Flood related loss and recovery of intertidal seagrass meadows in southern Queensland, Australia

Estuarine, Coastal and Shelf Science 60 (2004) 477e490 www.elsevier.com/locate/ECSS

Flood related loss and recovery of intertidal seagrass meadows in southern Queensland, Australia Stuart J. Campbella,b, Len J. McKenziea,b,) a

Marine Plant Ecology Group, Northern Fisheries Centre, Department of Primary Industries, PO Box 5396, Cairns, Queensland, Australia 4870 b CRC Reef, PO Box 772, Townsville, Queensland 4810, Australia Received 20 October 2003; accepted 17 February 2004

Abstract The loss and recovery of intertidal seagrass meadows were assessed following the flood related catastrophic loss of seagrass meadows in February 1999 in the Sandy Strait, Queensland. Region wide recovery rates of intertidal meadows following the catastrophic disturbance were assessed by mapping seagrass abundance in the northern Great Sandy Strait region prior to and on 3 occasions after widespread loss of seagrass. Meadow-scale assessments of seagrass loss and recovery focussed on two existing Zostera capricorni monitoring meadows in the region. Mapping surveys showed that approximately 90% of intertidal seagrasses in the northern Great Sandy Strait disappeared after the February 1999 flooding of the Mary River. Full recovery of all seagrass meadows took 3 years. At the two study sites (Urangan and Wanggoolba Creek) the onset of Z. capricorni germination following the loss of seagrass occurred 14 months post-flood at Wanggoolba Creek, and at Urangan it took 20 months for germination to occur. By February 2001 (24 months post-flood) seagrass abundance at Wanggoolba Creek sites was comparable to pre-flood abundance levels and full recovery at Urangan sites was complete in August 2001 (31 months post-flood). Reduced water quality characterised by 2e3 fold increases in turbidity and nutrient concentrations during the 6 months following the flood was followed by a 95% loss of seagrass meadows in the region. Reductions in available light due to increased flood associated turbidity in February 1999 were the likely cause of seagrass loss in the Great Sandy Strait region, southern Queensland. Although seasonal cues influence the germination of Z. capricorni, the temporal variation in the onset of seed germination between sites suggests that germination following seagrass loss may be dependent on other factors (eg. physical and chemical characteristics of sediments and water). Elevated dissolved nitrogen concentrations during 1999 at Wanggoolba Creek suggest that this site received higher loads of sediments and nutrients from flood waters than Urangan. The germination of seeds at Wanggoolba Creek one year prior to Urangan coincides with relatively low suspended sediment concentrations in Wanggoolba Creek waters. The absence of organic rich sediments at Urangan for many months following their removal during the 1999 flood may also have inhibited seed germination. Data from population cohort analyses and population growth rates showed that rhizome weight and rhizome elongation rates increased over time, consistent with rapid growth during increases in temperature and light availability from May to October. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: seagrass; loss; recovery; flood; water quality; mapping

1. Introduction Large-scale disturbance and loss of seagrasses associated with natural and human impacts have occurred both in Australia (Poiner et al., 1989; Preen et al., 1995; Seddon et al., 2000) and worldwide (Short and ) Corresponding author. Marine Plant Ecology Group, Northern Fisheries Centre, Department of Primary Industries, PO Box 5396, Cairns, Queensland, Australia 4870. E-mail address: [email protected] (L.J. McKenzie). 0272-7714/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecss.2004.02.007

Wyllie-Echeverria, 1996). Disturbance can occur on small localised scales (sewage outfalls, dredging, dugong grazing) or at large scales from flooding (Quammen and Onuf, 1993; Moore et al., 1997) and climatic related impacts (Seddon et al., 2000; Cabello-Pasini et al., 2002). Changes to land use patterns in the coastal zone have exacerbated the effects of sediment loading and eutrophication of large-scale flooding events on seagrass ecosystems (Preen et al., 1995; Terrados et al., 1998). Reductions in light (Bach et al., 1998; Halun et al., 2002) and increased concentrations of silt, organic matter and

478

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

nutrients (Kamp-Nielson et al., 2001) negatively affect seagrass growth by increasing organic matter decomposition and the bacterial oxygen demand in sediments (Terrados et al., 1998). The duration, frequency and type of such disturbance to coastal seagrass ecosystems are likely to influence the extent of loss and timing of recovery (Short and WyllieEcheverria, 1996). In areas disturbed by dugong grazing, propeller scars and other small-scale disturbance recovery can occur within weeks to months (Williams, 1988; Preen, 1995; Peterken and Conacher, 1997; Rasheed, 1999). Recovery of subtidal seagrass meadows from large-scale disturbance has been shown to take 2e4 years (Preen et al., 1995) or more than 5 years (Bulthuis, 1981; Birch and Birch, 1984; Onuf, 2000; Blake and Ball, 2001). Mapping exercises have documented recovery in areas of seagrass meadows over periods of 5e10 years, such as Posidonia oceania meadows in Cockburn Sound (Kendrick et al., 2002) and Heterozostera tasmanica in Westernport (Blake and Ball, 2001). Despite the known causes of widespread seagrass loss (Preen et al., 1995; Short and Wyllie-Echeverria, 1996) very few studies have documented post-flood recovery rates of intertidal meadows. The paucity of data on recovery of seagrass meadows is often due to the lack of data from long-term seagrass monitoring programs and because many seagrass meadows have either failed to or taken many years to recover following stress from declining water quality (Short and Wyllie-Echeverria, 1996). A recent study has shown rapid re-colonisation of Zostera marina meadows by seedling growth following disturbance from an anoxic event (Plus et al., 2003). Mechanisms likely to influence the recovery of seagrass meadows include the reproductive strategy (sexual or asexual) of a given species and the success of recruitment of seedlings to the disturbed area. Seeds can lie dormant in the sediments for some time and the success of germination of seeds is likely to be influenced by water and sediment conditions (Moore et al., 1993; Conacher et al., 1994). Optimal seed germination of Zostera capricorni has occurred at winter temperatures (16 (C) and reduced salinities (w15&) and in anaerobic sediments high in organic matter (Conacher et al., 1994; Brenchley and Probert, 1998). Although much is known about the factors (eg. light reduction, sedimentation) that cause seagrass decline (Onuf, 1994; Preen et al., 1995; Hall et al., 1999) very little information is known about early stages of recovery, including timing of germination, environmental cues for germination, population growth rates (Olsen and Sand-Jensen, 1994) and the time taken for seagrass meadows to recover to abundances recorded prior to disturbance. The time interval required for seagrass meadows to recover after severe loss and the ability to form meadows in perturbed areas are greatly influenced by seed dispersal and density (Inglis, 2000), germination

