,e,k and t which were com-
posed of fine-sorted sand.
Large temporal changes in community structure at the well-head site (cp) fol-
lowing drilling are evident in the species level MDS (unbroken line,Fig.4).Before
drilling samples taken at the cp site (cp1) plot close to e1 and t1,however,samples
taken two weeks after drilling (cp2) plot some distance away on the lower edge of the
ordination.Samples taken at site cp four months after drilling (cp3) plot equidistant
from cp1,and close to cp2 on the ordination.In comparison samples taken 11
months after drilling (cp4) plot an increased distance from all earlier samplings at
this site.These temporal shifts summarise many of the changes in species number
Species Stress=0.17
Family Stress=0.21
Class Stress=0.23
Phylum Stress=0.19
Fig.4.MDS ordinations of species abundance data from the Minerva gas field pooled at four separate
taxonomic levels (species,family,class and phylum).Character variables (i.e.,b,cp,d,e,k,t) denote site
and numerals (i.e.,1,2,3,4) sampling period before and after drilling.Sites b and d are shaded light and e,
k and t dark,to highlight spatial differences between sites at Minerva.A solid line plots temporal
movements in community structure of the well-head site (cp) following experimental drilling.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 225
and abundance observed at cp during the one year monitoring program:the large
movement between cp1 and cp2 reflecting significant declines in species diversity and
abundance two weeks after drilling;the proximity of cp2 and cp3 illustrating the
persistence of these changes four months after drilling;and the increased line length
between cp3 and cp4 depicting species recruitment events 11 months after drilling.
The large ordination distance between the pre-drilling samples (cp1) and samples
taken on the same site one year later (cp4),are due largely to the occurrence of 33
species in cp4 not previously collected from the well-head site.Almost half of these
species (15 of 33) were also absent from the 60 grab samples taken at the other five
Minerva sites during the course of the study.
MDS ordinations of the species abundance data pooled at family class and
phylum levels (Fig.4) show similar patterns.In all instances the MDS configuration
conveys a site dependent environmental gradient highlighting the sediment hetero-
geneity at Minerva.Samples taken through time at sites b and d (Fig.4,shaded light)
invariably plot at one edge of the ordinations,while those from sites e,k and t (Fig.
4,shaded dark) plot at the other.This gradation is,however,less obvious in the class
level MDS ordination.Temporal trajectories for samples taken from the well-head
site (cp) after drilling are also consistent across the four taxonomic groupings (Fig.
4),with cp2,cp3 and cp4 always plotting at the periphery of the ordination.At all
taxonomic levels,large shifts are evident between cp1 and cp2 resulting from a re-
duction in numbers of taxa and individuals coinciding with experimental drilling.
Lesser movements are generally apparent between cp2 and cp3,while greatest
movements occur between cp3 and cp4 in association with recruitments in taxa
number and abundance (Table 4).In all instances the MDS ordinations place c4
distant fromc1,indicating that the benthic community structure at the well-head site
remains modified 11 months after drilling.
3.4.Changes in individual species and species groupings
Abundances of all three common species found at Minerva (Apseudes sp.1,
Katlysia sp.1 and Prionospio coorilla) changed significantly (p < 0:05) following
drilling,with two species decreasing and one increasing.Densities of Apseudes sp.1
and Prionospio coorilla both declined at the well-head site in the sampling period
immediately following drilling (71% and 88%,respectively),while densities of Ka-
tlysia sp.1 increased by 640% at site d (200 m east of the well-head) over the same
period (Tables 5(A) and 6(A)).These changes varied in persistence between species,
