Movements of the western rock lobster (Panulirus cygnus ... - CiteSeerX

CSIRO PUBLISHING

www.publish.csiro.au/journals/mfr

Marine and Freshwater Research, 2008, 59, 603–613

Movements of the western rock lobster (Panulirus cygnus) within shallow coastal waters using acoustic telemetry L. D. MacArthurA,C , R. C. BabcockB and G. A. HyndesA A Centre

for Marine Ecosystems Research, School of Natural Sciences, Edith Cowan University, 100 Joondalup Drive, Joondalup, WA 6027, Australia. B CSIRO Marine and Atmospheric Research, Private Bag No. 5, Wembley, WA 6913, Australia. C Corresponding author. Email: [email protected]

Abstract. Understanding the residency and movement patterns of major consumers, such as lobsters, in coastal waters is important for the management of coastal habitats and their fisheries. In the present study, we tagged 34 Panulirus cygnus with acoustic transmitters on a shallow coastal reef in south-western Australia and monitored their movements using fixed and manual receivers between November and May 2005–2006 and 2006–2007. We determined the proportion of ‘white’ (migratory-phase) lobsters emigrating from the reef between November and January and also characterised the large-scale movements of ‘red’ (residential-phase) and white lobsters. We undertook tank experiments to determine the effect of tagging and handling on P. cygnus behaviour. Counter to our expectation, 50% of white lobsters were detected on the reef after the migration period, whereas only a small proportion (13.6%) of white lobsters were tracked leaving the reef and only one individual displayed directional offshore movement. This limited movement indicates that coastal no-take zones may build up legal-sized 4–5+-year-old lobsters because many of these are likely to remain resident over the migration season. Laboratory experiments and field observations suggest that tagging and handling affect lobster behaviour and movement for a few days post tagging, potentially confounding conclusions on dispersal and movement patterns in some studies. Additional keywords: acoustic tracking, population dynamics, spiny lobster. Introduction Spiny lobsters are often abundant in coastal ecosystems and can play an important role as a consumer group structuring benthic communities through top-down control (e.g. Pollock et al. 1979), as well as supporting large fisheries (Lipcius and Cobb 1994). Although numerous studies have examined the life history of spiny lobsters (Jernakoff 1990; Acosta and Robertson 2003; Castañeda-Fernández de Lara et al. 2005) and their role as consumers (e.g. Joll and Phillips 1984; BrionesFourzan et al. 2003; Langlois et al. 2005), there is far less understanding of their movement patterns. Yet, knowledge of lobster movement patterns is important for the management of the coastal habitats they utilise and for maintaining sustainable fisheries. Herrnkind (1980) classified lobster movements as either homing, nomadic or migratory. Homing movements represent nocturnal foraging excursions of tens to hundreds of metres over familiar ground, with lobsters returning to known shelters (Herrnkind 1980; Jernakoff 1987; MacDiarmid et al. 1991). Nomadic movements are those that appear to have no predictable direction or periodicity, and may cover several to hundreds of kilometres (Hernnkind 1980). In contrast, migratory movements are directional movements en masse during a confined period of time, such as the movement of sub-adult lobsters towards spawning grounds (e.g. Phillips 1983; Moore and MacFarlane 1984; Groeneveld and Branch 2002), seasonal return movements of © CSIRO 2008

lobsters towards feeding grounds (e.g. Kelly et al. 1999; Kelly 2001) or return movements of gravid females to specific areas for larvae release (Herrnkind 1980). There can be considerable variation within species in the proportion of lobsters undertaking nomadic or migratory movements and past researchers have identified ‘resident’ and ‘transient’ behaviour within single populations (Childress and Jury 2006). In addition, there may be considerable variation within species in the spatial extent, direction and timing of the movements displayed (Herrnkind 1980; Groeneveld and Branch 2002; Linnane et al. 2005). Measuring large-scale lobster movements has typically involved mark–recapture studies, with commercial fishers providing tag returns (e.g. Phillips 1983; Groeneveld and Branch 2002). However, there are several disadvantages with this method, including spatial and temporal heterogeneity of fishing effort introducing bias to the results and the varying effectiveness of baited pots over time and between individual lobsters (Herrnkind 1980). Acoustic telemetry offers several advantages over mark–recapture studies in that frequent position estimates of tagged lobsters can be made without disturbing the animals and the movement patterns observed are independent of fishing effort (e.g. Watson et al. 1999; Kelly 2001). The use of passive arrays of fixed receivers also allows for regular position estimates to be made over time and space, covering potentially large areas (see Heupel et al. 2006). Acoustic tracking, like all tagging techniques, requires the capture, handling and tagging of 10.1071/MF07239