rates (Orth et al., 2000) and rhizome elongation (Olsen and Sand-Jensen, 1994; Marba` and Duarte, 1998). In this study we expanded upon an existing monitoring program of seagrass meadows in the Hervey BayeSandy Strait region (Campbell and McKenzie, 2001) following the catastrophic loss of seagrass meadows in February 1999 from flooding. The broadscale spatial recovery rates of intertidal meadows following catastrophic disturbance were assessed by mapping seagrass abundance in the northern Great Sandy Strait region prior to and on 3 occasions after widespread loss of seagrass. Fine-scale spatial assessments of seagrass meadows focussed on two existing Zostera capricorni monitoring meadows in the region. Seagrass recovery and growth were studied at two sites in each meadow. The effect of the flooding on the timing of seed germination during initial recovery stages was also examined together with water quality data, to identify factors (eg. nutrient availability, suspended matter) that may influence seagrass growth and recovery. Morphological characteristics of Z. capricorni were characterised and grouped into population cohorts for estimation of rhizome elongation rates and increases in rhizome internodes during the initial recovery phase of growth.

2. Methods 2.1. Site description The Great Sandy Strait is an estuary between the mainland and Fraser Island and encompasses 93,160 ha (McKenzie et al., 2000) (Fig. 1). The area contains extensive seagrass meadows growing on intertidal sand and mud flats (McKenzie and Campbell, 2003). The Mary River flows into the northern region of the Great Sandy Strait draining a modified catchment of dryland grazing, agricultural crops, cleared land, forestry and both sewered and unsewered urban development (Rayment and Neil, 1997). For an ‘‘average’’ rainfall year, 21% of rainfall is exported as runoff into the Mary River and 268,000 tonnes of eroded sediments flow into nearshore seagrass ecosystems annually. Loads of nitrogen (1.7 kg/ha/y) and phosphorus (0.2 kg/ha/y) are supplied to the Great Sandy Strait passage each year (Schaffelke, 2002). 2.2. Mapping Intertidal seagrass distribution in the northern Great Sandy Strait was assessed using field surveys (boat, diver, foot and helicopter) and aerial photographs in areas exposed at low tide. The field surveys of intertidal seagrasses were conducted on four occasions. On 6e8 December 1998, seagrass distribution was mapped predominately by free-diving and by foot during low tides.

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

479

Fig. 1. The Great Sandy Strait is located on the eastern Australian coast in the southeastern corner of Queensland. The northern Great Sandy Strait is at the mouth of the Mary River between the mainland and Fraser Island.

Ground truthed sites were haphazardly selected across intertidal banks. On 15 April 1999, 23 November 1999 and 27 February 2002, distribution was mapped predominately by helicopter. During the flight, observers

interpreted the distribution of seagrass onto charts to aid interpretation when mapping on the Geographic Information System (GIS). The meadows were ground truthed by foot or while hovering less than a metre

480

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

above the ground, which enabled observers to visually estimate seagrass abundance, seagrass species composition, meadow landscape and sediment categories. Site positions were recorded using a differential Geographic Positioning System (dGPS). The boundary of some meadows was also mapped while hovering in a helicopter directly over the meadow edge and the position fixed using a dGPS. A GIS of the intertidal seagrass meadows was created in MapInfoÒ using the survey information. Errors in GIS maps include those associated with digitising and rectifying basemaps and with Global Positioning System (GPS) fixes for survey sites. Each seagrass meadow was assigned a qualitative mapping value, determined by the data sources and likely accuracy of mapping (McKenzie et al., 2001a). Mapping quality was based on the range of mapping information available for each meadow and associated estimates of reliability (R) in mapping meadow boundaries. Estimates of mapping quality ranged from 7.5 to 100 m. 2.3. Seagrass abundance and habitat characteristics Two Zostera capricorni meadows were monitored every 3 months for seagrass abundance over a 3-year period. The meadows were located at Urangan and Wanggoolba Creek (Fig. 2). Two sites at each meadow (UG1: 25( 18.0526# S, 152( 54.4087# E; UG2: 25( 18.1966# S, 152( 54.3643# E), (WC1: 25( 24.9658# S, 153( 0.3355# E; WC2: 25( 24.8394# S, 153( 59.0383# E) were studied. Seagrass abundance and habitat characteristics were investigated using a rapid assessment technique known as Seagrass-Watch (McKenzie et al., 2001b) developed in tropical Australia. At 2 sites within each meadow (Fig. 2), 3 ! 50 m transects (n ¼ 6) were positioned perpendicular to the shore and located 25 m apart. Visual estimates of percentage seagrass cover, seagrass species composition and sediment type were recorded along each transect from 0.25 m2 quadrats at 5 m intervals (n ¼ 33 per site). Standardised percentage cover photo-indices were used as a guide to reduce observer bias and increase observer consistency in visual estimation. At each quadrat 10 cm of sediment was examined for ‘living’ below-ground rhizomes and roots. Surface sediment type (!3 cm) was measured qualitatively using a visual and touch technique where the presence of 3 components (shell/ gravel (O1000 mm), sand (63 mm to 1000 mm) and mud (!63 mm)) was recorded in order of dominance. 2.4. Morphology and growth At Urangan (UG1) (Fig. 2) seagrass plants with intact rhizome and shoots were randomly harvested and transported to the laboratory on 28 July 2001 (n ¼ 44),

