Diversity of Polychaeta (Annelida) and other worm taxa in mangrove
habitats of Darwin Harbour, northern Australia
K.N. Metcalfea,⁎, C.J. Glasbyb
aSchool of Science, Faculty of Education, Health and Science, Charles Darwin University, Darwin, NT, 0909, Australia
bMuseum and Art Gallery of the Northern Territory, GPO Box 4646, Darwin, NT 0801, Australia
Received 4 October 2006; accepted 5 June 2007
Available online 30 June 2007
In this paper data on the diversity, distribution and abundance of polychaetes and other worm taxa in the mangroves of Darwin
Harbour, northern Australia, are presented and compared with those of other tropical mangrove areas. Aspects of the feeding guild
ecology and the effects of disturbance on mangrove worms are also examined. Data were collected over a period of four years,
across four mangrove assemblages. Samples were obtained using three sampling techniques: 1 m×1 m quadrat searches, epifauna
searches and a new infaunal sampling technique, the anoxic mat. A total of 76 species (68 polychaetes, 1 oligochaete, 1 echiuran, 3
sipunculans, 2 nemerteans, 1 turbellarian) were recorded from the four main mangrove assemblages. Of these, 30 species are
widespread, occurring in mangrove and non-mangrove habitats throughout the Indo-west Pacific. Only seven species (all
polychaetes) appear to be restricted to the mangroves of Darwin Harbour and northern Australia. Polychaetes are predominant,
comprising 80–96% of all worms sampled, with three families—Nereididae, Capitellidae and Spionidae—accounting for 46% of
all species. The highest diversity and abundance was recorded in the soft, unconsolidated substrates of the seaward assemblage,
with diversity and abundance decreasing progressively in the landward assemblages. Most of the worm fauna was infaunal (70%),
but the intensive sampling regime revealed a hitherto unknown significant percentage of epifaunal species (18%) and species
occurring as both infauna and epifauna (12%). Univariate analyses showed annual and seasonal differences in worm species
richness and abundance—presumably associated with the intensity of the monsoon and recruitment success. The worm fauna
differed between mangrove assemblages but the proportion of species in each feeding guild was relatively consistent across the four
assemblages studied. Herbivores were the most species-rich and abundant, followed by carnivores and sub-surface deposit feeders.
Multivariate analyses showed that the species composition of urbanised mangroves differed from that of undisturbed sites, with
surface deposit feeders more numerous in urbanised habitats. Overall, the findings demonstrate a dynamic spatial and temporal
variation in diversity and abundance, and provide insight on the range of microhabitats in which mangrove worms occur and their
response to anthropogenic disturbance.
© 2007 Published by Elsevier B.V.
Keywords: Polychaete worms; Macrobenthos; Feeding guild; Disturbance; Mangroves; Australia
Invertebrate fauna surveys conducted by the senior
author in the mangroves of Darwin Harbour, northern
Australia between 2001 and 2005 yielded a considerable
Journal of Sea Research 59 (2008) 70–82
E-mail address: email@example.com (K.N. Metcalfe).
1385-1101/$ - see front matter © 2007 Published by Elsevier B.V.
amount of data on the distribution, diversity and
abundance of macro-invertebrates (Metcalfe, 2004,
2005, 2007). Crustaceans, molluscs and worms (mainly
polychaetes) were the most species-rich groups sampled
during these surveys. This study presents the results for
the worm fauna, including diversity, abundance and
distribution, within the four main mangrove assem-
blages (hinterland margin, tidal flat, tidal creek and
seaward). Although worms comprise an ecologically im-
portant element of the macro-invertebrate fauna of
mangroves (e.g. Hutchings and Recher, 1982), they
have been relatively poorly studied or neglected (e.g.
MacIntosh et al., 2002). Previous studies in tropical
Australia include those of Wells (1983), who found 15
and 12 polychaete species (and one flatworm in each) in
Avicennia and Rhizophora assemblages respectively in
North-westCape, WesternAustralia;Hanley(1985), who
listed nine polychaete species from mangroves at several
sites in the Northern Territory; and Dittmann (2001),
who found 19 species of polychaetes and a couple of
oligochaetes at Missionary Bay, north Queensland.
The worm fauna of other tropical Indo-west Pacific
mangroves has been documented in several studies
including Sasekumar (1974), Frith et al. (1976), Kumar
(1995), and Guerreiro et al. (1996). All showed a diverse
e.g. Wells, 1983; Hanley, 1985). Kumar (1995) reported
higher faunal diversity during pre-and post-monsoon
months compared to the monsoon months of June and
July. Frith et al. (1976) found ‘a distinct and characteristic
mangrove fauna’ dominated by molluscs, crustaceans and
examined the worm fauna of mudflats immediately
adjacent to mangrove areas (e.g. Hsieh, 1995; Dittmann,
2002; López et al., 2002). Although these areas may
support a similar worm fauna, the habitat is sufficiently
different—especially in terms of insolation, exposure to
currents, and degree of inundation—that a comparison of
few studies have attempted to compile the invertebrate
Hutchings and Recher, 1982; Kumar, 2003), but they are
also not comparable with this study because of the
taxonomic inconsistency between source studies.