1323-1650/08/070603

604

Marine and Freshwater Research

the organism in question, and this may alter movement patterns, at least in the short-term (e.g. Atkinson et al. 2005). The western rock lobster, Panulirus cygnus, occurs along the lower west coast of Australia (Chittleborough 1970) and forms a major fishery in this region (de Lestang and Melville-Smith 2006). The distribution of P. cygnus is spatially segregated; 1– 5-year-old juveniles are most common on shallow reefs within coastal waters 6 years) generally dominate deeper (30–150 m), offshore waters (Chittleborough 1970; Chittleborough and Phillips 1975). This size distribution is consistent with the results of previous P. cygnus tagging studies that have identified an offshore movement of maturing 4–5-year-old lobsters from November to January, based on returns by commercial fishers (George 1958; Phillips 1983). These migrating individuals are characterised by a relative lack of pigmentation after moulting in October–November and are commonly referred to as ‘whites’ as opposed to the rest of the individuals in the population, which are referred to as ‘reds’ (George 1958). After moulting, the white colouration gradually changes to the typical red colour over the subsequent 2 months (George 1958). Whites, which are generally over 60 mm carapace length (CL) (George 1958), with many over the minimum legal size of 77 mm CL, form up to one-third of the total annual catch (Chubb and Barker 2005; de Lestang and Melville-Smith 2006). In contrast to whites, reds are believed to be mainly residential on shallow coastal reefs (Chittleborough 1970, 1974), although tagging studies have shown them to move between reef patches up to ∼100 m apart (Jernakoff 1987; Ford et al. 1988; Phillips 1990). As whites form a significant component of the commercial catch (Chubb and Barker 2005), knowledge of the proportion of whites emigrating from inshore reefs is important for the management of the fishery and for the spatial management of coastal habitats. Currently there is an assumption that the majority of 4–5-year-old white P. cygnus undertake this migration, although this may, in part, be confounded by the effect of fishing in removing whites from coastal reefs. In the present study we used acoustic telemetry to test the hypothesis that the majority of white P. cygnus emigrate from an inshore reef over the migration season. In addition, we characterised and compared the large-scale movements of lobsters to test the hypothesis that whites make larger movements than reds on our test reef. To test the hypothesis that tagging affects behaviour in the short-term and to help interpret the observed movement patterns in the field, we conducted tank experiments to compare the den occupancy and mussel consumption of tagged and control lobsters. Materials and methods Laboratory tagging experiments As handling and tagging have been shown to influence the movement and behaviour of spiny lobsters in the field (e.g. Jernakoff et al. 1987; MacDiarmid et al. 1991; Atkinson et al. 2005), we undertook tagging experiments in August 2006 to determine if the process of tagging affects the behaviour of Panulirus cygnus, and thus whether it has the potential to confound the results obtained in our field study. We examined food consumption (the number of mussels consumed per night) and den occupancy (the proportion of time spent outside shelter) to test whether tagging altered the behaviour of this species. These behaviours

L. D. MacArthur et al.

were chosen because experiments examining these behaviours could be conducted in relatively small enclosures, and although not measuring natural movements per se these experiments did measure aspects of natural behaviour related to movement. Lobsters (76–80 mm CL) were obtained from commercial fishers and allowed to acclimatise for a period of 2 weeks in holding tanks. Over this time they were fed Mytulis edulis mussels daily ad libitum. Six lobster pairs were then randomly allocated as the tagging treatment and six pairs remained untagged (serving as the control). Each pair was placed into one-half of a tank (100 cm diameter by 60 cm deep), divided with a mesh partition. Each half tank contained an identical plastic shelter, large enough to house both lobsters. Pairs rather than single lobsters were used because P. cygnus is gregarious at this size and solitary individuals have been found to exhibit abnormal behaviour, that is, reduced moult increments when held in captivity (Chittleborough 1975). Daily mussel consumption (the number of mussels consumed per day) was measured for 7 days before tagging. On the eighth day the lobsters allocated to the tagging treatment were tagged with ‘dummy’ tags that were the same size and weight as Vemco V13H (Vemco, Halifax, Nova Scotia, Canada) acoustic transmitters, that is, 36 mm × 13 mm in diameter and weighing 6 g in water. Dummy tags were attached to the dorsal surface of the carapace using two cable ties fastened on either side of the third pair of walking legs, a technique successfully used by Kelly (2001) on Jasus edwardsii. The entire tagging procedure took less than 2 min per lobster. Lobsters were returned to their half tanks and mussel consumption was measured for a further 14 days. To measure the proportion of time a lobster pair spent outside their shelter, two video cameras were set up above the three tanks to give a ‘birds-eye’ view of the shelter and the surrounding area. With the aid of red fluorescent lights that emitted wavelengths >575 nm (∼620 nm peak), which is well above the maximum sensitivity of the eye receptors in the congeneric Panulirus argus (Cummin et al. 1984), movements were recorded for a 10-h period onto long-play video cassettes from 1800 hours, that is, around sunset at the time of the experiment. At each 30-min interval point throughout the video recordings the number of lobsters occurring completely outside the shelter was tallied, giving a value between 0 and 40 sightings. This movement behaviour was recorded over three nights before tagging and for a further 6 days following tagging. A repeated-measures ANOVA (Statistica; StatSoft, Tulsa, OK, USA) was used to determine if tagging had an effect on behaviour and if there was an interaction between the factors of tagging and time. In particular, we were interested in determining if there was a significant change in mussel consumption or den residency immediately after tagging and for how long these differences persisted, that is, if there was a significant interaction between time and treatment. Before analysis, the data were shown to be normal using a Kolmogorov–Smirnov test. Levene’s test was used to test for equality of variance between the tagged and untagged lobster data on each day of the experiment. The consumption data displayed equal variances on 20 of the 21 days, while the den residency data displayed equal variances on eight of the nine experimental days. As unequal variance occurred only once for each data set, Levene’s test is sensitive to