16 August 2001 (n ¼ 37) and 13 October 2001 (n ¼ 33). Rhizome length, rhizome weight and leaf weight were measured for each plant. Each of these measures were grouped into cohorts for each sampling event. For each plant the number of internodes Rhizome Internodes (RI) was counted. The age of individual plants was determined by measuring the number of additional internodes grown between sampling events (28 July to 16 August). This was achieved by determining frequencies of RI cohorts and following their growth over the 19 day study period. For each plant the increase in the number of RI over the 19 day period was used to calculate ages of individual seagrass plants by the following equations: Age ¼ number of rhizome internodes ! plant
1 =RI d
1 RI d
1 ¼ RIt1
RIt0 d
1 where RIt1 Z number of rhizome internodes at second sampling event (16 August) and RIt0 Z number of rhizome internodes at first sampling event (28 July), and d Z number of days between sampling events. Growth rates for each parameter were calculated from the growth of the two youngest population cohorts between 28 July and 16 August, according to the following equation: Growth rate d
1 ¼ Pt1
Pt0 d
1 where Pt1 Z parameter measured for each plant (i.e., rhizome length (mm), rhizome weight (g dry wt.), leaf weight (g dry wt.)) at the second sampling event (16 August) and Pt0 Z parameter measured at the first sampling event (28 July), and d Z number of days between sampling events. Rhizome, shoot and RI growth rates were calculated from differences in mean morphological values at 28 July, 16 August and 13 October to obtain JulyeAugust and AugusteOctober growth rates for each harvested seagrass population, according to the following equation: Growth rate d
1 ¼ Pt1
Pt0 d
1 where Pt1 Z parameter measured for each plant (ie. rhizome length (mm), rhizome weight (g dry wt.), leaf weight (plant
1)) at second sampling event and Pt0 Z parameter measured at first sampling event, and d Z number of days between sampling events. 2.5. Water quality Secchi depth (m) was measured from the boat using a Secchi disc. Temperature ((C), salinity (PSS), and water column concentrations of suspended matter (non-filterable residue) (mg L
1), ammonium, nitrate and organicN (mg L
1) were recorded at each site monthly from July

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

481

Fig. 2. Distribution of northern Great Sandy Strait seagrass meadows pre-flood (December 1998) and post-flood (April 1999, November 1999 and February 2002) with positions (-) of the Zostera capricorni recovery monitoring sites: Urangan (UG1, UG2) and Wanggoolba Creek (WC1, WC2).

1998 to May 2002 using standard protocols (APHA, 1998). Nutrient samples were analysed in a segmented C flow analyser (TechniconÔ) for N as NHC 4 and NO3
1 (mg L ) using standard protocols (APHA, 1998).

Two-way ANOVA was used to test for site (n ¼ 2) and time (n ¼ 5) differences in suspended matter, Secchi depth (m) and nutrient concentrations pooled over the annual sampling periods. Prior to all analyses

482

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

3. Results

complete loss of seagrass (0% cover) occurred at all 4 sites (Fig. 3). There was an absence of below-ground rhizomes and roots. Post-flood observation of Zostera capricorni seagrass seedlings first occurred at Wanggoolba Creek in August 2000, and in May 2001 at Urangan (Fig. 3). Recovery of seagrass meadows to percentage cover values O20% recorded prior to February 1999 did not occur until February 2001 at Wanggoolba Creek and February 2002 at Urangan (Fig. 3).

3.1. Seagrass distribution

3.3. Morphology and growth

Approximately 90% of the seagrass meadows in the northern Great Sandy Strait were Zostera capricorni dominated communities (plants generally of a small morphology with a canopy height !5 cm). The remainder consisted of a mix of Halodule uninervis and Halophila ovalis. Mapping surveys showed that the total area of seagrass meadows in the northern Great Sandy Strait in December 1998 was greater than that shown by postflood surveys conducted in April 1999 and November 1999 (Table 1, Fig. 2). Approximately 90% of intertidal seagrass cover in the northern Great Sandy Strait disappeared after the February 1999 flooding of the Mary River. In February 2002 the total area of seagrass throughout the northern Great Sandy Strait increased to 3712 G 524 ha (Table 1) representing over 100% recovery since the total loss of seagrass meadows in 1999 (Fig. 2).

Mean (GS.E.) morphological characteristics for each sampling event (28 July, 16 August, 13 October 2001) are presented in Table 2. Leaf (above ground) to rhizome (below ground) ratio decreased with age, whereas all other parameters increased over the 19 day period. Significant linear relationships between rhizome length and both number of shoots per plant and plant age (Fig. 4) demonstrate a capacity to predict seagrass growth dynamics from simple measures of rhizome length. Growth rates for rhizome, shoot and RI calculated from differences in mean population values at 28 July, 16 August and 13 October show that growth rates of rhizome length (mm) and rhizome weight (g dry wt.) were higher in AugusteOctober than in JulyeAugust. Growth of shoots and rhizome internodes were lower in AugusteOctober (Table 3).

3.2. Seagrass abundance

3.4. Population dynamics

Meadows of Zostera capricorni occurred on intertidal mud banks at each site chosen for monitoring. Each site consisted of an intertidal bank and was exposed at low tide for an equal period of time (2e3 h) during a daily tidal cycle. Maximum tidal height is the same at each site throughout the year, ranging from 2 to 3 m. On sampling days water depth over the intertidal banks varied from 0.01 m at midday (1200e1400) low-tides to 2.50 m at high tide. In July 1998 the abundance of seagrass ranged from 3 to 38% at Urangan sites and 5 to 80% at Wanggoolba Creek sites. Following the flood in February 1999 a

The frequency of cohorts based on rhizome length on 28 July and 16 August showed two dominant cohorts (i and ii) (Fig. 5). The dominant cohorts could not clearly be identified in October. Rhizome length growth rates (per plant) (between 28 July and 16 August) for the two youngest cohorts were calculated and were higher for the older cohorts compared with the younger cohorts (1.06 and 1.57 mm shoot
1 d
1 for cohorts i and ii, respectively). Six RI cohorts were identified on both 28 July and 16 August sampling events (Fig. 6). The youngest cohort (i), (RI ¼ 1), increased by 3 RI over the 19 day period.

assumptions of homogeneity of variance and independence were tested (Genstat vers 5). Suspended matter data were loge ðxC1Þ transformed and ammonium-N data were square root transformed prior to analysis to meet assumptions of homogeneity of variance. The significance level used was p ! 0:05. LSDs were used to make post-hoc multiple comparisons among treatment means from significant ANOVA tests.

