Why Carcinus maenas cannot get a grip on South Africa's wave-exposed coastline

The European green crab Carcinus maenas has established considerable breeding populations in harbours and sheltered bays in the South-Western Cape, South Africa, but appears unable to flourish on the wave-exposed coastline. This study compares the abilities of C. maenas and those of an indigenous rocky-shore crab, Plagusia chabrus, to resist hydrodynamic forces. C. maenas had less than half the vertical tenacity of P. chabrus (371.5g and 780.5g respectively) and was unable to grip against as fast a unidirectional flow (0.23m s−1 vs 0.53m s−1) as P. chabrus. C. maenas also has significantly shorter and lighter limbs than P. chabrus and the dactyls of its walking legs are poorly adapted to grip onto rocky substrata. We conclude that C. maenas is poorly adapted to survive in wave-swept conditions and hence unlikely to displace indigenous crab species along the open wave-exposed coastline of South Africa. However, it may invade other sheltered locations, particularly Saldanha Bay and False Bay.

If C. maenas is limited by wave-exposure, the obvious question is: why? Other species native to South Africa's coastline, such as Plagusia chabrus, the most abundant native crab found in sites colonised by C. maenas, do not seem to be limited by wave action and are found clinging to rock surfaces in extremely wave-swept locations (Griffiths et al. 1992).
We compare the abilities of C. maenas and P. chabrus to actively grip a rock surface against a vertical force and to withstand unidirectional flow. The relative leg length (stance), leg mass (strength) and shape of the dactyls are also compared between the two species. On the basis of these results, we speculate on the potential for future spread of C. maenas.

Material and Methods
C. maenas were collected from Table Bay Harbour using crab traps baited with sardine Sardinops sagax. P. chabrus were collected in the same way, or by divers, from Kalk Bay Harbour in False Bay. Only individuals with all limbs intact were included in the analysis. Crabs were returned to the University of Cape Town, where they were kept in a closedcircuit re-circulating aquarium at 12°C. If held for more than 12h, they were fed with sardine or mussels. Ten C. maenas and 10 P. chabrus were used in the tenacity experiments, and a different 10 crabs of each species in the flume experiments. Thirty-one crabs of each species were frozen overnight and used for morphological analysis.

The European green crab Carcinus maenas has established considerable breeding populations in harbours and sheltered bays in the South-Western Cape, South
Africa, but appears unable to flourish on the waveexposed coastline. This study compares the abilities of C. maenas and those of an indigenous rocky-shore crab, Plagusia chabrus, to resist hydrodynamic forces. C. maenas had less than half the vertical tenacity of P. chabrus (371.5g and 780.5g respectively) and was unable to grip against as fast a unidirectional flow (0.23m s -1 vs 0.53m s -1 ) as P. chabrus. C. maenas also has significantly shorter and lighter limbs than P. chabrus and the dactyls of its walking legs are poorly adapted to grip onto rocky substrata. We conclude that C. maenas is poorly adapted to survive in wave-swept conditions and hence unlikely to displace indigenous crab species along the open wave-exposed coastline of South Africa. However, it may invade other sheltered locations, particularly Saldanha Bay and False Bay.
Keywords: Carcinus maenas, invasive species, tenacity, wave exposure Vertical tenacity was measured by tying a loop of nylon line around the body of each crab. Crabs were then allowed approximately 30 seconds to settle on a large rough rock, submerged in seawater. They were then pulled vertically until they were dislodged. A spring balance, accurate to the nearest 50g and attached to the nylon line, measured the maximum force required to dislodge the crab. Five replicate readings were taken from each of 10 crabs of each species.
Horizontal tenacity was measured in a seawater flume around which water was pumped in one direction. Crabs were tethered to the top of the flume chamber using nylon line tied around their bodies between the second and third pairs of walking legs. The tether allowed the crabs movement, but restricted them to the central portion of the flume chamber. A single sheet of sandpaper was used to standardise the surface inside the chamber and was regularly replaced with another sheet of the same make and abrasiveness.
Each crab was allowed to settle in the flume for 30 seconds before the current was switched on. The current was then increased by 5% at successive five-minute inter- Each of the eight walking legs of 31 crabs per species was straightened and measured from the base to the tip of the dactyl, using Vernier callipers. Carapace width (CW) was measured at the widest point. Average leg length was divided by carapace width to determine a leg-length to body-width ratio. The same crabs were used to determine a leg-mass to body-mass ratio. Each leg was separated from the crab and weighed individually on a scale (± 0.01g) and average leg mass calculated. The body of the crab was weighed independently and leg-mass to body-mass ratios calculated. The dactyls of the walking legs were sketched with a camera-lucida.

