NOTE Nolan and Danilowicz: Use of crest nets for sampling presettlement larvae of reef fishes in the Caribbean Sea 
219 
A Abundance 
(no. of individuals per unit of water volume) 
o 
< 
rinn 1 Ini Ln I In 
01 
Lag (days) 
Figure 4 
Auto-correlation plots of the average nightly abundance (A) and species richness (B) of larvae caught in both crest and 
channel nets for flow-corrected data. These plots identify any significant periodicity in the combined catch of crest versus 
channel nets. Auto-correlation plots are similar to cross-correlation plots but, unlike cross-correlation (which provides a 
comparison of two data sets), auto-correlation allows a comparison of one data set to itself. Lag refers to the number of 
days by which one copy of the data set is offset from the other when the correlation is calculated. Data sets are aligned 
for correlation on the same day at a lag of zero; one data set leads the other by one day for correlation at a lag of +1, etc. 
The auto-correlation function (ACF, correlation coefficient between the data sets at each lag) is on the ordinate. Values 
of the ACF above and below zero represent positive and negative correlations, respectively; the horizontal lines above 
and below the abscissa indicate the upper and lower 95% confidence limits, respectively. 
tively, the larvae used this increased flow to facilitate 
their movement to the reef and the darker conditions to 
improve predator avoidance. Given that larval fish near 
the time of settlement possess impressive swimming 
and sensory abilities, the effect of flow could simply 
be viewed as an interesting variable that masks true 
larval abundance in the water column. 
As reported previously (Shenker et al., 1993; Thor- 
rold et al., 1994; Kingsford and Finn, 1997), rather 
than deploying a net continuously, deploying a net 
around the new moon with an onshore wind would 
optimize collection efforts. The measurements of wind 
speed and direction at the international airport on 
mainland Belize were positively correlated with abun- 
dance and species richness of fish larvae at Turneffe 
Atoll. Because water temperature was found to be 
negatively correlated with the capture of almost all 
species, it is possible that the emptying of warm water 
from the lagoon negatively affects the arrival of larvae. 
All of these factors (lunar period, water temperature, 
and prevailing wind) may be further considered when 
trying to optimize the collection of fish larvae in sta- 
tionary nets. 
In assessing the effort required to install, maintain, 
and deploy the codend of each type of net, we found that 
channel nets were far easier to work with. Because of 
the position of crest nets, they are subject to high wave 
energy and strong currents. Therefore more effort is 
required to anchor the frame to the substrate and more 
time is needed to repair the unavoidable wear and tear. 
Channel nets, on the other hand, are quick to retrieve 
in the case of a storm and require very little ongoing 
maintenance. 
Researchers need to be aware of the additional effort 
required to set and maintain crest nets in comparison 
to other types of nets. The importance of flow has also 
been highlighted, and great care should be taken to 
evaluate this variable when making comparisons of 
larval catch among times and locations. Environmen- 
tal factors which alter this rate of flow seem to have 
the greatest influence on the catch of both stationary 
net types. Given the results of this study, there are 
no obvious obstacles to the use of crest nets in other 
parts of the Caribbean Sea where appropriate sites ex- 
ist, i.e., shallow reef crest with mainly unidirectional 
water flow. Given the greater water flow through the 
