Giménez & Dick: Shore crab settlement and transport mechanisms 
163 
The period of low colonisation rate was dominated by 
strong SW winds; predicted residual currents were to 
wards the NE, with high speeds. This situation changed 
a few days before the increase in colonisation, with 
winds relaxing and shifting between SE and NW, and 
predicted residual currents tending to move west 
wards. For 2004 (Fig. 4b), spectral analysis of colonisa 
tion showed peaks at periods of 10 and 38 d, while 
short-term variation (<5 d) was not important. 
In 2005, the colonisation rate was mostly moderate, 
with several peaks of high density (>15 ind. d' 1 ) and 
with only short periods (1 to 3 d) of low density (Fig. 6a); 
the only long period of low density was at the end of the 
settlement season (after Day 220). Winds showed low 
variations in magnitude: they were mostly from the 
NW, with short periods (<4 d) from the SE and NE. Pre 
dicted residual currents also showed low variation in 
magnitude,- they flowed towards the NW and SW dur 
ing most of the settlement season, with short periods in 
a NE or SE direction. Spectral analysis of the 2005 data 
showed peaks at periods of 5, 10 and 32 d (Fig. 4c). 
Relationships among variables 
DISCUSSION 
Using correlative analysis, this study evaluated the 
importance of transport processes on the colonisation 
rate of natural substrata by megalopae of the shore 
crab Casein us maenas. In the absence of a manipula 
tive experimental design, causality may be tested by 
repeating observations in space and time. For this rea 
son, a 3 yr data set was used, thus obtaining more 
robust conclusions. The data showed variability among 
and within years, with patterns observed in one year 
not necessarily being repeated in the following year. 
For instance, July of 2003 and 2005 was characterised 
by high colonisation rates, while in 2004, this month 
had lower rates. Within years, 2003 and 2004 were 
characterised by periods of 5 to 10 d of very low coloni 
sation, while in 2005 low colonisation occurred only at 
the end of the setdement season. 
The observed variability in the colonisation rate of 
megalopae may be due to a combination of pre- 
setdement processes (e.g. larval transport affecting 
larval supply) and by larval movements, development 
and mortality in the intertidal. In Carcinus maenas, lar- 
In 2003, colonisation rate was significantly 
correlated with tidal cycle, suggesting 
maximal colonisation 3 to 4 d after spring tide 
(Table 1). Correlations were also statistically 
significant after northern winds (lag = 0 to 5 
d), suggesting that maximum colonisation 
followed north winds and minimum colonisa 
tion followed south winds. Significant corre 
lations between colonisation and predicted 
residual currents suggested that minimum 
colonisation followed after eastwards cur 
rents (lag = 0lo5d). 
In 2004, colonisation rate was not signifi 
cantly correlated with tidal cycle (Table 1). 
Correlations with wind direction were sig 
nificant, suggesting maximum colonisation 
after N-NE winds and minimum colonisa 
tion after SW winds. Correlations with pre 
dicted residual currents suggested that 
minimum abundance followed after N-E 
currents, with a lag of 4 d. 
In 2005, relationships between colonisa 
tion rates and tidal cycle were statistically 
significant, suggesting maximum values 1 
to 3 d after spring tide (Table 1). Correla 
tions with wind were negative, suggesting 
minima after north winds (lag = 5 d); corre 
lations with residual currents were positive, 
suggesting maxima after current relaxation 
or slightly eastward currents. 
Fig. 5. Time series for year 2004. (a) Colonisation of traps by Carcinus mae 
nas megalopae; (b) wind direction; (c) predicted residual currents. Further 
details as in Fig. 3