apparently due to differential periodicity in natural recruitment and mortality (Fig.
5).In the case of Apseudes sp.1 and Prionospio coorilla,initial declines in abundance
at the well-head site were indistinguishable four months after drilling.In contrast,
elevated numbers of Katlysia sp.1 persisted at site d for between 4 and 11 months
following drilling.
Spatial and temporal patterns of change in abundance coinciding with drilling
were replicated in groupings of species at higher taxonomic orders (Tables 5(B) and
6(B)).This result largely follows the high degree of numerical dominance of the three
most abundant species in their respective phylogenetic groups (i.e.,Apseudes sp.
226 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
Table4
Totalabundancesof13benthicinvertebratephylaidentifiedfromeachofsixsites(b,cp,d,e,k,t)duringfoursamplingperiods(1,)1week;2,+2weeks;3,
+4months;4,+11months)beforeandafterexperimentaldrillingatsitecp
Phylumb1b2b3b4cp1cp2cp3cp4d1d2d3d4e1e2e3e4k1k2k3k4t1t2t3t4
Crustacea151188905011744489177741375811113710914512857312221814812572
Mollusca214339120400721113246441916271810164261093182
Polychaeta92126192410265321203234274327472419354927173615
Echinodermata006081383141767283438580
Coelenterata0001011900020000000000000
Chordata011000040011001100180011
Nemertea021011000011121001000000
Aschelminthes000100070001000010000000
Bryozoa000000000000073000000000
Sipunculoidea010100000000000002000001
Unidentified000000010010000100100000
Platyhelminthes000000000001000000000001
Porifera000000000000000000100000
Totalnumber
ofindividuals
1812561631921545796171122208424141165211175214170126157188263179201172
Totalnumber
ofphyla
366655474488567656754456
Totalphyla
sharedwith
control(e)
34635436446545554454
Totalphylanot
sharedwith
control(e)
24160351123422411224
Numbersarepooledvaluesfor3·0.1m
2
replicateSmith–McIntyregrabsamples.Totalnumbersofphylasharedandunmatchedbetweeneachsiteandthe
controlsite(e)alsogiven.Noteallforaminiferaexcludedfromcounts.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 227
Table5
ProbabilitylevelsofANOVAsfortemporalchangesintheabundanceof(A)thethreemostcommoninfaunalspecies,and(B)fivephylumgrouping,fol-
lowingdrilling
SitebSitecpSitedSitekSitet +2
weeks
+4
months
+11
months
+2
weeks
+4
months
+11
months
+2
weeks
+4
months
+11
months
+2
weeks
+4
months
+11
months
+2
weeks
+4
months
+11
months
(A)SpeciesApseudessp.10.40410.77390.87880.00380.60460.43020.72810.86830.94780.59570.92530.75980.33050.14030.2065
Katlysiasp.10.24760.72830.51450.83750.76860.91230.00480.03850.08980.20630.15990.46460.49330.33080.3522
Prionospio
coorilla
0.71200.88430.91210.00400.73750.51980.75710.57000.53640.35500.79060.85560.07060.65300.4509
(B)PhylumCrustacea0.75880.89410.73680.01680.13710.15060.44670.90720.86230.07480.08010.08770.28910.06590.2956
Polychaeta0.96030.68890.62570.04870.66670.74100.30240.98770.93640.06080.91630.99980.06930.75160.5112
Mollusca0.30380.72860.51530.48460.40060.85100.00880.04430.11660.20220.12670.49090.59850.35700.4171
Echinodermata0.56360.49090.24690.31410.10110.70420.75920.77950.82520.59540.31190.49590.98780.69510.9987
Others0.60130.17620.89860.18790.96710.76300.08410.75150.91250.30210.28600.89110.08410.43480.6475
Siteewasusedasacontrolinallanalyses.Forthe+4and+11monthsanalyses,site·timeinteractionsaretestedagainstsite·period(time)terms.All
abundancedataweretransformedusinglog
10
(N+1).
228 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
1¼23% of crustaceans,Katlysia sp.1¼94% of molluscs,and Prionospio coo-
rilla ¼47% of polychaetes).Crustaceans and polychaete abundances both declined
significantly (p < 0:05) at the well head site in the period immediately following
drilling (73% and 66%,respectively),but differences for these groupings were not
Site b
1
0
0
A S O N D J F M A M J J A S O N D
Apseudes sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Katlysia sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Prionospio coorilla