Spiny lobster movement


(a)

605

(b) N

1 km

North West Cape

1

Western Australia

Tagging site 2

Jurien Bay

3

Cape Naturaliste 4

6

5

Jurien Bay

Sanctuary Zone

Tagging site

8

9

7 N

1 km (c)

N

1 km

16

30°20S

17 1

2

18

Tagging site

3

4

19

5 6

8

7

30°21S 10

9

11

12 15

Receiver with range circle

13

14 20

Reef edge 115°00E

115°01E

115°02E

Fig. 1. Map of the study site at Jurien Bay, Western Australia, for acoustic tracking of Panulirus cygnus showing (a) the position of the study site in relation to the Boullanger Island Sanctuary Zone, (b) the position of the array between November 2005 and May 2006 and (c) the position of the array between November 2006 and May 2007.

unequal variance (Quinn and Keough 2002), and transformations did not improve the data, we used untransformed data for the analyses. For both the consumption and den residency data, Mauchly’s test of sphericity was violated so a Huynh–Feldt correction was applied to reduce the degrees of freedom when determining the P-values. Post-hoc Tukey’s HSD comparisons were carried out between tagged and untagged lobster data to determine on which days any significant differences existed.

Field site The field study was conducted in the Boullanger Island Sanctuary Zone (30◦ 18.6594 E, 30◦ 21.1740 S) within the Jurien Bay Marine Park, on the mid-west coast ofAustralia (Fig. 1a). Fishing for P. cygnus is prohibited in this zone. Conducting the study in this zone allowed the residency and movements to be studied without the potential for lobsters to be caught by, or their movements influenced by the presence of, baited pots. Within the zone

606


there are two main areas of reef; an area of inshore reef within 600 m of the shore at the south-eastern corner of the zone and a line of offshore reef running roughly parallel with the shore and marking the western boundary of the sanctuary zone, 3–4 km from shore (Fig. 1c). As an accurate map of the inshore reef system was not available, video tows were conducted over the area to identify and map the location of the reef edge and the nature of the surrounding benthic habitat. A video was recorded with a GPS overlay and the benthic habitat was later determined during playback. In addition, a high-relief reef edge could be identified from the surface in favourable wind conditions and was marked by GPS. This information was used to construct a map of the area using mapping software (OziExplorer; D & L Software, Brisbane). SCUBA surveys of the inshore reef revealed a complex topography, with reef patches of varying sizes with vertical faces of up to 3 m from the sea floor and undercut ledges at the base of the reef that were occupied by P. cygnus. The habitat around the reef consisted mainly of Amphibolis spp. seagrass meadows at depths of 3–7 m. Posidonia spp. and patchy Halophila ovalis seagrass meadows together with areas of bare sand were present further offshore at depths of up to 10 m. This area of inshore reef was typical for the marine park and is representative of habitats of the mid-west coast of Australia and is in the centre of the western rock lobster’s distribution (CALM 2005). Tagging and acoustic monitoring A total of 34 individual P. cygnus were tagged between December 2005 and March 2006 and between November and December 2006 (Fig. 2). All but two lobsters (51 and 57 mm CL) were between 60 and 85 mm CL. The lobsters were caught using hand snares and were brought to the boat for tagging. Lobsters were tagged in the same manner as in the laboratory with eitherV9H or V13H coded tags emitting on 69 kHz (Vemco). Carapace colour (i.e. red or white), sex and CL were recorded. Red lobsters had a section of one of their pleopods removed, which was used to determine moult stage in the laboratory (Dall and Barclay 1977). White lobsters were assumed to have recently moulted and to be in intermoult condition because only hard-shelled lobsters were tagged. The tagging process took less than 2 min and the lobsters were then allowed to recover in fresh, shaded seawater for a period of 5–30 min. Lobsters were returned to the water within 50 m of where they were captured. The majority of lobsters in the present study (60–80 mm CL) would be expected to undergo two moults per year, one before mid November and another between February and April (Melville-Smith et al. 1997). Lobsters tagged in November and December should not have lost tags as a result of moulting until at least February. The lobsters tagged on 8 March 2006 may have already moulted or may have been due to moult in the following months. It was, however, confirmed that these lobsters were not in pre-moult condition. The premoult condition lasts ∼30 days for 2–3-year-old lobsters (Dall and Barclay 1977), and longer for lobsters of the size tagged in the present study. This suggests that the lobsters were likely to retain their tags at least until 10 April 2006. An array of nine fixed receivers (VR2; Vemco) was deployed around the inshore reef system during November 2005 (Fig. 1b). These receivers were placed between 400 and 700 m apart to cover an area of ∼3.46 km2 . Field trials to test the range of the