Polychaetes in general are good subjects for research
Giangrande et al., 2005), because of their highly diverse
range of feeding and reproductive strategies, which give
them different potentials for responding to disturbance.
and recently examined in the context of environmental
assessment by Pagliosa (2005). This study represents an
opportunity to further examine the relationship between
feeding guild and anthropogenic disturbance, in this case
potential differences between disturbed and undisturbed
mangrove habitats. It is based on two data sets from
Darwin Harbour: a one-year survey in 2001 of three
relatively pristine mangrove sites and four sites affected
directly or indirectly by anthropogenic disturbance
(Metcalfe, 2007); and a three-year study (2003–2005) of
2001 study—which was part of a mangrove monitoring
program examining invertebrate biodiversity (Metcalfe,
2004, 2005). Collectively these data were compared with
specimen records in the database of the Museum and Art
Gallery of the Northern Territory (NTM)—built from
various baseline surveys and environmental assessment
projects—and an annotated list of polychaete and other
worm species was compiled. Where possible, the feeding
guild, ecology and distribution of each species in the
wider context of the tropical Indo-Pacific region were
The specific aims of the study were:
1. To describe the spatial and temporal changes in the
diversity and abundance of worm fauna in mangrove
habitats of Darwin Harbour;
2. To test for the effects of anthropogenic disturbance
on diversity, abundance and trophic composition;
3. To compare the diversity of worm fauna of Darwin
Harbour with that of other tropical mangrove regions
in northern Australia and the wider Indo-Pacific and
assess the level of endemism.
2. Study area
All data were collected within Darwin Harbour,
situated on the north-western coastline of the Northern
Territory between latitudes 12°20′ and 12°40′ S and
longitudes 130°45′ and 131°05′ E. Darwin Harbour is
bounded to the west and east by Charles Point and East
Point, respectively, and contains ∼ 20400 ha of healthy
and relatively intact mangrove and saltflat habitat—
representing one of the largest tracts of mangroves in the
Northern Territory (Fig. 1).
Darwin's climate is tropical, seasonally humid, with
mean annual temperature of 28 °C and 54% relative
humidity. Annual rainfall is ∼ 1713 mm, with wet
summer monsoon and dry winter seasons (Bureau of
Meteorology, 2006). The region is macrotidal, with a
maximum tidal range of 7.8 m and strong bi-directional
tidal velocities. Tides are diurnal (two per day) with a
71K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
(Woodroffe, 1995). Despite strong tidal currents, the
Darwin Harbour estuary is relatively poorly flushed
(Williams et al., 2006), which contributes to characteris-
tically high turbidity levels, particularly during the wet
season. Mangrove substrates generally comprise root-
structured and bioturbated mud and muddy sands with
fine-grained, unconsolidated marine muds in the seaward
assemblages (Semeniuk, 1985).
Over 30 species of mangrove are known from the
Darwin area, with 21 shrub and tree species commonly
tively they form dense mangrove forests that comprise a
number of distinct habitats, indicated by a predictable
assemblages are confined to discrete elevation ranges
(Semeniuk, 1985; Woodroffe and Bardsley, 1987;
Metcalfe, 1999). Four of the ten assemblages recognised
in Darwin Harbour (Brocklehurst and Edmeades, 1996)
occupy 88.2% of the total mangrove area (Fig. 2). The
primary factors influencing the distribution and extent of
mangroves in Darwin Harbour (Woodroffe and Bardsley,
At around mean sea level, open woodlands with
Sonneratia alba occur in soft, unconsolidated sub-
strates. Further landward, tall Rhizophora stylosa forests
occur between ∼ 0.5 and 2 m Australian Height Datum
(AHD). These two assemblages occupy the lower
intertidal zone of the mangrove and are largely shaped
by marine processes—including wave action, tidal
currents and two high tides per day. Only 58% of
annual tides exceed 2 m AHD, however, and assem-
blages in the upper intertidal zone are inundated only by
spring tides, for one week of every fortnight (Metcalfe,
1999). The landward assemblages are thus more
influenced by terrestrial rather than marine processes,
including freshwater seepage, seasonal deposition of
sediments and desiccation. Dense, low (2 to 4 m high)
thickets of Ceriops australis occur in this habitat, partly
in response to increasing soil salinities.
Sampling for the 2001 study was conducted at three
undisturbed mangrove sites (E1, E2 and M3) during wet
Fig. 1. Map showing location of Darwin in the Northern Territory, Australia (left) and the distribution of mangroves (shaded in black) within Darwin
Harbour (right). The seven study sites (E1, E2, M1, M3, W1, W2 and W3) and four disturbed sites (BV, DE, DP and DM) are indicated.
72 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
and dry seasons, and at four disturbed mangrove sites
(BV, DE, DP and DM) during the dry season of the same
housing development situated on 103 ha of reclaimed
mangroves (site BV), on bulldozed tracks through
relatively pristine mangroves (site DE), adjacent to the
(site DP), and adjacent to an earthen channel for a prawn
farm in Middle Arm (site DM). Sampling for the three-
year monitoring program (2003–2005) was done at six
sites (W1, W2, W3, E1, M3 and M1) during the wet and
dry seasons (Fig. 1).
At each site, transects traversing each of the four main
assemblages were established from the landward to
tidal flat, tidal creek and seaward (Fig. 2). Transect length
one permanent study plot was established in each
assemblage (Fig. 2). Study plots for the 2001 study
were 50 m×50 m in size and two transects were estab-
lished at each site. For the monitoring study, study plots
were20 m×20m insizeand one transectwas established
per site. In all other respects, the methodology for both
studies was identical.
Worms were sampled from four distinct micro-
habitats: (i) on the mud surface, (ii) within the substrate,
(iii) on the surface of tree trunks, roots and rocks, and
(iv) within rotting logs. Worms in the first two micro-
habitats were considered infauna and those in the latter
two, epifauna. Within each 0.25 ha or 0.04 ha study plot,
three randomly selected sampling stations were located
at which one pitfall trap, one 0.05 m2‘anoxic mat’ and
one 1m × 1m quadrat were used. The quadrat was placed
number co-ordinates. Active searches of the quadrat and
all surfaces of the tree to a height of 2 m (including roots
and foliage) were conducted and quadrats were also dug
to a depth of ∼ 5–10 cm.