Table 1 Area (ha) (GR) of seagrass meadows mapped in December 1998, April 1999, November 1999 and February 1999 at Urangan, Wanggoolba Creek and the northern Great Sandy Strait Regions

3 months, pre-flood (December, 1998)

2 months, post-flood (April, 1999)

9 months, post-flood (November, 1999)

36 months, post-flood (February, 2002)

Urangan Wanggoolba Creek Northern Great Sandy Strait

91.48 G 9.22 119.5 G 49.9 1995 G 524

0 4.37 G 3.2 913.3 G 190.7

0 0 98.1 G 40.1

92.26 G 12.24 134.7 G 55.6 3712 G 524

483

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

(a)

60

Seagrass cover (%)

50 40

Flood

30 20 10 0

(b)

50

Seagrass cover (%)

40

Flood 30

20

10

Apr-03

Jan-03

Oct-02

Jul-02

Apr-02

Jan-02

Oct-01

Jul-01

Apr-01

Jan-01

Jul-00

Oct-00

Apr-00

Oct-99

Jan-00

Jul-99

Apr-99

Jan-99

Oct-98

Jul-98

0

Fig. 3. Temporal change of total seagrass cover (%) at (a) Urangan (B, UG1; -, UG2) and (b) Wanggoolba Creek (B, WC1; -, WC2) from July 1998 to April 2003. Values represent means (GS.E.) (n ¼ 33) of percentage cover for each month. The time of the Mary River flood in February 1999 is shown.

Cohorts ii to vi increased by 4e5 RI over the 19 days (Fig. 6). Applying an increase of 4 RI over the 19 days (i.e. 0.211 RI d
1), ages of individual seagrass shoots (RI/0.211 d
1) were calculated. The oldest plant found on 28 July had a shoot age of 18 RI and its rhizome length was 109.4 mm. Based on

Table 2 Mean (GS.E.) shoot characteristics for Zostera capricorni from Urangan on 28 July, 16 August and 13 October 2000 Parameter

July 28

August 16

October 13

Rhizome length (mm) 51.79 G 6.06 91.83 G 15.01 313.21 G 12.13 4.20 G 0.89 8.21 G 1.40 60.99 G 4.92 Rhizome weight (mg DW plant
1) Leaf weight 5.09 G 0.53 6.91 G 1.18 53.55 G 6.37 (mg DW plant
1) Leaf/rhizome ratio 2.22 G 0.56 0.97 G 0.09 0.89 G 0.08 No. shoots plant
1 6.83 G 0.63 8.91 G 1.08 13.63 G 0.44 No. leaves plant
1 14.83 G 1.34 19.56 G 2.40 35.00 G 1.24 RI 8.77 G 0.71 13.24 G 1.25 24.27 G 0.97 Plant age (d) 41.55 G 3.35 62.76 G 5.92 115.04 G 4.52

RI the age of the plant is estimated to be 85.3 d (18 RI/ 0.211 RI d
1). Using this estimate a germination date of 5 May is proposed. Alternatively, by using rhizome length and estimates of rhizome growth rates (1.06 and 1.57 mm d
1) to estimate age, an estimate of 70d (18 RI ! 1.06) and 103d (18 RI ! 1.57) was found, respectively. Assuming this is one of the first seedlings to germinate at the Urangan site we estimate that initial germination of seedlings occurred between 17 April and 20 May. As no seedlings were found at the site on 2 May this places initial germination for this plant sometime after 2 May and before 20 May.

3.5. Sediment and water quality parameters Sediments at both Urangan and Wanggoolba Creek sites were predominately mud prior to February 1999 (Fig. 7). Immediately post-flood, Urangan sediments were dominated by sand with a small component of mud. In contrast, Wanggoolba Creek sediments remained

484

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490 250

25 i

(a)

20

150 100 50

y = 0.2908x + 28.858 R2 = 0.7925

Frequency (%)

Age (days)

200

28 July

ii

15 10 5

0 0

100

200

300

400

500

0

600

Rhizome length (mm) 25 25

(b)

20

Frequency (%)

15 10

15

ii

i

10 5

y = 0.0309x + 4.9581 R2 = 0.625

0

0 600

25

Rhizome length (mm)

Table 3 Rhizome growth rates, shoot growth rates and RI expansion rates of Zostera capricorni from Urangan, Hervey Bay (daily growth was calculated from increases in mean parameters from JulyeAugust and from AugusteOctober) Growth parameter

JulyeAugust

AugusteOctober

Rhizome length (mm plant
1 d
1) Rhizome weight (mg DW plant
1 d
1) Shoots (no. shoots plant
1 d
1) Rhizome internodes (RI plant
1 d
1)

2.107

3.817

0.253

0.910

0.109 0.236

0.081 0.190

15 10

450+

350+

400+

300+

250+

200+

0

180+

5

160+

predominately mud with a small sand component. Urangan sediments generally had a higher sand component throughout the study period than Wanggoolba Creek sediments (Fig. 7), however mud was present from February 2000 onwards. No difference in temperature or salinity could be detected between sites (Fig. 8). A decline in salinity was evident during and following the flood in 1999. Twoway ANOVA showed a significant time (year) effect on all water quality parameters measured and a site effect for nitrate-N concentration (Table 4). The time effect was due to higher ammonium-N, nitrate-N and organicN concentrations at Wanggoolba Creek in 1999 compared to other years examined (1998, 2000, 2001, 2002) (Fig. 8). At Urangan, nitrate concentrations in 1999 were higher than in 1998 and ammonium concentrations

20

Frequency (%)

Fig. 4. Relationship between Zostera capricorni rhizome length and (a) age of plant (n ¼ 126), and (b) number of shoots (n ¼ 126).

13 October

140+

500

120+

400

80+

300

100+

200

60+

100

40+

0

20+

5

0+

No. shoots

20

16 August

Rhizome length (mm) Fig. 5. Percentage frequency of Zostera capricorni rhizome lengths collected from Urangan (UG2) on 28 July 2001 (n ¼ 47), 16 August 2001 (n ¼ 45) and 13 October 2001 (n ¼ 33). Length frequency cohorts (i) and (ii) are identified.

were higher in 1999 than in 2000, 2001 and 2002. An interaction effect (site ! year) for nitrate concentration was explained by higher nitrate concentrations at Wanggoolba Creek than at Urangan in 1999 and 2000 but not in other years. Although no site effect was found for other parameters, monthly post-flood concentrations were often lower at Wanggoolba Creek than at Urangan, and mean monthly Secchi depth was generally higher during 2001and 2002 at Wanggoolba Creek than at Urangan (Fig. 8).