receivers revealed a maximum range of 350 m for receivers over a flat bottom and at similar depths to the array location (i.e. 3– 10 m). However, the acoustic reception range of the receivers is a dynamic variable that depends on several environmental conditions, such as sea state, turbidity and bottom structure (Simpfendorfer et al. 2002; Heupel et al. 2006). The receivers were assumed to have adequate coverage to detect lobsters in open habitat away from reef, however, they were not expected to continuously detect lobsters while they were sheltering under high-relief reef, which can create a barrier to acoustic signals. In an attempt to gather more information on the presence/ absence of lobsters that may not have been detected by the fixed receivers, we used a manual receiver (VR100; Vemco) with a directional hydrophone (VH100; Vemco) during the nights of 11, 12, 15 and 22 February 2006 and on 3 and 7 March 2006. Lobster position estimates were marked by GPS. As the tags were coded and, therefore, emitted signals relatively infrequently (i.e. between 15 and 180 s), it was difficult to use the directional capability of the hydrophone to pinpoint the exact position of the lobsters. However, preliminary tests using the manual receiver revealed that tag signals around the high-relief reef could only be reliably detected within 100 m of the transmitter (L. D. MacArthur, unpubl. data) and, thus, as a conservative estimate, these positions were likely to be accurate to within 100 m. Data captured by the fixed receivers were downloaded in January and June 2006 after which the receivers were retrieved. In early November 2006, an array of 20 VR3 receivers (Vemco) was deployed around the same inshore test reef and extended to include coverage of a section of the offshore reef system encompassing a total area of 7.31 km2 (Fig. 1c). These new receivers allowed lobster movements to be tracked over a larger area, had a similar range to the VR2 receivers and provided better memory storage and data retrieval. Data captured by this array were downloaded in June and July 2007. Additional information on lobster presence/absence was recorded using the manual receiver and directional hydrophone on 12 March 2007. Data analysis Presence/absence data from the fixed receivers was used to determine the percentage of red and white lobsters detected by multiple (>1) receivers. A Fisher’s exact test was used to test for significant differences in the proportion of red v. white lobsters observed. A time-line was constructed for each lobster showing the number of receivers that had registered a transmission from that lobster per week, and whether each lobster had been detected by the manual receiver (Fig. 2). Using both fixed and manual receiver data we determined the percentage of white lobsters detected on the inshore reef system after the migration season, that is, from February onwards. When more than one receiver had detected a lobster within a short time interval (1–3 h) or bin, a mean position algorithm from Simpfendorfer et al. (2002) was used to estimate the average position, or activity centre, of the lobster in that bin. This was done by calculating the mean of the logged receiver’s latitude and longitude, weighted by the number of transmissions each receiver detects in each bin (Simpfendorfer et al. 2002). Mean positions were plotted onto a map of the study site using a GPS mapping program (OziExplorer; D & L Software) and presented visually to indicate the direction and size of the movement.

M M M F F F F F F F M F F F M M M M M F F M M M M

M M F F F M F M M

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

26 27 28 29 30 31 32 33 34

76 77 74 75 79 84 78 75 75

57 62 51 72 77 67 75 76 75 80 75 63 74 74 65 76 70 81 72 70 71 81 76 74 73

CL

White White White White White White White White White

Red Red Red Red White White White White White White White White White White White White White Red Red Red Red Red Red Red Red