Results
The vertical forces needed to dislodge P. chabrus were consistently three times greater than for C. maenas of equivalent carapace width (Figure 1a). There was a strong positive relationship between tenacity and CW in P. chabrus -crabs of 60mm CW having a grip more than 2.5 times that of crabs of 40mm (1 107g and 417g respectively) (t = 5.59, df = 8, p = 0.00). By contrast, there was no significant relationship between CW and vertical tenacity in C. maenas, and tenacity ranged between 200g and 400g (t = 1.21, df = 8, p = 0.26). When vertical tenacity was expressed as a function of body mass (Figure 1b), P. chabrus again showed a much stronger grip per unit mass (mean tenacity 10.1 times body mass) than C. maenas (6.8 times). Expressed as a function of mass, tenacity decreased steeply with increasing size in C. maenas (t = -2.81, df = 8, p = 0.02), but not for P. chabrus (t = -2.00, df = 8, p = 0.08). P. chabrus withstood nearly two times higher horizontal flow velocities than C. maenas (Figure 2), but there was no significant relationship between size and flow velocity in either species (P. chabrus: t = -1.03, df = 8, p = 0.33; C. maenas: t = -0.65, df = 8, p = 0.53).

Discussion
Crabs in a wave-swept environment are constantly in danger of being overturned or swept away as a result of hydrodynamic forces (Martinez et al. 1998, Dickinson et al. 2000, Lau and Martinez 2003. P. chabrus is found on exposed and sheltered shores, whereas C. maenas is conspicuously absent from exposed shores. Our aims were to test if C. maenas is physically less able to deal with hydrodynamic forces than P. chabrus and, hence, to explain why its distribution is limited to sheltered bays, harbours and lagoons in the South-Western Cape.
The vertical force required to dislodge C. maenas was indeed less than half that required for P. chabrus of the same carapace width. Active grasping, especially on rugose 10mm (a) (b) Figure 4: Dactyls of periopods 1-4 of (a) C. maenas and (b) P. chabrus of carapace widths of 61.2mm and 55.0mm respectively rocks, is an important behavioural adaptation to life in wave-swept environments (Martinez 2001), so this may well explain why C. maenas are unable to tolerate the wave forces experienced by P. chabrus on exposed rocky coasts. The negative tenacity to body mass relationship in C. maenas (but not in P. chabrus) further suggests that larger C. maenas will have even more difficulty in gripping onto rocks when hydrodynamic forces increase compared to smaller individuals and P. chabrus. Our second experiment tested the ability of crabs to withstand unidirectional flow, which generates forces of both drag and lift. P. chabrus was clearly able to better withstand unidirectional flows than C. maenas, but the absolute values need to be treated with caution. In nature, crabs have to deal with the added complications of acceleration reaction of the water resulting from wave action (Lau and Martinez 2003) and also need to contend with a wide range of substrata. Moreover, they have to remain mobile in order to mate, feed and escape predation. The flume experiment used only a unidirectional flow and a standardised artificial substratum. Whereas ideal for comparative purposes, these do not represent natural conditions. It is also worth noting that the angle at which crabs face a current may influence the velocity they can withstand (Lau and Martinez 2003). Live crabs were used in our experiments (unlike the work done on Grapsus tenuicrustatus and Pachygrapsus crassipes by Martinez 2001, Lau andMartinez 2003), and thus assumed that the crabs would naturally optimise their stance to avoid losing contact with the surface (Dickinson et al. 2000, Martinez 2001, Lau and Martinez 2003. The leg-length to body-width ratio gives an indication of the maximum lateral stance available to an animal. Longer legs allow for a more lateral stance to increase manoeuvrability (Dickinson et al. 2000) and stability (Martinez et al. 1998, Martinez 2001. Martinez (2001) showed that G. tenuicrustatus changed its posture depending on whether it was moving through water or on land. In water, it widened its stance, increasing its stability and therefore decreasing the possibility of being overturned or washed away. This posture also decreased the drag experienced while underwater (Martinez 2001). P. chabrus has considerably longer legs than C. maenas (Figure 3a), implying that it can adopt wider postures to enhance tenacity (Figures 1, 2). It also has heavier legs than C. maenas, indicating greater muscle volume and hence grip. Thus, both stance and strength will favour P. chabrus when hydrodynamic forces increase.
The curved ends and thick spines on the dactyls of P. chabrus also appear better adapted to grip the substratum than the blunter less spinose dactyls of C. maenas, and the sharp tips may aid grip in small holes and crevices. The dactyls of the last pair of legs in C. maenas are flattened, a characteristic typical of the Portunidae (swimming crabs), where the last pair of legs is usually held horizontally, and is rarely used for walking. By contrast, P. chabrus, a grapsid (shore) crab, has similar dactyls on all walking legs and uses all four pairs to grip the substratum.
Our study suggests that C. maenas is less able than P. chabrus to resist both vertical and horizontal hydrodynamic forces, and that the root of these differences lie in its morphology. We suggest that C. maenas is therefore likely to remain confined to protected harbours and bays along the South African coast. P. chabrus is able to inhabit areas that are unsuitable to C. maenas, suggesting that the indigenous species will not be competitively excluded from the open coast (as it appears to have been in Table Bay Docks -unpublished data from Robinson et al. 2005).