4. Discussion The key finding of the present study was that recovery of intertidal sub-tropical seagrass populations to pre-flood abundances can occur within 2 years

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490 50

28 July

Frequency (%)

40 30 20

iv

10 i

ii

v

iii

vi

0 50

16 August

Frequency (%)

40 30 20 10

i

ii

iii

vi iv

v

0

13 October

50

Frequency (%)

40 30 20

0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20+ 25+ 30+ 35+ 40+

10

RI Fig. 6. Percentage frequency of rhizome internodes (RI) from Zostera capricorni plants collected from Urangan (UG2) on 28 July 2001 (n ¼ 47), 16 August 2001 (n ¼ 45), and 13 October 2001 (n ¼ 33). RI cohorts (i to vi) are identified.

following flood related loss. Seedling growth occurred after an 18 month absence of seagrass populations, with full recovery taking between 6 and 9 months after initial seed germination. Re-colonisation of subtidal meadows (O5 m) has also been reported to occur within 2 years of initial loss (Preen et al., 1995; McKenzie et al., 2000). The likely cause of seagrass loss was the loss of light from floodwaters containing high concentrations of sediments and nutrients. The duration and intensity of short-term increases in turbidity is known to inhibit seagrass photosynthesis, carbohydrate concentrations and leaf and rhizome growth (Moore and Wetzel, 2000; Halun et al., 2002; Peralta et al., 2002). The survival period of seagrass below its minimum light requirement is shorter in small species with low carbohydrate storage capacity than in larger species (Longstaff et al., 1999; Peralta et al., 2002). Experimental

485

evidence has shown that low light caused by long duration turbidity events can kill Halophila ovalis after 30 days (Longstaff et al., 1999) and Halodule pinifolia after 100 days (Longstaff and Dennison, 1999). In other studies Zostera noltii was found to survive severe (below 2% SI) reductions in light availability for only 2 weeks (Peralta et al., 2002) and Zostera capricorni survived 30 days of low light (5% SI) with reduced productivity (Grice et al., 1996). During the Mary River flood in February 1999, high turbidity capable of attenuating light lasted for a considerably shorter period than 30 days and resulted in loss of seagrass meadows (McKenzie et al., 2000). This suggests light reduction alone is unlikely to fully explain seagrass loss and other factors associated with the flood (eg. sediment deposition, sediment disturbance, salinity reduction) may also be responsible for the loss of seagrass in the region. At both study sites in the northern Great Sandy Strait the complete loss of seagrass meadows coincided with 2e3 fold increases in turbidity and nutrient concentrations during the 6 months following the flood. Higher concentrations of nitrate were detected at Wanggoolba Creek in 1999, because of its relative proximity to the mouth of the Mary River where loads of sediments and nutrients from floodwaters were deposited. Previous manipulative studies have demonstrated that Zostera capricorni growth, canopy height, shoot density and the proportion of leaf biomass to below-ground biomass can increase with increasing nutrient availability (Udy and Dennison, 1997a,b). High concentrations of inorganic and organic nitrogen can promote seagrass growth and may explain the faster recovery rates of Z. capricorni at Wanggoolba Creek compared with Urangan. Udy and Dennison (1997a,b), however, concluded that growth and morphological characteristics of Z. capricorni were influenced more by light availability and other environmental parameters than nutrient availability. At Wanggoolba Creek relatively high light availability in the years following the 1999 flood and the high availability of organic and inorganic nitrogen suggests that growth was unlikely to be light or nutrient limited. In contrast, germination and growth of Z. capriconi at Urangan is likely to have been both light and nutrient limited. Udy et al. (1999) suggests that in oligotrophic waters an increase in nutrient availability could cause an increase in seagrass and algal biomass. The present study suggests that it is possible that in shallow embayments, where light availability is lower than oligotrophic waters, increased nutrient availability may also enhance seagrass growth. The duration of light limitation is likely to be the critical factor controlling growth responses of seagrasses whereas nutrient availability most likely determines their maximum growth potential (Udy et al., 1999).

486

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

Percentage composition

(a)

Mud/sand Sand/mud

Aug-02

May-02

Feb-02

Nov-01

Aug-01

May-01

Feb-01

Nov-00

Aug-00

May-00

Feb-00

Nov-99

Aug-99

Apr-99

Jul-98

Percentage composition

(b)

Fig. 7. Percentage frequency of occurrence of sand dominated sediments , and mud dominated sediments - at (a) Urangan and (b) Wanggoolba Creek monitoring sites between July 1998 and August 2002.

The re-appearance Zostera capricorni and Halophila ovalis as seedlings following an 18 month absence indicates that seeds of these species can remain viable for a number of years. Short-lived seagrass species (ie. Z. capricorni, H. ovalis, Halodule uninervis) are known to have relatively long-lived seeds (Vermaat et al., 1995) and long dormancy periods may explain the abundance of viable seeds that germinated at both meadows 18 and 27 months post-flood (Wanggoolba Creek and Urangan, respectively). Onset of Z. capricorni seedling growth varied considerably in space and time (different years), occurring between May and August, a finding consistent with previous work (Peterken and Conacher, 1997). Highest seed densities for this species have been found between December and May (Peterken and Conacher, 1997), preceding high numbers of germinating seeds from April to July. At Urangan high germinating seed densities were also observed from April to July (pers. obs.) suggesting that in south-east Queensland the timing of flowering, seed production and seed germination are strongly influenced by seasonal factors. Optimal conditions for Z. capricorni seed germination have been found at winter temperatures (16 (C) (Conacher et al., 1994; Brenchley and Probert, 1998), which is consistent

with germination of seeds at both sites from May to August when water temperatures ranged from 16 to 20 (C. Germination of Z. capricorni seeds has also been shown to favour low salinities (1e10 PSS) (Conacher et al., 1994; Brenchley and Probert, 1998) suggesting that at times low saline floodwaters may promote seed germination. There was no difference in salinity between the meadows studied, suggesting that seed germination and seedling growth were likely to be affected by seasonal forces (temperature, daylength). Between site variation in the onset of seedling growth suggests that seedling growth following seagrass loss may also depend on factors such as physical and chemical characteristics of sediments and water. Differences in sediment characteristics between meadows have been used to explain the timing of re-colonisation of unvegetated sites (Orth et al., 2000). The onset of germination and seedling growth at Wanggoolba Creek in AprileJune 2000 coincided with relatively high nutrient and low suspended sediment concentrations. These factors known to enhance germination and seedling growth of Zostera communities (Moore et al., 1993). The germination of Zostera capricorni seeds is also favoured by anaerobic sediments rich in organic