Colour

1 1 2 2 1 1 4 2 1

5 1 0 1 8 6 6 1 2 0 0 1 0 0 1 0 0 1 1 3 1 1 1 1 2 November 2006 1 1 1 1 1 1

# fixed receivers November logging lobster 2005

1 1 1 1 1 1

2 x x 1 x x 1 x x x x 1 x x 1 x x

x x 1 x x 1 x x

x x x x x 1 x x

1 1 2 1

1 2 2 1

1 1 2 1 1 x 1

x

1 1

x x

x

x 1 1 1 x 1 x x

x

x

x

x x x x x x x x x x x x x x x x x 1 1 x 1 x 1 x x x x x x x 1 February

x x x x x x x x

1 1 1 1 x x x x x x 2 1 1 1 1 1 1 1 x 2 1 1 1 1 1 x x

January 2007

2 x x 1 x x 1 x x

x 1

x x x 1 x x 1 x x

x

1 x x 1 x x 1 x x

x

x

x

1 1 x x x

x x

February

1 x x x x

January 2006

x x 1 1 1 1 1 1 x x x x x x x x x x 1 1 1 1 1 1 1 x x

December

5 1 x 1 8 6 6 1 1 x x 1 x x x x x

December

x

x

x

x

1 x 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 x x 1 2 1 1 March

x

March

1 3 1 1 1 x 2

x

x

1 1 1 1 1 1 x x 1 1 April

1

x

April

1 1 1 1 1

x

1 1 1 x May

1

1 1 1 1 1 1 1 1 1 1 1

1

May

Fig. 2. Details of tagged Panulirus cygnus with the detection time-line aggregated into months and weeks. Note: the last ‘week’ of each month contains 7–10 days for convenience. Grey boxes indicate one or more fixed receiver receptions in that week. Numbers inside the boxes indicate the number of receivers hit that week. Black boxes indicate a manual receiver reception that week. ‘×’ indicates no receptions that week, but receptions to follow. CL, carapace length (mm).

Sex

Lobster #

Spiny lobster movement Marine and Freshwater Research 607

608



Table 1. Repeated-measures ANOVA showing the before and after impact of tagging Panulirus cygnus on mussel consumption and den occupancy over the experimental period in which three lobster pairs were tagged and three pairs remained untagged at a designated time in the experiment P < 0.05 highlighted in bold

N 1 km Receiver Actual tag track Predicted tag track

Test variable

Factor

Mussel consumption Tagging Residual (tagging) Time Tagging × time Residual (time)

MS

d.f.

F

P

67.18 1 2.48 0.190 27.06 4 9.53 11.98A 2.15 0.031 0.107 11.98A 2.24 0.024 0.022 47.92A

Reef Den occupancy

A Huynh–Feldt

Fig. 3. Map showing the actual path of a tag towed through the study site for 1 h and the path of that tag predicted by the mean position algorithm using fixed receiver data analysed in 5-min bins.

The accuracy of this method was tested in field trials by towing a tag through the array near the seabed at a speed of 2.7 km h−1 for 1 h while logging the boat’s position using GPS. The boat’s position, averaged over 5-min bins, was then compared with the position estimates for those bins using data logged on the receivers and the mean position algorithm. For the 12 bins, the accuracy ranged from 62 to 370 m and averaged 193.8 ± 30.8 m (Fig. 3). The summed straight-line distance between 5-min positions was 2.6 km for the mean position estimates and 2.7 km for the actual tag path. The path predicted by the mean position equation followed the same general direction as the actual tag path (Fig. 3). Results Laboratory experiment There was a significant interaction between tagging and time on the number of mussels consumed by lobsters per night, but not for the den occupancy of lobsters (Table 1; Fig. 4). A post-hoc Tukey’s HSD test revealed that all pair-wise differences between mussel consumption of tagged and untagged lobsters occurred between 0 and 3 days after tagging (P < 0.05). Therefore, tagged lobsters exhibited significantly decreased consumption of mussels compared with untagged lobsters over the 4-day period immediately after tagging, but increased consumption to levels similar to the untagged lobsters thereafter. Detection of lobsters by fixed and manual receivers The proportion of all lobsters that were detected by one fixed receiver or were not detected by any fixed receivers was greater than the proportion of lobsters detected by two or more receivers (67.6 v. 32.4%; Fig. 5). The proportion of red and white lobsters detected by multiple receivers (25.0 and 36.4% respectively) did not differ significantly (P = 0.705, two-tailed Fisher’s exact test). Lobsters detected by only one receiver were always detected by

Tagging 118.52 Residual (tagging) 72.76 Time 53.88 Tagging × time 21.24 Residual (time) 26.55

1 1.63 0.271 4 3.63A 2.03 0.147 3.63A 0.80 0.534 14.53A

reduced degrees of freedom.