The anoxic mat comprised a plastic disc that tem-
porarily created a localised area of anoxic mud. It was
placed flat on the mud surface, covered with a mound of
mud to maintain anoxic conditions and left for up to
24 h. The following day the mat was peeled back and
specimens collected, by eye, from the mud surface. The
mud beneath the mat was also searched, by digging
with a trowel to a depth of approximately 5 cm. The
technique was very effective at capturing small poly-
chaetes overlooked by other methods, but possibly
under-sampled taxa having a high tolerance of anaerobic
conditions (e.g. some oligochaetes); the effectiveness of
the anoxic mat compared to other benthic sampling
devices is described elsewhere (Metcalfe, 2007). All
specimens sampled were preserved in 70% ethanol, and
lodged with the NTM for identification (references
specimens were registered with the NTM).
Specimens identified to genus or species levels were
included in statistical analyses and species tallies.
Specimens identified only to family (usually because
they were in poor condition) or phylum level (lack of
taxonomic expertise) were omitted from the analyses,
unless they represented the only member of that family,
obtained during all surveys were omitted. Each of the
analysed taxa was assigned to one of 22 feeding guilds
and five trophic categories—herbivore, carnivore, filter
Fig. 2. Permanent study plots were placed within the four assemblages on paired transects from the landward to seaward margin (left). Sampling was
conducted at three randomly placed subplots within each study plot (middle) using quadrats and anoxic mats (right).
73 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
feeder, surface deposit feeder and subsurface deposit
Pagliosa (2005)]—in order to assess trophic character-
istics of the worm fauna.
For the three-year study, species richness and
abundance data were compared between years, seasons,
sites and assemblages using a 4-factor ANOVAwith all
factors fixed. Tests for ANOVA assumptions were run
prior to the analysis, by examination of normal plots of
within-cell residuals and plots of means versus standard
deviations, before and after transformation. Abundance
data was transformed (log10(x+1)) but transformation
was not necessary for species richness data. Analyses
were conducted using either Statistica or the General
Linear Model in Minitab. By convention, significance
levels were set at pb0.05.
All comparisons of disturbed and undisturbed man-
groves were based on the dry season survey of 2001. The
in 2003–2005, as paired transects were sampled at each
location. ANOVA for species richness, abundance and
feeding guild datainvolved 4-factor, nestedanalyseswith
inwhich all factors werefixed excepttransect, which was
random and nested in location and disturbance.
The two data sets from undisturbed sites were
merged for multivariate analyses, to permit examination
of community data spanning four years. Ordination by
non-metric multi-dimensional scaling (nMDS) proce-
dures in PRIMER (Clarke and Warwick, 1994; Clarke
and Gorley, 2001) was used to examine community
patterns in worm diversity and abundance, and for
comparison of disturbed and undisturbed sites. Ordina-
tions were generated using Bray Curtis dissimilarity on
untransformed data after 50 random restarts.
A total of 216 records were obtained for worms
during three surveys in 2001 (one wet season, two dry
seasons) and 810 records for the three-year survey of six
sites (three wet and three dry seasons).
4.1. Diversity, distribution and habitat
from mangrove habitats comprising 68 polychaetes,
1 oligochaete (Annelida), 1 echiuran, 3 sipunculans,
naturalsciences/annelids.html). In any one sample,
polychaetes were predominant, comprising 80–96%
of all taxa. Only one species, Mastobranchus sp.,
previously reported from Darwin Harbour mangroves
was not recollected in this study. Seven species appear
and northern Australia – the polynoid Lepidonotus sp.
1, three capitellids (Heteromastus sp. 1, sp. 2 and
Mastobranchus sp.) and three nereidids (Ceratonereis
sp. NTM6742, Namalycastis nicolea and Paraleon-
nates bolus). Thirty-three species occur in mangrove
and non-mangrove habitats throughout the Indo-West
Pacific; the remainder are too poorly known taxonom-
ically for distributions to be analysed.
In terms of microhabitat, the majority of species are
infaunal (70%), but a substantial portion also occurs
as epifauna (18%) and about 12% of species occur
as both (http://www.nt.gov.au/nreta/museums/magnt/
Most species avoided the sediment surface, the excep-
tions being the large nereidid Paraleonnates bolus and
the phyllodocid Phyllodoce sp., which mostly was
collected in pitfall traps. Epifaunal species mostly
occurred under the bark of mangrove trees, but also in
fallen timber. Particularly productive microhabitats
were beneath the large flakes of bark on the lower
trunks of Sonneratia alba in the seaward assemblage
and within rotting, burrow-structured roots and limbs
of Rhizophora stylosa in the tidal creek assemblage.
Certain species including Lepidonotus sp. 1, Neanthes
cf. biseriata and Perinereis singaporiensis were almost
exclusively sampled from the trunks of Sonneratia alba.
4.2. Species richness and abundance
Within the mangroves, the overall species richness
and abundance of worms at the six sites sampled during
Mean species richness and abundance of worms per m2(±SE) in the
four mangrove assemblages during wet and dry seasons over three
AssemblageDry Season Wet season
0.1±0.03 0.06±0.030.1±0.05 0.13±0.05
Data from 1×1 m quadrats and epifaunal counts were used to calculate
74K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
the three-year monitoring program were not significant-
ly different. Also, diversity levels for each of the four
assemblages were reasonably consistent between sites.