487

0.50

40

0.45

36

0.10

10

0.08

8

Jul-02

Jan-02

Apr-02

Jul-01

Oct-01

Jan-01

Apr-01

Jul-00

Oct-00

Jan-00

Apr-00

Jul-99

Oct-99

Jan-99

Jul-99

Oct-99

Jan-00

Apr-00

Jul-00

Oct-00

Jan-01

Apr-01

Jul-01

Oct-01

Jan-02

Jul-99

Oct-99

Jan-00

Apr-00

Jul-00

Oct-00

Jan-01

Apr-01

Jul-01

Oct-01

Jul-99

Oct-99

Jan-00

Apr-00

Jul-00

Oct-00

Jan-01

Apr-01

Jul-01

Oct-01

Jul-02

Jan-99

Apr-99

Jul-98

0.30

Apr-99

2 0

Apr-02

Jan-20

Jul-01

Oct-01

Jan-10

Apr-01

Oct-00

Jul-00

Jan-00

Apr-00

Oct-99

Jul-99

Apr-99

Jan-99

0.00

Oct-98

0.02

4

Jan-99

0.04

6

Apr-99

0.06

Apr-02

Secchi depth (m)

12

Jan-99

Jan-02

4 0

Apr-02

Oct-01

Jul-01

Jan-01

Apr-01

Jul-00

Oct-00

Apr-00

Jan-00

Jul-99

Oct-99

Jna-99

Oct-98

8

0.12

Jul-98

Total nitrate-N (mg L-1)

Jul-98

0.00

Apr-99

0.05

12

Apr-99

0.10

16

Jul-98

0.15

20

Oct-98

0.20

24

Jul-98

0.25

28

Oct-98

0.30

32

Jul-98

0.35

Oct-98

0.40

Oct-98

Suspended matter (mg L-1)

Total nitrogen (mg L-1)

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

50 40 Temperature °C

Total organic N (mg L-1)

45 0.25 0.20 0.15 0.10

35 30 25 20 15 10

0.05

Jan-02

Apr-02

Jul-02 Jul-02

45

Jan-02

50

0.07

Apr-02

Apr-02

Jan-02

Jul-01

Oct-01

Jan-01

Apr-01

Oct-00

Jul-00

Jan-00

Apr-00

Oct-99

Jul-99

Apr-99

Jan-99

Jul-98

0

0.08

40 Salinity (PSS ‰)

0.06 0.05 0.04 0.03 0.02

35 30 25 20 15 10

Apr-02

Jan-02

Jul-01

Oct-01

Apr-01

Jan-01

Oct-00

Jul-00

Jan-00

Apr-00

Oct-99

Jul-99

Apr-99

0

Jan-99

5

0

Jul-98

0.01 Oct-98

Total ammonium-N (mg L-1)

Oct-98

5 0.00

Fig. 8. Mean monthly concentrations, of total N, nitrate, organic N, ammonium, suspended matter, Secchi depth, temperature and salinity from July 1998 to July 2002 for sites in the northern Great Sandy Strait adjacent to monitoring sites: Urangan (B, UG1) and Wanggoolba Creek (-, WC2).

matter (Brenchley and Probert, 1998), and the absence of muddy sediments at Urangan for many months following their removal during the 1999 flood may have inhibited seed germination. The replacement of muddy sediments by sandy sediments may also inhibit germination due to the physical movement of sand causing negligible trapping of organic matter and low

nutrient availability (Walker et al., 1999). In turn a low supply of nutrients may limit growth of Z. capricorni (Udy and Dennison, 1997a,b). Although meadow specific nutrient, sediment and microtopographic characteristics may have strongly limited the timing of seed germination, the timing of seed supply from distant source populations may also be

488

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

Table 4 Two-way ANOVA examining effects of site and time on Secchi depth (m), suspended matter (mg L
1), total nitrogen (mg L
1), organic N (mg L
1), ammonium-N (mg L
1) and nitrate-N (mg L
1) (suspended matter data were loge ðxC1Þ transformed and ammonium-N data were square root transformed prior to analysis)

Secchi depth (m) Site Year Site ! Time Error

df

F

P-value

1 4 4 88

3.19 11.95 0.312

0.077 0.001 0.312

1.39 4.49 0.20

0.242 0.002 0.938

1.22 10.29 3.89

0.001 0.048 0.001

0.03 18.41 0.88

0.869 0.001 0.480

Suspended matter (mg L
1) Site 1 Year 4 Site ! Time 4 Error 88 Nitrate-N (mg L
1) Site Year Site ! Time Error

1 4 4 88

Ammonium-N (mg L
1) Site 1 Year 4 Site ! Time 4 Error 88 Organic-N (mg L
1) Site Year Site ! Time Error

1 4 4 88

1.72 5.19 0.89

0.194 0.001 0.474

Total-N (mg L
1) Site Year Site ! Time Error

1 4 4 88

2.15 7.26 1.76

0.147 0.001 0.145

responsible for inter-meadow differences in seedling recolonisation. Seed supply from distant seagrass populations is possible for Halophila ovalis seeds as they are capable of floating. Long distant seed dispersal over many kilometres has been reported for Zostera marina seeds via rafting of reproductive shoots (Harwell and Orth, 2002) but Zostera capricorni seeds can float for only short distances and are unlikely to be dispersed this way. Transport and deposition of seeds by dugong over many kilometres may also allow the supply of seeds from distant sources (ie intertidal or subtidal). Such factors may also explain the differences in the onset of seedling growth between Urangan and Wanggoolba Creek meadows in the northern Great Sandy Strait. The cohort analyses and population growth rates presented here show that rhizome elongation rates increased over time, most likely due to temperature and light increasing from winter to spring. Recruitment of new shoots from seeds increased between May and July but the weight of shoots (ie. leaves) for a given plant

remained relatively stable for the first 20 day growth period, suggesting that carbohydrate reserves are directed to rhizome growth for rapid expansion of the seagrass shoots, consistent with other reports (eg. Alcoverro et al., 1999; Rasheed, 1999; Plus et al., 2003). The second growth phase was characterised by increased shoot production with an increase in leaf surface area suggesting a re-direction of energy reserves to new shoots. The growth of new shoots within meadows is, however, strongly dependent on rates of rhizome elongation and the production of shoots on side branches to occupy available space. This would augment the growth in above-ground canopy and surface area of seagrass shoots beneficial for capturing light and increasing habitat complexity as a function of available space. The strong linear relationships between rhizome length and both plant age and shoot number demonstrate that rhizome length may be used to predict aboveground shoot characteristics. The determination of plant age from rhizome length provides a useful tool to resolve population cohorts and recovery rates for this species following disturbance. The information presented in this paper provides an accurate estimate of the time it takes for seagrasses to recolonise and commence recovery following loss from an acute impact at the region and meadow scale. It is the first information available to suggest that recovery of tropical intertidal seagrass meadows over tens of kilometres can occur within 3 years. Recovery of seagrass populations to pre-flood abundances took between 6 and 9 months after initial seed germination at both Wanggoolba Creek and Urangan. The data also provide new evidence that intertidal seagrass recovery can occur within 2 years following flood related loss. Re-colonisation of subtidal meadows (O5 m) has also been reported to occur within two years of initial loss (Preen et al., 1995; McKenzie et al., 2000). The time interval required to recover after severe loss and the ability to form meadows in perturbed areas are greatly influenced by light quality, although nutrient availability and sediment characteristics are likely to promote seed germination.