the receiver closest to the tag site, that is, receiver 3 in 2005– 2006 or receiver 5 in 2006–2007. Those lobsters that were not detected by any fixed receivers were all detected by the manual receiver during February 2006 (Fig. 2) and were located within ∼300 m radius of the tagging site. Patterns of movement All red lobsters were found to be in intermoult phase. Three red lobsters (lobsters 1, 20 and 25) made detectable movements within the array and two of these lobsters (lobsters 1 and 20) moved out of the array (Fig. 6). On 30 March 2006, 52 days after tagging, lobster 20 (70 mm CL ) was detected to have moved south and to the seaward edge of the reef. After this period of time this lobster moved several hundred metres further offshore away from the inshore reef. No further detections within the array were logged for this lobster. In contrast, lobster 1 (57 mm CL ) moved south out of the array immediately after tagging between 1400 and 1900 hours on 4 December 2006 and appeared to move along the seaward edge of the inshore reef habitat. After 1900 hours on this day the lobster was not detected again. On several occasions during March 2006, lobster 25 (73 mm CL ) was detected by receivers 3 and 5, which were ∼550 m apart (Figs 2 and 6). These movements between receivers were detected at night and represent a movement south and offshore from the original tag site. Eight white lobsters (lobsters 5, 6, 7, 9, 28, 29, 33 and 32) were detected by two or more receivers, which allowed their movement paths to be plotted. Lobsters 5 (77 mm CL ), 6 (67 mm CL ) and 7 (75 mm CL ) all moved away from the tagging site immediately after tagging on 2 December 2005 (Fig. 6); however, they did not immediately move out of range of the array. All lobsters showed complex movement patterns, travelling through offshore seagrass meadows for ∼48 h and frequently backtracking over routes previously travelled (Fig. 6). Position changes took place during both day and night and the paths predicted by the mean position estimates ranged between ∼4.5 and ∼7 km in length (Fig. 6). It is possible that some of these movements took the lobsters outside the range of the array and there were times,



100

(a) 10

0–1 Receivers 2 Receivers

Tagged

80

Control

8

Proportion of lobster (%)

Mean no. mussels consumed

609

6

4

2

60

40

20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

0

Experiment day number

(b)

All lobsters n 34

20

Mean no. sightings outside den

Tagged Control 15

10

5

0 1

2

3

4

5

6

7

8

9

Experiment day number Fig. 4. Influence of tagging on Panulirus cygnus in a before/after impact experiment on (a) nightly mussel consumption (the mean number of mussels consumed ± 1 s.e.) over 21 days and (b) den occupancy (the mean number of sightings outside shelter ± 1 s.e.) over 9 days. Three lobster pairs were tagged and three pairs remained untagged. The day of tagging is indicated by a dashed line.

usually during the night, when no position estimates could be made over the 3-h bins. This, together with the fact that the paths between mean position estimates are simplified straight lines, suggests that these calculated distances are underestimates. The speed of lobsters over these calculated distances ranged between 94 and 146 m h−1 . Lobster 5 was last detected near receiver 3 on 4 December 2005 (Fig. 6), whereas lobsters 6 and 7 were last detected near receiver 8 on 22 December 2005 and 30 April 2006, respectively, and may have moved out of the array at its southern edge at these times (Fig. 6). Lobster 9 (75 mm CL ) moved between receivers 3 and 5, which were spaced ∼550 m apart, during December 2005 and January 2006 (Figs 2 and 6). Lobsters 28 (74 mm CL ), 29 (75 mm CL ) and 33 (75 mm CL ) were detected by receivers 5 and 7 (∼600 m apart) at various stages during December 2006 and January 2007 (Figs 2 and 6). These receivers covered the

‘Reds’ n 12

‘Whites’ n 22

Fig. 5. Proportion of all Panulirus cygnus, ‘reds’ and ‘whites’, logged by 0–1 or 2+ receivers during the tracking study.

northern and the north-western edge of the reef, indicating that these lobsters moved south and offshore from the original tag sites. Lobster 32 (78 mm CL ) displayed a substantial directional movement between the inshore and offshore reef systems between 13 and 18 December 2006; a straight-line distance between the receivers of 3.37 km (Fig. 6). This represents a speed of ∼560 m day−1 . Between 13 and 17 December 2006, this lobster was detected by receivers 3 and 4, west of the tagging site in a predominantly Posidonia sinuosa seagrass meadow. Between 17 and 18 December, lobster 32 moved between receiver 3 and receiver 18 over an area dominated by a large sand-flat. The four receivers that logged lobster 32 formed a line that ran approximately east–west and thus the movement of this lobster followed a consistent westerly direction (Fig. 6). After 18 December no further receptions from this lobster were logged, suggesting movement further offshore beyond the range of the receiver array. Residency and migration of white lobster Of the 22 white lobsters tagged, 11 (50%) were detected after the proposed November–January migration period (Fig. 2). Of the other 11 white lobsters that may have potentially moved off the test reef, three were found to have moved from the tag site to the edge of the array (lobsters 6, 7 and 32; Fig. 6). Thus, three of the 22 white lobsters (13%) showed evidence of movement away from the inshore reef. Discussion Migration of white P. cygnus from inshore reefs The present study indicates that an offshore migration by the majority of 4–5-year-old white P. cygnus during November– January does not hold true for all inshore reefs. It is possible that a high proportion, perhaps the majority, of sub-adult, maturing lobsters on some inshore reefs remain resident over this period.