However, within each site univariate analysis showed
significant annual and seasonal differences in species
richness and abundance between assemblages. The
seaward assemblage had the highest diversity of
worms, with species richness and abundance decreasing
progressively to landward, with few worms sampled in
the hinterland margin (Table 1; Fig. 3). This distribution
pattern reflects the frequency of tidal inundation and the
suitability of substrates—muds become increasingly
moist and unconsolidated to seaward and the habitat
opportunities for infauna increase.
Significant year×assemblage and season×assem-
blage interactions were also found, indicating distinct
annual differences amongst assemblages and that the
effects of season on diversity are dependent on
assemblage. For instance, species richness is higher in
the seaward assemblages in the dry season, whereas it
decreases in the two landward assemblages in the dry
season (Fig. 4). The mean squares indicated that the
most significant factor determining worm species
richness and abundance was mangrove assemblage.
Overall mean abundance decreased by almost 50% in
the seaward assemblage during the wet (monsoon)
season, when rainfall, erosion and wave action peak. In
contrast, abundance increased in both the tidal flat and
hinterland margin assemblages during the wet season
Fig. 5). Worm abundance varied between years but this
was determined by assemblage, as was the effect of
season. Overall worm abundance showed an apparent
increase from 2001 to 2005 (Fig. 5), a pattern that was
mirrored by other invertebrate groups, e.g. crustaceans
and molluscs (Metcalfe, 2007).
Multivariate analyses showed that mangrove assem-
blage is the primary determinant of species composition.
thepresenceandabundance ofworms(Fig.6).The worm
fauna of the tidal creek assemblage shows similar affinity
between the different locations sampled, but there is also
Fig. 4. Annual (and seasonal) variation in worm species richness (left) and abundance (right) in the four main assemblages. Points are means per
sampling station (±SE); data are pooled over six sites.
Fig. 3. Variation in mean worm species richness (left) and abundance (right) per sampling station (±SE) in the four main mangrove assemblages. Data
are pooled over six sites and three years.
75 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
some overlap with study sites from both other assem-
blages. The tidal flat and hinterland margin assemblages
havea distinctwormfauna,althoughnot asprolificasthe
seaward assemblages (Fig. 6). Typical species in the
hinterland margin and tidal flat are the nereidids Nama-
nereis malaitae and Paraleonnates bolus and Phyllodoce
sp. (Phyllodocidae). The tidal creek and seaward
assemblages are characterised by a greater number of
species including the ampharetid Isolda pulchella, the
lumbrinerid Scoletoma sp. 1, the magelonid Magelona
sp. 1, the nereidids Nereis sp. 1 and Perinereis
singaporiensis, the orbiniid Leitoscoloplos latibranchus
and the polynoid Lepidonotus sp. 1. Only two species—
the nereidid polychaete Perinereis aibuhitensis and the
sipunculan Phascolosoma arcuatum—occurred in all
four mangrove assemblages, and both have a widespread
Indo-west Pacific distribution.
of occurrence of worms in different microhabitats—few
infauna in the tidal creek and seaward assemblages.
Typical epifaunal species include the eunicid Nematoner-
eis sp., the nereidids Ceratonereis australis, Neanthes cf.
biseriata, Perinereis singaporiensis, the scaleworm Lepi-
donotus sp. 1, the serpulid Pomatoleios kraussii and a
syllid, Syllis sp. 1. A few infaunal species are also
characteristic of the seaward assemblage including Isolda
pulchella (Ampharetidae), the lumbrinerids Arabelloneris
broomensis and Scoletoma sp. 1, Nephtys mesobranchia
(Nephtyidae) and Nereis sp. 1 (Nereididae).
4.3. Feeding guild
All five trophic categories and 13 of the 22 feeding
guilds were identified among the worm taxa collected in
well represented (http://www.nt.gov.au/nreta/museums/
The proportion of worms per feeding guild remained
the exception of the tidal flat (Ceriops australis)
assemblage where subsurface deposit feeders were more
numerous than carnivores and herbivores. Overall,
herbivores were the most numerous, with carnivores,
subsurface deposit feeders and surface deposit feeders
in decreasing order of abundance (Fig. 7). Although
herbivores were treated together in one group in this
analysis, they can be subdivided further into diatom-and
macrophyte-feeders (Fauchald and Jumars, 1979). Prob-
ably the majority of herbivores in this study are of the
former type (i.e., microphagous), and feed not only on
diatoms, but also algae and detritus. Polychaetes are also
potentially capable of breaking down whole mangrove
leaves (e.g. Camilleri, 1992) so it is likely that some
species would also utilise this abundant food source.
Fig. 5. Mean worm abundance (±SE) indicating seasonal and annual variation. Differences in mean abundance in the four assemblages (left) and
during wet and dry season during four years of sampling (right). Points represent mean numbers of worms per sampling station, averaged over three
locations (2001 data) or six locations (2003–2005 data) shown from seaward (left) to landward (right).
Fig. 6. nMDS ordination of 107 study plots surveyed over four years
based on the abundance of 69 polychaete species indicating the
similarity of study plots in different assemblages. Points represent data
pooled for each sampling technique at three replicate sampling stations
per study plot.