Acknowledgements Water quality data was provided courtesy of Queensland Environment Protection Agency and the Queensland Parks and Wildlife Service, collected as part of a state-wide monitoring program. Thank you to Karen Kirk, Juanita Bite´ and Meredith Campey for collecting, sorting and morphometric measurement of field samples. Field surveys were conducted with assistance from the Hervey Bay Dugong and Monitoring Program, Steve Winderlich, Michael Ford, Bill Alston, Mark Brammar, and Kai Yeung (Queensland Parks and

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

Wildlife Service) and Rob Coles, Anthony Roelofs and Scott McKinnon (DPI). Rudi Yoshida (DPI) provided data entry and data management expertise. References Alcoverro, T., Zimmerman, R.C., Kohrs, D.G., Alberte, R.S., 1999. Resource allocation and sucrose mobilization in light-limited eelgrass Zostera marina. Marine Ecology Progress Series 187, 121e131. APHA, 1998. Standard methods for the examination of water and wastewater. In: Eaton, A.D., Clesceri, L.S., Greensberg, A.E. (Eds.), American Public Health Association. United book Press, Baltimore, Maryland, pp. 1e155. Bach, S., Borum, J., Fortes, M.D., Duarte, C.M., 1998. Species composition and plant performance of mixed seagrass beds along a siltation gradient at Cape Bolinao, the Philippines. Marine Ecology Progress Series 174, 247e256. Birch, W.R., Birch, M., 1984. Succession and pattern of tropical intertidal seagrasses in Cockle Bay, Queensland, Australia, a decade of observations. Aquatic Botany 19, 343e367. Blake, S., Ball, D., 2001. Victorian marine habitat database seagrass mapping of Westernport. Marine and Freshwater Resources Institute, Report No. 29, 36 pp. Brenchley, J.I., Probert, R.J., 1998. Seed germination responses to some environmental factors in the seagrass Zostera capricorni from eastern Australia. Aquatic Botany 62, 177e188. Bulthuis, D.A., 1981. Distribution and summer standing crop of seagrasses and macroalgae in Westernport, Victoria. Proceedings of the Royal Society of Victoria 92, 107e112. Cabello-Pasini, A., Lara-Turrent, C., Zimmerman, R.C., 2002. Effect of storms on photosynthesis, carbohydrate content and survival of eelgrass populations from a coastal lagoon and the adjacent open ocean. Aquatic Botany 74, 149e164. Campbell, S.J., McKenzie, L.J., 2001. Community-based monitoring of intertidal seagrass meadows in Hervey Bay and Whitsunday, 1998e2001. Queensland Department of Primary Industries Information Series No. QI01090 (QDPI, Brisbane), 30 pp. Conacher, C.A., Poiner, I.R., Butler, J., Tree, D.J., 1994. Germination, storage and viability testing of seeds of Zostera capricorni Aschers. from a tropical bay in Australia. Aquatic Botany 49, 47e58. Grice, A.M., Loneragan, N.R., Dennison, W.C., 1996. Light intensity and the interactions between physiology, morphology and stable isotope ratios in five species of seagrass. Journal of Experimental Marine Biology and Ecology 195, 91e110. Hall, M.O., Durako, M., Fourqurean, J.W., Zieman, J.C., 1999. Decadal changes in seagrass distribution and abundance in Florida Bay. Estuaries 22, 445e459. Halun, Z., Terrados, J., Borum, J., Kamp-Nielsen, L., Duarte, C.M., Fortes, M.D., 2002. Experimental evaluation of the effects of siltation-derived changes in sediment conditions on the Philippine seagrass Cymodocea rotundata. Journal of Experimental Marine Biology and Ecology 279, 73e87. Harwell, M.C., Orth, R.J., 2002. Long-distance dispersal potential in a marine macrophyte. Ecology 83 (12), 3319e3330. Inglis, G.J., 2000. Variation in the recruitment behaviour of seagrass beds: implications for population dynamics and abundance in Florida Bay. Estuaries 22, 445e459. Kamp-Nielson, L., Vermaat, J., Wesseling, I., Borum, J., GertzHansen, O., 2001. Sediment properties along gradients of siltation in South-east Asia. Estuarine, Coastal and Shelf Science 56, 127e137. Kendrick, G.A., Aylward, M.J., Hegge, B.J., Cambridge, M.L., Hillman, K., Wyllie, A., Lord, D.A., 2002. Changes in seagrass coverage in Cockburn Sound, Western Australia between 1967 and 1999. Aquatic Botany 73, 75e87.