610


N

N

Lobster 20

1

2


2

3

4

6

5

Lobster 7

2

3

4

6

5

6 5

8 9

N

1 km

7

1

8

9

1 km

7

1

3

4

8

N

Lobster 1

1

N

Lobster 5

1 km

7

1

N

Lobster 6

9

1

Lobsters 28, 29 and 33

2

3 2

3

4

4

2

6

4

6 5

5

7

8

9

8 1 km N

1

2

10

8

11

9

12

1 km

N

Lobsters 25 and 9

3

7

6

1 km

7

5

3

Lobster 32

16

17

1 2

18 4

6 5

3 4

19 8 7

1 km

5

9 1 km

6

7

8

Fig. 6. Movement patterns of tagged Panulirus cygnus constructed by joining the mean position estimates (solid lines) or receivers detecting lobsters (dashed lines). Three-hour bins were used for lobsters 5, 6, 7 and 20 and 1-h bins were used for lobster 1.

Of the 22 whites (63–84 mm CL) that were tagged over 2 years in late November or early December, 11 (50%) of these were detected on the inshore reef system during February, after the migration period. Furthermore, only three lobsters (13.6%) were detected passing the outer receivers of the inshore array during the migration period, and only one (4.5%) of the tagged whites displayed a clear offshore movement typically associated with migration (Phillips 1983). Thus, the proportion of whites migrating offshore from our study reef lies between a minimum of

4.5% and a maximum of 50%, the greater part of this discrepancy relating to those lobsters that were not detected after the migration period. We suggest that undetected lobsters were more likely to have been present on the reef, but in acoustic shadows created by the reef, than to have emigrated undetected from the reef system and, therefore, the actual proportion that migrated from the reef system is likely to be closer to the stated minimum. Phillips (1983) used a mark and recapture study to investigate the number and movement patterns of lobsters leaving



611

an inshore reef over the migration period. Of the 878 lobsters returned with complete data in that study, as many as 354 (40.3%) were recaptured away from the tag site during the fishing season. As the lobsters were tagged before moulting into a white or red, the results cannot be directly compared with our estimates of the proportion of whites migrating. However, assuming that these lobsters were whites, the results from our study suggest that a smaller proportion of lobsters migrated. The identification of resident and transient behaviour by white lobsters in the present study supports the results reported in previous studies. Panulirus argus has been shown to exhibit resident and transient behaviour (Herrnkind 1980; Cox and Hunt 2005), as has the clawed lobster Homarus americanus (Bowlby et al. 2007). A similar behavioural dichotomy has been observed in fish populations (Egli and Babcock 2004). As a significant proportion of white P. cygnus are of legal size (77 mm CL) and form up to one-third of the fishery landings (Chubb and Barker 2005), this has important consequences for the management of the fishery and for the management of habitats in shallow coastal waters. Our study suggests that a high proportion of lobsters is likely to remain resident on coastal reefs in the absence of fishing pressure, at least over the summer–autumn period covered in the present study. Therefore, coastal no-take zones have the potential to increase the number of legal-sized lobsters on shallow coastal reefs. Babcock et al. (2007) demonstrated that a shallow-water lobster no-take zone at Rottnest Island, south and offshore of Jurien Bay, had significantly greater numbers of large, mature P. cygnus than adjacent fished areas. Whether coastal no-take zones could similarly build up numbers of resident mature adults is difficult to predict, however, anecdotal evidence suggests that large, mature P. cygnus (>6 years old) were historically a conspicuous component of inshore reefs, but are now uncommon (Ottaway et al. 1987). Further monitoring of current coastal notake zones along the west coast of Australia will be required to test the hypothesis that some P. cygnus will remain as long-term permanent residents on shallow coastal reefs. The present study indicates that acoustic arrays can be a useful tool for monitoring the residency of lobsters and other marine organisms inhabiting discrete structured habitats. Although highly structured habitats, such as reefs, block acoustic signals, fixed receivers placed around such habitats can be used to quantify the number and timing of animals emigrating. The ability of an acoustic array to consistently monitor movement over time and space also reduces the chance of missing any movements that occur over confined periods of time. In contrast to acoustic tagging, only a small proportion of tagged lobsters is recaptured using traditional mark–recapture techniques and these recaptures are influenced both by the spatial and temporal dynamics of the fishing fleet and the effectiveness of the fishing gear (Herrnkind 1980).