76 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
4.4. Fauna of disturbed mangroves
The one-year survey of three undisturbed and four
the worm faunal assemblages in mangroves directly or
indirectly affected by anthropogenic development, with
undisturbed sites. Although mean diversity was lower
and abundance slightly higher in disturbed mangroves,
univariate analyses found no significant differences in
overall mean species richness and abundance between
disturbed and undisturbed mangroves. A significant
disturbance × assemblage interaction for worm abun-
dance indicated that abundance in disturbed and
undisturbed sites differed between assemblages. Mean
worm abundance is apparently higher in disturbed sites
in the tidal creek and to a lesser extent the seaward
assemblage, but is apparently lower than undisturbed
sites in the landward assemblages (Fig. 8).
Multivariate analyses illustratedthattheworm fauna of
urbanised mangroves differs somewhat from that of
undisturbed sites (Fig. 9). Seventeen species, of the total
dry seasons, revealed a wider distribution for many of
those species, including undisturbed sites. Although not
always exclusive to disturbed habitats, the surface deposit
feeders Aphelochaeta sp. 1, Leonnates stephensoni, Sco-
lelepis sp. 1, Terebellides kowinka, and Terebella tanta-
biddycreekensis and the herbivore (detrital) feeder,
Simplisetia cf erythraensis appear characteristic of urba-
Fig. 7. Mean abundance of worms in the five main feeding guilds at six sites during 2003–2005. Means per sampling station (±SE) in the four
assemblages are shown from seaward (L) to landward (R).
Fig. 8. Mean abundance of worms (±SE) in disturbed (D) and
undisturbed (U) sites, in the four assemblages. Means are pooled
across three undisturbed and two disturbed sites from one dry season
survey in 2001.
Fig.9. Ordination of 13disturbedand14undisturbedstudyplotsbased
on worm taxon presence/absence. Dry season data from three replicate
sampling stations were pooled for each study plot.
77 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
The abundance of worms in three of the four main
feeding guilds varied little between disturbed and
undisturbed sites (Fig. 10). Univariate analyses found
that the abundance of surface deposit feeders differed
between disturbed and undisturbed sites; this trophic
disturbance × assemblage interaction also indicated that
pronounced in the seaward assemblage (Fig. 10). These
results were, however, based on one dry season survey
during 2001 and further sampling is required to
substantiate these findings.
5.1. Diversity and abundance
The diverse and extensive mangrove environments
of Darwin Harbour, provide habitat for a surprisingly
rich worm fauna in which polychaetes predominate,
comprising 80–96% of all worms sampled. Three
polychaete families—Nereididae, Capitellidae and
Spionidae—accounted for 46% of all species. The
dominance of these three polychaete families coincides
with the results of other studies of mangrove systems
(Frith et al., 1976; Hutchings and Recher, 1982; Kumar,
1995). The total number of taxa reported in the study
(76 species; 56 genera) is, however, far in excess of any
other mangrove study in the Indo-west Pacific, but this
is likely to be the result of the more extensive sampling
regime over a longer period of time. Studies of one-year
duration or less and those that consider only the infauna
have yielded less than half the number of genera (e.g.
Sasekumar, 1974; Guerreiro et al., 1996; Dittmann,
2001). Multi-year studies, therefore, are more likely to
sample short-lived species (one year or less), which may
be absent altogether in years following poor recruitment.
The four-year period of this study was sufficient in
duration to sample almost all of the polychaete species
currently known from Darwin Harbour mangroves,
based on literature records and polychaete specimen
records (spanning over 30 years) in the database of
NTM. The only species known to occur in Darwin
Harbour mangroves but not sampled in this study is the
capitellid, Mastobranchus sp., which was also reported
by Hanley (1985) as Heteromastus sp. A; it may have a
specialised habitat (tailings of mud lobster burrows).
Of the other worm taxa recorded in this study,
nemerteans were also abundant (8.5% of all records) but
the taxonomy of this group is poorly known and apart
from the large, rarely encountered Cerebratulus species
of the seaward zones, other species could not be
identified. Sipuncula were also abundant (6.5% of all
records) with most belonging to the widely distributed
Indo-west Pacific estuarine species, Phascolosoma
arcuatum; two other, possibly undescribed, species
were also found in the tidal creek zone.
The spatial patterns in diversity and abundance were
remarkably consistent between locations in Darwin
Harbour such that the species richness and abundance
Fig. 10. Mean abundance of worms in the four main feeding guilds in disturbed (D) and undisturbed (U) mangroves (left). Means represent average
abundance per sampling station, pooledacross 78 disturbed and 72 undisturbed replicates. Mean abundance of surface deposit feeders (±SE)indisturbed
(D) and undisturbed sites (U) in the four assemblages (right). Means are pooled for two disturbed (BV, DP) and two undisturbed sites (E2 , M3).
78 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
predictable—with highest diversity and abundance in the
seaward mangrove assemblage, decreasing progressively
to landward, with few worms sampled in the hinterland
margin. This distribution pattern presumably reflects the
frequency of tidal inundation and the suitability of
substrates—muds become increasingly moist and uncon-
solidated to seaward and the habitat opportunities for
epifauna increase. In addition, frequency of tidal
inundation directly influences recruitment such that the
more frequently inundated seaward zones have a greater
potential to receive larvae, and for larvae to survive. For
example, species occurring in the tidal creek assemblage
(∼ 1 m AHD) are inundated by 93% of annual tides,
(N 2 m AHD) are only bathed by 58% of tides. High dry
conditions which further add to the harshness of the mid-
during the wet season moderate the harsh environmental
conditions in the tidal flat and the hinterland margin,
which may contribute to the higher diversity and
abundance observed in landward zones during the wet
season. Desiccation and high salinity are likely to be
important factors limiting worm populations in these
On the contrary, the levels of diversity in the seaward
assemblages decreased during the wet season. Mon-
soonal conditions during the wet season generate swell
and wave action, typically leading to erosion of surface
sediment in the seaward assemblages during this period
(K. Metcalfe, pers. obs.). Monsoonal conditions may
drastically alter sediment characteristics, such as particle
size and can exert a strong seasonal impact on the
macrobenthos (Alongi and Sasekumar, 1992), especial-
ly polychaetous worms (Sarkar et al., 2005). Kumar
(1995) also reported lessened faunal diversity during the
monsoonal period in the mangroves of Cochin, India.