Site cp
1
0
0
A S O N D J F M A M J J A S O N D
Apseudes sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Katlysia sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Prionospio coorilla

Number of Individuals / 0.1m2


Site d



↓ ↓











1
0
0
A S O N D J F M A M J J A S O N D
Apseudes sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Katlysia sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Prionospio coorilla



Site k
1
0
0
A S O N D J F M A M J J A S O N D
Apseudes sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Katlysia sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Prionospio coorilla



Site t
1
0
0
A S O N D J F M A M J J A S O N D
Apseudes sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Katlysia sp.1

1
0
0
A S O N D J F M A M J J A S O N D
Prionospio coorilla

Sampling Month
Fig.5.Change in the abundance of the three most common species on the control site e (––) and impact
sites b,cp,d,k and t (_ _ _ _ _) before and after drilling.Arrows indicate time of experimental drilling.
Error bars show standard errors.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 229
Table6
Percentagechangeinabundanceof(A)thethreemostcommonspecies,(B)fivephylumgroupings,ateachsite2weeksafterdrilling
Mean
N/0.1m
2
%Differencefromcontrolsitepre-drilling%ChangefollowingexperimentaldrillingPoweroftestD50%,a¼0:10
bcpdktbcpdktbcpdkt
(A)Species
Apseudessp.116.0+42)25)21)52+210)140)71
)30)11)2040.860.700.630.97
Katlysiasp.16.6+5)950)47)58+181)5+640

+41+110.760.640.710.74
Prionospio
coorilla
3.3)70+70)60+40+70)1)88
+13)58)1170.470.450.520.68
(B)Phylum
Crustacea37.0+36+5)31+14+96+74)73
)15)52)880.970.650.790.99
Polychaeta9.0)67)11)22)110+16)66
)31)45)600.810.780.830.86
Mollusca6.3+11)79+11)47)47+158)21+596