reports of the residency of reds and the migratory nature of whites (George 1958; Chittleborough 1970; Phillips 1983), and indicates that colour alone is not necessarily a good indicator of the likelihood of a lobster moving large distances. Our study indicated that regardless of colour, most of the 4–5-year-old lobsters tagged were likely to reside close to the tag site over the summer–autumn period. The offshore movement of a 78-mm CL white lobster is consistent with the offshore migration of whites studied by Phillips (1983). This lobster moved 3.37 km between the inshore and the more offshore reef between 13 and 18 December in a westerly direction. This represents a straight line rate of movement of 560 m day−1 , which falls within the range of 359–622 m day−1 calculated by Phillips (1983) for similar-sized lobsters migrating offshore. These estimates are likely to be underestimates of true speed because lobsters are unlikely to be constantly moving in a straight line. Six lobsters, both red and white between 70 and 75 mm CL, were detected to have moved towards the outer edge of the test reef between January and March. One of these lobsters, a 70-mm CL red, then left the array in an offshore direction in March. These detections occurred at night and may represent movements made during nocturnal foraging excursions. The common offshore component of these movements is similar to that observed for P. cygnus moving between shallow patch reefs (Ford et al. 1988; Phillips 1990), and to the gradual offshore dispersal of P. argus from settlement habitats (Herrnkind 1980). This type of gradual or stepwise movement of P. cygnus could account for the reported distribution of small juveniles at water depths up to 10 m and larger juveniles at depths up to 20 m (Chittleborough 1970; Chittleborough and Phillips 1975), and may represent an important mechanism for the dispersal of P. cygnus in shallow waters. The seemingly nomadic movements of three whites, 67– 77 mm CL, covering between 4.5 and 7 km in 48 h, were probably caused by the act of handling and/or tagging. These lobsters moved over primarily Posidonia spp. habitat at approximate speeds of 94–146 m h−1 , quicker than the speeds recorded for P. cygnus foraging over seagrass and sand (∼60 m h−1 ; Jernakoff 1987). Together with the fact that movements occurred during the day, this suggests that the observed movements were not related to foraging. Tank experiments also point to these movements being initiated by the tagging disturbance (see below). Interestingly, the whites were observed to backtrack over previous routes, suggesting that navigational cues were being used at quite large scales, as has been shown for other spiny lobsters (Herrnkind 1980; Nevitt et al. 1995; Boles and Lohmann 2003). Similar to the movements of these whites, the movement of a red lobster out of the array immediately after tagging was also likely to be triggered by the disturbance involved with the tagging.

Movement patterns of P. cygnus inhabiting inshore reefs All tagged lobsters in the present study were found to be in intermoult condition and were likely to be displaying maximal locomotor activity (Lipcius and Herrnkind 1982). The proportion of red v. white (25 v. 36.4%) lobsters undergoing large-scale movements between fixed receivers was not found to be significantly different in the present study. This is in contrast to previous

Effect of handling and/or tagging on P. cygnus movement and behaviour Like many studies on the movements of lobsters, the movement data collected during the present study relied on capturing, handling and tagging animals, thereby inducing a level of disturbance to the lobsters. Therefore, it is necessary to assess interpretations of lobster movement behaviour in light of these

612


manipulations. Laboratory experiments indicated that tagging can modify the foraging behaviour of P. cygnus by reducing mussel consumption for up to 3 days after tagging. This validated the field results, which showed that the act of tagging and/or handling lobsters has a short-term effect on their behaviour. Four of the 11 detected movements (36.4%) were initiated immediately after tagging at around midday and persisted throughout the following 48 h during the day and night. As P. cygnus is known to be nocturnally active, sheltering under reef ledges by day (Cobb 1981; Jernakoff 1987), this represents behaviour counter to that exhibited by the greater population. Abnormal movement patterns immediately after tagging have also been demonstrated by several other researchers studying spiny lobster movements (e.g. Jasus edwardsii, MacDiarmid et al. 1991; Jasus lalandii, Atkinson et al. 2005; Palinurus elephas, Giacalone et al. 2006). The finding of large-scale movements initiated directly after tagging and handling raises some serious issues on the interpretation of tagging data collected from P. cygnus and other spiny lobsters. Using a conventional mark–recapture technique or acoustic tracking with infrequent position estimates, it is impossible to determine the exact timing of movements and thus to separate movements made immediately after tagging with those made later. Abnormal or exaggerated movements made after tagging are likely to be interpreted as typical for the untagged population and could potentially lead to overestimates on the degree of dispersal from study sites. Intensive manual tracking, frequent diver re-sightings of acoustically tagged animals or automated monitoring of movement using arrays of receivers can help overcome this issue. Chittleborough (1974) and Melville-Smith and Cheng (2002) found that displacing P. cygnus large distances (>500 m) away from their site of capture can induce larger than normal movements. The present study suggests that, even without large displacement of lobsters (i.e.