Polychaetes may be seriously affected by erosion and
reduced salinity (Kurian, 1984), while other phyla are
not (Nandi and Choudhury, 1983 as cited in Alongi and
Sasekumar (1992)). Recent research in north Queens-
land indicates that during intense, short-term freshwater
inundations, the majority of benthic species in mudflats
just seaward of mangroves are lost and do not return; a
small remnant fauna remains—comprising euryhaline
‘resident’ species—which slowly recover to pre-distur-
bance levels (J. Sheaves, pers. comm., 2006).
The gradual increase in diversity and abundance
observed during 2001 to 2005 may be a response to the
season of 2003–2004 had an above average rainfall and
2001 values were, however, even lower than 2003.
Nevertheless, the observed increase may represent
gradual recovery of invertebrate populations during
years in which the monsoon was more moderate. It is
unlikely, given the methodological consistency that the
increase is due to an artefact of sampling or improved
discrimination in the field or laboratory. Determination of
the factors influencing worm populations is, however,
beyond the scope of this project. Forthcoming surveys
may provide further insight into the long-term patterns in
diversity and abundance of mangrove worms.
The higher number of subsurface deposit feeders in
the tidal flat may be associated with the number of mud
lobster mounds, which provide excellent habitat (soft,
reworked mud) for worms. The capitellids – Hetero-
mastus sp. 1, Mediomastus sp. 1 and Notomastus sp. –
appear to be the main taxa responsible for this pattern.
The higher number of surface deposit feeders in
disturbed sites is the only significant difference detected
between disturbed and undisturbed mangroves, but it
was only based on one year's sampling and therefore
requires corroboration. If supported, it suggests that a
shift in polychaete trophic assemblages, such as the
sudden dominance of surface deposit feeders, could be a
good indicator of disturbance, such as increased
5.2. Effect of anthropogenic disturbance
Probably the most significant type of anthropogenic
disturbance to the mangroves of Darwin Harbour is
associated with urbanisation. Urbanisation can affect
sediment properties when runoff, currents, tidal flow
and the ability of mangrove trees to capture sediments
are altered (Kaly et al., 1997). By contrast, organic
enrichment associated with pollution is anticipated to be
relatively minor, if present, at the disturbed study sites in
Darwin Harbour. Sediment properties are a primary
factor determining polychaete populations, especially
grain size (Alongi, 1987; Pagliosa, 2005; Sarkar et al.,
2005) and silt and clay content (Hsieh, 1995). Increased
populations of polychaetes at several sites and increases
in the abundance of surface deposit feeders may have
been related to changes in the sediment.
disturbance need to be interpreted with caution,
however, as the work did not document the direct res-
ponse of the worm fauna to disturbance. The faunal
differences observed in disturbed sites may, to some
extent, also be due to intrinsic environmental differences
between sites. Substrates at the port site, for example,
79K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
may naturally have been more sandy, gravely or rocky
than at the undisturbed sites studied—with a specialised
worm fauna that reflected this. Pre-disturbance surveys
these studies have provided valuable baseline informa-
tion and further research on the response of mangrove
polychaetes to anthropogenic disturbance is seen as a
priority for environmental assessment and management
of mangrove communities in Darwin Harbour. The
preliminary results obtained here suggest that of all the
invertebrate groups studied in mangrove environments,
polychaetes may be the most useful as key indicators of
5.3. Mangrove worms: characteristic or specialised
Hutchings and Recher (1982: p. 102) point out that
‘Relatively few animals are restricted to mangroves or
show specific adaptations to the mangrove environ-
ment.’ This appears to be the case for the majority of
species encountered in this study. Of the 76 worm
species reported in this study from Darwin Harbour
mangroves, only seven (∼10%) may be restricted to this
environment. The remainder that are well enough
known (at least 33 species) are also present on adjacent
mudflats and channels, and other intertidal non-
mangrove shallow coastal habitats in northern Australia.
Several species including the polychaetes Marphysa
mossambica, Dendronereides heteropoda, Namalycas-
tis abiuma, Perinereis aibuhitensis, Simplisetia cf ery-
thraensis, and the sipunculan, Phascolosoma arcuatum
have been reported from other mangrove areas in
northern Australia and the Indo-west Pacific and are
characteristic members of the Indo-west Pacific man-
No previous study has identified an endemic or spe-
cialised mangrove worm fauna, perhaps because the
knowledge of polychaetes and other worms (especially
of the tropics) is not mature enough to know with any
degree of confidence the taxonomic limits and distribu-
tions of each species. This is also true of the seven
species identified here as possible endemics (http://www.
naturalsciences/annelids.html). Of these seven species,
the three most likely to be endemics are: Mastobranchus
sp., which appears to be confined to the mounds of the
mud-lobster Thalassina squamifera. This species was not
collected in the present study, possibly because Thalas-
sina mounds were rarely sampled in the sampling
strategy, which placed 1 m×1 m quadrats against
randomly selected trees. It was one of only two species
found by Hanley(1985)tobe exclusively associated with
Thalassina mounds; the other one Neanthes sp. B
(=Perinereis aibuhitensis) was found here to occur more
widelyacross all mangrove assemblages bothas epifauna
and infauna. Another nereidid, the epifaunal Namaly-
castis nicoleae, is mainly found under the bark of Son-
is only known from Darwin Harbour mangroves and a
drainage channel of reclaimed mangroves near Brisbane
1 also appears to be restricted to Darwin Harbour
mangroves. The occurrence of endemic polychaetes in
spatial extent, floristic diversity and the habitat complex-
known from other animal groups occurring in tropical
mangroves (Hutchings and Recher, 1982: p.102).