+47+40.770.510.630.72
Echinodermata2.3+14)57+14+14)97)26)64)310.490.130.470.47
Others0.300)78)670.100.10
Siteeisusedasacontrolforallcomparisons.Meandensities(N/0.1m
2
)areforsiteeduringthepre-drillingsamplingperiod.
Powertodetectchangesof50%areshown,wherechangeswerenon-significant.
*
p<0:05.
**
p<0:01.
230 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
detectable after four months.Molluscs numbers in contrast increased significantly at
site d (p < 0:05) by 596% in the sampling period immediately following drilling,and
remained relatively higher at this site for between 4 and 11 months following drilling.
Densities of two additional phylum groupings (Echinoderms and Others),did not
change significantly at any site following drilling (Tables 5(B) and 6(B)),however,it
appears that abundances for these groupings are simply too low and variable to
determine with high statistical power any changes in abundance coinciding with
drilling.
4.Discussion
The ecological consequences of discharging drill cuttings on to the sea bed depend
on the quantity of material discharged,the physical and chemical nature of the
discharge,the depth of water,and prevailing hydrographic conditions (Davies &
Kingston,1992).Impacts must also depend on the vulnerability of the benthic
communities themselves (Currie &Parry,1999).Irrespective of type,piles of cuttings
have been shown to bury and smother sediments immediately below the discharge
point (Davies & Kingston,1992 cited in Dow,Davies,& Raffaelli,1990;Menzie,
Maurer,& Leathem,1980;Neff,1981).Hydrocarbons may persist in these deposits
and rapid recovery of the seabed may only occur if the cuttings are widely dispersed
or buried by fresh marine sediments (Dow et al.,1990).
A review of studies undertaken to date suggests that the effects of offshore oil and
gas production on benthic communities are both localised and minor (Black et al.,
1994).The general consensus for North Sea operations is that effects are found out to
3 kmfromboring platforms or production rigs with rather severe effects being found
within a 500 m radius (Gray,Bakke,Beck,& Nilssen,1999;Kingston,1987).Such
effects are characterised by changes in benthic community structure and falls in
species diversity the nearer an operational platform is approached (Davies &
Kingston,1992;Olsgard,Somerfield,&Carr,1997).These changes are believed to be
a consequence of the increasing degree of environmental disturbance around the
point of cuttings discharge,and higher impact distances are generally apparent when
oil-based drilling muds are employed or oily cuttings discharged (Bakke,Gray,&
Reiersen,1990;Grant & Briggs,2002).
At Minerva,water based drilling muds and associated cuttings apparently modify
population densities of species at sampling sites up to 200 m distance from the well
head.The most pronounced effects were evident within 100 m of the well-head,
where declines in density of most abundant species exceed 70% immediately fol-
lowing drilling.Such declines in abundance were also conferred on rarer taxa here,
and expressed as reduced species diversity (i.e.,36% decline in the number of species
at the well-head) immediately following drilling.Not all population responses to
drilling were negative however,one bivalve mollusc (Katlysia sp.1) increased in
abundance following drilling,apparently due to preferential recruitment 200 m east
of the well-head.
MDS ordinations summarise many of the changes noted in the abundance and
diversity of taxa at each of the Minerva sampling sites.The ordinations show that
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 231
changes in community structure due to drilling are greatest at the well-head with
lesser changes detectable 100 m west and north,and up to 200 m east of the well-
head site.The decrease in size of drilling effect on community structure with distance
fromthe well head is thought to be a reflection of reduced sedimentation loads at the
more distant sites.A number of studies report that water based drilling muds barely
show an impact beyond that of physical disturbance in which smothering by the
cuttings pile appears to be the most important factor (Menzie et al.,1980;Neff,1981;
Petrazzuolo,1981 cited in Davies & Kingston,1992).
Changes in community structure due to drilling at sites 100 and 200 m distant
from the well-head were small in relation to seasonal changes in community struc-
ture.The effects of drilling on the community structure at these sites does not persist
beyond four months as natural species recruitments swamp residual effects over the
same period.In contrast,benthic communities at the well-head site remain modified
11 months after drilling,in spite of recoveries in species diversity and abundance.It
is most probable that this persistent community difference is due to the physical
modification of the sediment at this site by drill cuttings discharge.
This study has documented the size and duration of impacts to the benthos re-
sultant from exploratory drilling in the Minerva Field.Unfortunately,there is little
ecological theory that allows an assessment of whether such changes are excessive.
Previous attempts to provide generalized models of biological responses to oil and
gas production have concentrated on longer-term impacts of production facilities
(Peterson et al.,1996),and have not considered the short-term impacts of explor-
atory drilling.Given that the abundances of most common benthic species recover
