The study was conducted at Dunwich, North Stradbroke Island, Queensland (27⁰29’45”S, 153⁰24’2”E) (Figure 1) on the 14^th and 15^th of May, 2016. On the shoreline of Moreton Bay, the tidal flats of Dunwich are of sandy to muddy sediments (Kruck, et al., 2009). This area is a semi-diurnal tidal zone, with an average tidal range of 1.1m during the neap tides (half-moon) and 1.8m during spring tides (full or new moon) (Morton, et al., 1987). Both days field work was conducted there was sunny weather, temperatures 15 to 25⁰C, barometer was between 1020 and 1024 mbar and winds from the south west direction at less than 10 knots. The moon was at its first quarter on the 14^th of May. Low tides occurred at 1036 (with a height of 0.67m) and 2210 (0.89m) with high tides at 0322 (2.20m) and 1624 (1.75m) on the 14^th May. The 15^th May had low tides at 1132 (0.71m) and 2322 (0.86m) and high tides at 0426 (2.05m) and 1729 (1.85m).

Crabs were collected on the sand flats between 2 hours either side of the low tides. Crab size was the distances between the third pair of appendages (Figure 2). The crabs were kept with other crabs of similar sizes in buckets of 40cm diameter. The buckets had moist sand and crabs were kept in the shade until tested. At any time the crabs had a maximum of 30 crabs per bucket. All crabs were released and all examination was complete on the day. Crabs were separated into three groups; small (3.0-8.0mm), medium (8.1- 14.0mm) and large (>14.1mm).

A pen of 5.5m x 1.0m was set up on the intertidal edge of the sand flat (Figure 3). A crab pen was set up so soldier crabs could not backtrack on their paths. Crabs were selected as either individuals or groups of 5 and were placed on a 0.5m square piece of wood covered by shelter of 15cm x 15cm. After a stabilising period of 30 seconds, a person walked to the pen and removed the cup to simulate a predator. Timing started once the crabs walked off the wood and was stopped when crabs commenced burying. For groups of 5, burying time started once 3 crabs had left the wood and stopped when 3 crabs had commenced burying. Scurrying distance was measured by the straight line distance from the front centre of the wood to where the crab had buried itself. In groups of 5 the group distance was determined as the average of the 3 median soldier crab distances. If crabs walked in the side of the pen their trial was not counted.

Individual crab results were categorized into size categories as to match with the group data. ANOVA tests were conducted between groups of different crab sizes for both burying time and scurrying distance.

Overall 88 individual crabs and 75 crabs in groups were tested. Individual crabs were both analysed as individual and in size categories to be compared with group crab data. When separated, individual crabs had 42, 22 and 24 in the small, medium and large crab categories respectively. Individual sizes of M. longicarpus ranged from 3.0mm to 20.3mm. For the groups, each category had 5 groups of crabs in them.

Within individual crabs, burying time varied from 0.9 seconds to 36.7 seconds. A linear relationship was placed between the crab size and burying time (Figure 4). It was seen that with one millimetre of size, crab burying time increases by 0.85 seconds. The linear relationship explained 30.98% of variance within the data (R²=30.98) (Figure 4). Crab burying time was not significantly different between individual crabs and groups of crabs for each size category and an overall comparison (Figure 5). Data had large standard deviations away from the mean (Figure 5). Groups had burying times from 7-22 seconds in the small category, 9-33 seconds in the medium category and 14-32 seconds in the large category.

Scurrying distance of individual crabs had a range of 1cm to 430cm. A linear relationship explaining 42.34% of the data (R²=42.34%) exhibited that with every one millimetre of crab length, M. longicarpus s are likely to travel an extra 13.01cm (Figure 6). Groups of crabs were found to not differ significantly to the scurrying distance of individual crabs (Figure 7). Data had large standard deviations (Figure 7). Group means varied from 60-194.3cm in the small category, 51.6-203cm in the medium category and 79.6-282cm in the large category.

With warm sunny conditions, the M. longicarpus were at a peak activity level (Cameron, 1966). Burying time and scurrying distance was found to increase with the size of crab (Figure 4,6). The ability to burrow quickly is likely to avoid predation (Takeda, et al., 1996). Knowing this and considering that crabs of larger sizes have the confidence to take longer to bury and scurry further it can be suggested that crab size effects the rate of predation. Smaller crabs are more vulnerable to a larger group of predators. For example a past study found that soldier crabs of smaller sizes are more susceptible to predation by fiddler crabs (Takeda, 2010). Larger crabs also had the ability to scurry at a faster pace than the smaller crabs, hence are more likely to escape a predator without having to bury. Behavioural adaptations exhibited by soldier crabs include crabs raising themselves up to appear larger (Takeda, 2010). This is likely to be as a form of predator defence (Takeda, 2010) and can lower the chances of being preyed upon.

Similar to the individual data, crabs groups of larger sizes took longer to bury and scurried further. Scurrying distance was not found to be significant between individual crabs and crabs in groups of 5. This suggests that when crabs swarm they do not act differently than when crabs are found individually. However the burying time between individual crabs and groups was found to significantly different suggesting that crab confidence increases when surrounded by other individuals. Supporting this, Murakami and colleagues (2014) showed that when soldier crabs where in larger groups (15 crabs) they were more likely to cross a pool of water then when in smaller groups (5 crabs). This suggests that soldier crabs may find safety in numbers. Behaviour similar to this can be seen all over the animal kingdom. For example fish are seen to shoal when predators are present (Pitcher, 1992), birds flocking in Hawaii (Hart & Freed, 2005) and impala form herds in South Africa (Smith & Cain, 2009).

There are many advantages for congregating in groups such as group size discourages predators (Pitcher, 1992), lower predation on individual (Watt, et al., 1997) and an increase number of eyes to identify predators (Smith & Cain, 2009). However, discovered by Watt and team (1997) with a study on tadpoles, the total strike right of predators on a group increased with larger group sizes but the strike right per individual decreases. This means individuals benefit but the group as a whole does not. Animals sometimes have lower vigilance when in large groups (Smith & Cain, 2009), though for the study this is not considered as M. longicarpus burying time started once the crabs had identified the predator and started to scurry away. As the scurrying distance did not vary amongst individuals and groups, the soldier crabs stayed within the presence of a predator for longer periods of time. Hence it is suggested that M. longicarpus behave similar to the tadpoles when a predator is present. Thus the individual is favoured and not the group.

The greatest limitation for the data was time. Only two days of field work were conducted and although the study managed to find a significant difference for burying time between individual and groups of crabs the standard deviation was large. Insignificant variation were observed throughout all the data indicating that a larger sample size should be to determine whether this is crab behaviour or is by random chance. Acknowledging this, crabs behave different in various temperature and climate conditions (Cameron, 1966; Kelemec, 1979), therefore to keep a fair test field work would have to be conducted in similar conditions, as done for this study.

The group size of crabs was only five as observers were unable to following each individual crab in larger groups. Future research can be conducted using video cameras and image processing software similar to the study by Murakami and team (2014). Using image processing the scurry path and burying time of each individual would be able to be recorded, thus larger groups of soldier crabs would be able to be studied.

Other studies can conducted to determine changing crab behaviour for individuals and groups such as changing the sediment type and sand temperature.