At the generic level, a similar suite of ‘characteristic’
worm taxa exists. Polychaete genera common to both
Darwin Harbour mangroves and the mangroves of other
Indo-west Pacific mangroves include, the ampharetid
Amphicteis, the nereidids Composetia, Dendroner-
eides, Dendronereis, Neanthes, Nereis, Perinereis and
Simplisetia, the onuphid Diopatra, the maldanid Eu-
clymene, the glycerid, Glycera, the capitellids Hetero-
mastus and Mediomastus, the polynoid Lepidonotus,
the eunicid Marphysa, the phyllodocid Phyllodoce, the
lumbrinerid Scoletoma, the spionids Polydora, Prio-
nospio and the orbiniid, Scoloplos. The only non-
polychaete worm so far reported from more than one
Indo-Pacific mangrove area is the sipunculan, Phasco-
losoma, which is a genus typical of Indo-west Pacific
hard substrates (Cutler and Cutler, 1990). Most of these
are species-rich genera whose members occupy a wide
variety of habitats globally. Thus, while these genera can
be considered characteristic of mangroves, they are not
mangrove specialists. The high number of genera (56) in
Darwin Harbour mangroves that have not been
previously reported from other Indo-west Pacific
mangroves probably reflects the poor state of taxonomic
knowledge of polychaetes, particularly in the tropics.
1. The diverse mangrove worm fauna of Darwin
Harbour is dominated by polychaetes, especially
Nereididae, Capitellidae and Spionidae.
2. The majority of species, and many genera, are
characteristic of mangrove areas across the Indo-
Pacific; only about 10% of species may be endemic
to the mangroves of northern Australia, but further
studies are required to test this hypothesis.
80K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
3. The distribution and abundance of species in Darwin
years and seasons) and space (between mangrove
to site, however, with a reasonably predictable suite of
species occurring at particular tidal elevations.
4. The microhabitats from which worms were sampled
differed markedly from the seaward assemblages.
5. Theseasonalpattern ofincreaseddryseasondiversity
and abundance to seaward is reversed to landward,
where it declined in response to desiccation.
6. A different species and trophic composition between
disturbed and undisturbed mangroves is also sug-
gested by the data.
7. Herbivores are the most abundant trophic group
are relatively more common.
We thank Julie Crawford and Joe Darcy for field
assistance, Keith McGuinness and Tim Glasby for
comments on the manuscript and Joe Lee for the
invitation to present the paper at the Second Interna-
tional Mangrove Macrobenthos Meeting, Coolangatta,
25–30 June 2006. In particular, we acknowledge the
proponents of the mangrove monitoring program for
permission to publish these valuable baseline data. This
work was undertaken with support from an Australian
Postgraduate Award and the Parks and Wildlife
Commission of the Northern Territory.
Alongi, D.M., 1987. Intertidal zonation and seasonality of meio-
benthos in tropical mangrove estuaries. Mar. Biol. 95, 447–458.
Alongi, D.M., Sasekumar, A., 1992. Benthic communities. In:
Robertson, A.I., Alongi, D.M. (Eds.), Tropical Mangrove Ecosys-
tems.American Geophysical Union,Washington,DC,pp.137–172.
Harbour, Northern Territory, Australia. Land Conservation Unit,
Conservation Commission of the Northern Territory, Darwin, NT.
Bureau of Meteorology, 2006. Climate averages for Station 014015
Darwin Airport, NT during the period 1941–2006, Common-
wealth Bureau of Meteorology Website www.bom.gov.au. 2006.
Camilleri, J.C., 1992. Leaf-litter processing by invertebrates in a
mangrove forest in Queenland. Mar. Biol. 114, 139–145.
Clarke, K.R., Gorley, R.N., 2001. Primer v5: User Manual. Plymouth
Clarke, K.R., Warwick, R.M., 1994. Change in Marine Communities:
An Approach to Statistical Analysis and Interpretation. Plymouth
Marine Laboratory, Primer-E Ltd.
Cutler, N.J., Cutler, E.B., 1990. Revision of the genus Phascolosoma
Dittmann, S., 2001. Abundance and distribution of small infauna in
mangroves of Missionary Bay, North Queensland Australia. Rev.
Biol. Trop. 49, 535–544.
Dittmann, S., 2002. Benthic fauna in tropical tidal flats of
Hinchinbrook Channel, NE Australia: diversity, abundance and
their spatial and temporal variation. Wetlands Ecol. Manag. 10,
Fauchald, K., Jumars, P.A., 1979. The diet of worms: a study of
polychaete feeding guilds. Oceanogr. Mar. Biol. Ann. Rev. 17,
Frith, D.W., Tantanasiriwong, R., Bhatia, O., 1976. Zonation of
macrofauna on a mangrove shore, Phuket Island. Res. Bull. -
Phuket Mar. Biol. Cent. 10, 1–37.
Giangrande, A., Licciano, M., Musco, L., 2005. Polychaetes as
environmental indicators revisited. Mar. Pollut. Bull. 50,
1. Taxonomy and Phylogeny. Rec. Aust. Mus., Suppl. 25, 1–129.
Guerreiro, J., Freitas, S., Pereira, P., Paula, J., Macia Jr., A., 1996.