following their next recruitment,it is considered unlikely that exploration drilling at
Minerva causes marked changes in the composition of higher trophic groups.It is
nonetheless noteworthy that pulse impacts from experimental drilling can have a
subtle and persistent influence on benthic community structure,even where the
bedforms are unconsolidated and subject to large and sustained natural impacts
from wave action.
Acknowledgements
This work would not have been possible without financial support from BHP
Petroleum.I am also grateful for the logistical support provided by Sinclair Knight
Merz and the Marine and Freshwater Resources Institute,Queenscliff.In particular,
I thank Brett Kettle,Paul Goldsworthy Simon Conron and the crew of the MV
Starfire for their professional assistance with the field work.Finally,I thank two
anonymous referees for their constructive comments on earlier drafts of this paper.
References
Bakke,T.,Gray,J.S.,& Reiersen,L.O.(1990).Monitoring in the vicinity of oil and gas platforms:
Environmental status in the Norwegian sector in 1987–1989.In:Proceedings of the first international
symposium on oil and gas exploration and production (pp.623–633).US EPA,Louisiana.
232 D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233
Belbin,L.(1990).PATN technical reference manual.Canberra:CSIRO Division of Wildlife and Ecology.
Black,K.P.,Brand,G.W.,Grynberg,H.,Gwyther,D.,Hammond,L.S.,Mourtikas,S.,Richardson,B.
J.,& Wardrop,J.A.(1994).Environmental implications of offshore oil and gas development in
Australia-production activities (pp.217–396).Sydney:Australian Petroleum Exploration Association.
Bray,J.R.,&Curtis,J.T.(1957).An ordination of the upland forest communities of southern Wisconsin.
Ecological Monographs,27,325–349.
Clarke,K.R.(1993).Non-parametric multivariate analyses of changes in community structure.Australian
Journal of Ecology,18,117–143.
Currie,D.R.,& Parry,G.D.(1996).The effects of scallop dredging on a soft-sediment community:A
large scale experimental study.Marine Ecology Progress Series,134,131–150.
Currie,D.R.,& Parry,G.D.(1999).Impacts and efficiency of scallop dredging on different soft
substrates.Canadian Journal of Fisheries and Aquatic Sciences,56,539–550.
Davies,J.M.,&Kingston,P.F.(1992).Sources of environmental disturbance associated with offshore oil
and gas developments.In W.F.Cairns (Ed.),North sea oil and the environment.Developing oil and gas
resources,environmental impact and responses (pp.417–439).Amsterdam:Elsevier.
Dow,F.K.,Davies,J.M.,& Raffaelli,D.(1990).The effects of drill cuttings on a model marine sediment
system.Marine Environmental Research,29,103–134.
Faith,D.P.,Minchin,P.R.,& Belbin,L.(1987).Compositional dissimilarity as a robust measure of
ecological distance.Vegetatio,69,57–68.
Gamito,S.,& Raffaelli,D.(1992).The sensitivity of several ordination methods to sample replication in
benthic surveys.Journal of Experimental Marine Biology and Ecology,164,221–232.
Grant,A.,&Briggs,A.D.(2002).Toxicity of sediments fromaround a North Sea oil platform:Are metals
or hydrocarbons responsible for ecological impacts?Marine Environmental Research,53,95–116.
Gray,J.S.,Bakke,T.,Beck,H.J.,& Nilssen,I.(1999).Managing the environmental effects of the
Norwegian oil and gas industry:Fromconflict to consensus.Marine Pollution Bulletin,38(7),525–530.
Gray,J.S.,Clarke,K.R.,Warwick,R.M.,&Hobbs,G.(1990).Detection of initial effects of pollution on
marine benthos:An example from the Ekofisk and Eldfisk oilfields,North Sea.Marine Ecology
Progress Series,66,285–299.
Kingston,P.F.(1987).Field effects of platform discharges on benthic macrofauna.Philosophical
Transactions of the Royal Society of London,B316,545–555.
Menzie,C.A.,Maurer,D.& Leathem,W.A.(1980).An environmental monitoring study to assess the
impact of drilling discharges in the mid-Atlantic.IV.The effects of drilling discharges on the benthic
community.In:Symposium on research on environmental fate and effects of drilling fluids and cuttings
(pp.499–540).Lake Buena Vista,FL.
Neff,J.M.(1981).Fate and biological effects of oil well drilling fluids in the marine environment:A
literature review.Report to US Environmental Protection Agency,Report No.15077.
Olsgard,F.,Somerfield,P.J.,&Carr,M.R.(1997).Relationships between taxonomic resolution and data
transformations in analyses of a macrobenthic community along an established pollution gradient.
Marine Ecology Progress Series,149(1–3),173–181.
Peterson,C.H.,Kennicutt,M.C.,Green,R.H.,Montagna,P.,Harper,D.E.,Powell,E.N.,&Roscigno,
P.F.(1996).Ecological consequences of environmental perturbations associated with offshore
hydrocarbon production:A perspective on long-term exposures in the Gulf of Mexico.Canadian
Journal of Fisheries and Aquatic Sciences,53,2637–2654.
Petrazzuolo,G.(1981).Preliminary report:An environmental assessment of drilling fluids and cuttings
released onto the outer continental shelf.Prepared by the Industrial Permits Branch,Office of Water
Enforcement and Oceans Programs Branch,Office of Water and Waste Management,US EPA.
Stewart-Oaten,A.,Murdoch,W.M.,& Parker,K.R.(1986).Environmental impact assessment:
‘Pseudoreplication’ in time?Ecology,67,929–940.
Underwood,A.J.(1991).Beyond BACI:Experimental designs for detecting human environmental
impacts on temporal variations in natural populations.Australian Journal of Marine and Freshwater
Research,42,569–587.
D.R.Currie,L.R.Isaacs/Marine Environmental Research 59 (2005) 217–233 233