489

Longstaff, B.J., Dennison, W.C., 1999. Seagrass survival during pulsed turbidity events: the effects of light deprivation on the seagrasses Halodule pinifolia and Halophila ovalis. Aquatic Botany 65, 105e121. Longstaff, B.J., Lonerragan, N.R., O’Donohue, M.J., Dennison, W.C., 1999. Effects of light deprivation on the survival and recovery of the seagrass Halophila ovalis (R. Br.) Hook. Journal of Experimental Marine Biology and Ecology 234, 1e27. Marba`, N., Duarte, C.M., 1998. Rhizome elongation and seagrass clonal growth. Marine Ecology Progress Series 174, 269e280. McKenzie, L.J., Roder, C.A., Roelofs, A.J., Lee Long, W.J., 2000. Post-flood monitoring of seagrasses in Hervey Bay and the Great Sandy Strait, 1999. Queensland Department of Primary Industries Information Series No. QI00059 (QDPI, Brisbane), 46 pp. McKenzie, L.J., Finkbeiner, M.A., Kirkman, H., 2001a. Methods for mapping seagrass distribution, (Chapter 5). In: Short, F.T., Coles, R.G. (Eds.), Global Seagrass Research Methods. Elsevier Science B.V., Amsterdam, pp. 101e122, 473 pp. McKenzie, L.J., Campbell, S.J., Roder, C.A., 2001b. Seagrass-Watch: Manual for Mapping & Monitoring Seagrass Resources by Community (citizen) volunteers. Queensland Department of Primary Industries (QDPI, Brisbane), 100 pp. McKenzie, L., Campbell, S., 2003. Seagrass resources of the Booral Wetlands and the Great Sandy Strait: February/March 2002. Queensland Department of Primary Industries Information Series No. QI03016 (QDPI, Brisbane), 28 pp. Moore, K.A., Wetzel, R.L., 2000. Seasonal variation in eelgrass (Zostera marina L.) responses to nutrient enrichment and reduced light availability in experimental ecosystems. Journal of Experimental Marine Biology and Ecology 244, 1e28. Moore, K.A., Orth, R.J., Nowak, J.F., 1993. Environmental regulation of seed germination in Zostera marina L. (eelgrass) in Chesapeake Bay: effects of light, oxygen and sediment burial. Aquatic Botany 45, 79e91. Moore, K.A., Wetzel, R.L., Orth, R.J., 1997. Seasonal pulses of turbidity and their relations to eelgrass (Zostera marina L.) survival in an estuary. Journal of Experimental Marine Biology and Ecology 215, 115e134. Olsen, B., Sand-Jensen, K., 1994. Demography of shallow eelgrass (Zostera marina) populationsdshoot dynamics and biomass development. Journal of Ecology 82, 379e390. Onuf, C.P., 1994. Seagrasses, dredging and light in Laguna Madre, Texas, U.S.A. Estuarine, Coastal and Shelf Science 39, 75e91. Onuf, C.P., 2000. Seagrass responses to and recovery (?) from seven years of brown tide. Pacific Conservation Biology 5, 306e313. Orth, R.J., Fishman, J.R., Harwell, M.C., Marion, S.R., 2000. Seeddensity effects on germination and initial seedling establishment in eelgrass Zostera marina in the Chesapeake Bay region. Marine Ecology Progress Series 250, 71e79. Peralta, G., Pe´rez-Lorens, J.L., Hernandez, I., Vergara, J.J., 2002. Effects of light availability on growth, architecture and nutrient content of the seagrass Zostera noltii Hornem. Journal of Experimental Marine Biology and Ecology 269, 9e26. Peterken, C.J., Conacher, C.A., 1997. Seed germination and recolonisation of Zostera capricorni after grazing by dugongs. Aquatic Botany 59, 333e340. Plus, M., Deslous-Paoli, J., Dagault, F., 2003. Seagrass (Zostera marina L.) bed recolonisation after anoxia-induced full mortality. Aquatic Botany 77, 121e134. Preen, A.R., Lee Long, W.J., Coles, R.G., 1995. Flood and cyclone related loss, and partial recovery, of more than 1000 km2 of seagrasses in Hervey Bay, Queensland, Australia. Aquatic Botany 52, 3e17. Preen, A., 1995. Impacts of dugong foraging on seagrass habitats: observational and experimental evidence for cultivation grazing. Marine Ecology Progress Series 124, 201e213. Poiner, I.R., Walker, D.I., Coles, R.G., 1989. Regional studies d seagrasses of tropical Australia. In: Larkum, A.W.D., McComb,

490

S.J. Campbell, L.J. McKenzie / Estuarine, Coastal and Shelf Science 60 (2004) 477e490

A.J., Shepherd, S.A. (Eds.), Biology of Seagrasses. Elsevier, New York, pp. 279e296. Quammen, M.L., Onuf, C.P., 1993. Laguna Madre: seagrass changes continue decades after salinity reduction. Estuaries 16 (2), 302e310. Rasheed, M.A., 1999. Recovery of experimentally created gaps within a tropical Zostera capricorni (Ashers) seagrass meadow, Queensland Australia. Journal of Experimental Marine Biology and Ecology 235, 183e200. Rayment, G.E., Neil, D., 1997. Sources of material in river discharge. In: The Great Barrier Reef: Science, Use and Management, vol. 1, James Cook University, Townsville, pp. 42e59. Schaffelke, B., 2002. A review of water quality issues influencing the habitat quality in dugong protection areas. Great Barrier Reef Marine Park Authority, Research Publication No. 66, 78 pp. Seddon, S., Connolly, R.M., Edyvane, K.S., 2000. Large-scale seagrass dieback in northern Spencer Gulf, South Australia. Aquatic Botany 66, 297e310. Short, F.T., Wyllie-Echeverria, S., 1996. Natural and human induced disturbances of seagrasses. Environmental Conservation 23, 17e27. Terrados, J., Duarte, C.M., Fortes, M.D., Borum, J., Agawin, N.S.R., Bach, S., Thampanya, U., Kamp-Nielsen, L., Kenworthy, W.J., Geertz-Hansen, O., Vermaat, J., 1998. Changes in community

structure and biomass of seagrass communities along gradients of siltation in SE Asia. Estuarine, Coastal and Shelf Science 46, 757e768. Udy, J.W., Dennison, W.C., 1997a. Growth and physiological responses of three seagrass species to elevated sediment nutrients in Moreton Bay, Australia. Journal of Experimental Marine Biology and Ecology 217, 253e277. Udy, J.W., Dennison, W.C., 1997b. Physiological responses of seagrasses used to identify anthropogenic nutrient inputs. Marine and Freshwater Research 48, 605e614. Udy, J.W., Dennison, W.C., Lee Long, W.J., McKenzie, L.J., 1999. Responses of seagrasses to nutrients in the Great Barrier Reef, Australia. Marine Ecology Progress Series 185, 257e271. Vermaat, J.E., Agwin, N.S.R., Duarte, C.M., Fortes, M.D., Marba, N., Uri, J.S., 1995. Meadow maintenance, growth and productivity of a mixed Philippine seagrass bed. Marine Ecology Progress Series 124, 215e225. Walker, DI, Dennison, WC, Edgar, G, 1999. Status of Australian seagrass research and knowledge. In: Butler, A., Jernakoff, P. (Eds.), Seagrass in Australia: Strategic review and development of an R & D plan. CSIRO, Collingwood, pp. 1e24. Williams, S.L., 1988. Disturbance and recovery of a deep water Caribbean seagrass bed. Marine Ecology Progress Series 42, 63e71.