Sediment macrobenthos of mangrove flats at Inhaca Island,
Mozambique. Cah. Biol. Mar. 37, 309–327.
Hanley, J.R., 1985. Why are there so few Polychaetes (Annelida) in
northern Australian mangroves? In: Bardsley, K.N., Davie, J.D.S.,
Woodroffe, C.D. (Eds.), Coasts and Tidal Wetlands of the
Australian Monsoon Region. North Australia Research Unit,
Australian National University, Darwin, NT, pp. 239–250.
Hsieh, H.-L., 1995. Spatial and temporal patterns of polychaete
communities in a subtropical mangrove swamp: influences of
sediment and microhabitat. Mar. Ecol. Prog. Ser. 127, 157–167.
Hutchings, P., Recher, H.F., 1982. The fauna of Australian mangroves.
Proc. Linn. Soc. N. S. W. 106, 83–121.
Kaly, U.L., Eugelink, G., Roberston, A.I., 1997. Soil conditions in
damaged north Queensland mangroves. Estuaries 20, 291–300.
Kumar, R.S., 1995. Macrobenthos in the mangrove ecosystem of
Cochin backwaters, Kerala (southwest coast of India). Ind. J. Mar.
Sci. 24, 56–61.
Kumar, R.S., 2003. A checklist of polychaete species some mangroves
of Asia. Zoos' Print J. 18, 1017–1020.
Kurian, C.V., 1984. Fauna of the mangrove swamps in Cochin estuary.
In: Soepadmo, E., Rao, A.N., MacIntosh, D.J. (Eds.), Proc. Asian
Symp. Mangrove Environments: Research and Management.
University of Malaya and UNESCO, Kuala Lumpur, pp. 226–230.
López, E., Cladera, P., San Martín, G., Laborda, A., Aguado, M.T.,
2002. Polychaete assemblages inhabiting intertidal soft bottoms
associated with mangrove systems in Coiba National Park (East
Pacific, Panama). Wetlands Ecol. Manag. 10, 233–242.
MacIntosh, D.J., Ashton, E.C., Havanond, S., 2002. Mangrove
rehabilitation and intertidal biodiversity: a study in the Ranong
Metcalfe, K., 1999. Mangrove litter production, Darwin Harbour. A
study of litter fall as a measure of primary productivity in the
mangrove communities of Darwin Harbour. Master of Tropical
Environmental Management Thesis, Faculty of Biological and
Environmental Science, Northern Territory University, 84 pp.
Metcalfe, K., 2004. Mangrove Invertebrate Fauna. Darwin LNG
ProjectEnvironmental MonitoringProgram, Darwin,NT. Prepared
for URS Australia Pty Ltd, 55 pp.
Metcalfe, K., 2005. Mangrove Invertebrate Fauna. Darwin LNG
Project Mangrove Monitoring Program, Annual Report 2004–
2005. Prepared for URS Australia, 94 pp.
Metcalfe, K., 2007. The biological diversity, recovery from distur-
banceandrehabilitationof mangroves,Darwin Harbour, NT. Ph.D.
81 K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82
Thesis, Faculty of Education, Health and Science, Charles Darwin Download full-text
University, Darwin, Northern Territory, Australia.
Pagliosa, P.R., 2005. Another diet of worms: the applicability of
polychaete feeding guilds as a useful conceptual framework and
biological variable. Mar. Ecol. 26, 246–254.
Saenger, P., Specht, M.M., Specht, R.L., Chapman, V.J., 1978. Mangal
(Ed.), Ecosystems of the World. Wet Coastal Ecosystems, vol. I.
Elsevier, Amsterdam, pp. 293–345.
Sarkar, S.K., Bhattacharya, S.G., Bhattacharya, B., Sarkar, D., Nayak,
D.C., Chattopadhaya, A.K., 2005. Spatiotemporal variation in
benthic polychaetes (Annelida) and relationships with environ-
mental variables in a tropical estuary. Wetlands Ecol. Manag. 13,
Sasekumar, A., 1974. Distribution of macrofauna on a Malayan
mangrove shore. J. Anim. Ecol. 43, 51–69.
Territory: the physical framework and habitats. J. R. Soc. West.
Aust. 67, 81–97.
Wells, F.E., 1983. An analysis of marine invertebrate distributions in a
mangrove swamp in northwestern Australia. Bull. Mar. Sci. 33,
Wightman, G.M., 1989. Mangroves of the Northern Territory.
Conservation Commission of the Northern Territory, Palmerston.
Williams, D., Wolanski, E., Spagnol, S., 2006. The hydrodynamics of
Darwin Harbour. In: Wolanski, E. (Ed.), The Environments in Asia
Pacific Harbours. Springer, Dordrecht, Netherlands, pp. 461–476.
Woodroffe, C.D., 1995. Response of tide-dominated mangrove
shorelines in northern Australia to anticipated sea-level rise.
Earth Surf. Proc. Landf. 20, 65–85.
Woodroffe, C.D., Bardsley, K.N., 1987. The distribution and
productivity of mangroves in Creek H, Darwin Harbour. In:
Larson, H.K., Michie, M.G., Hanley, J.R. (Eds.), Proc. Workshop
on Research and Management Held in Darwin, 2–3 September
1987. Mangrove Monograph, vol. 4. North Australia Research
Unit, Australian National University, Darwin, pp. 82–122.
82K.N. Metcalfe, C.J. Glasby / Journal of Sea Research 59 (2008) 70–82