4 
Fishery Bulletin 11 6(1) 
Table 1 
Number and mean lengths (with standard deviations [SD]) for individuals of the sea 
cucumber Holothuria arguinensis sampled by visual census and random sampling 
inside the Ria Formosa, a coastal lagoon in southern Portugal, between November 
2012 and March 2014. The 9 localities sampled (see Fig. 1) were Armona (ARM), 
Barinha (BAR), Cacela (CAC), Culatra (CUL), Fuzeta (FUZ), Praia de Faro to the 
west of the main bridge (PFW), Praia de Faro to the east of the main bridge (PFE), 
Quinta do Lago (QTL), and Tavira (TAV). N signifies number of sampled individuals. 
Locality 
N 
Mean length 
(cm) 
SD 
Minimum 
(cm) 
Maximum 
(cm) 
ARM 
222 
26.51 
6.42 
10.00 
45.00 
BAR 
78 
17.54 
3.31 
12.00 
23.00 
CAC 
1 
23.00 
- 
23.00 
23.00 
CUL 
12 
23.42 
5.47 
13.00 
34.00 
FUZ 
78 
23.38 
6.29 
10.00 
38.00 
PFE 
433 
22.46 
6.27 
4.00 
35.00 
PFW 
332 
21.63 
5.18 
10.00 
48.00 
QTL 
25 
17.48 
4.78 
15.00 
25.00 
TAV 
17 
17.18 
6.27 
9.00 
30.00 
Total 
1198 
23.10 
6.28 
4.00 
48.00 
seasonal von Bertalanffy growth model (Pauly, 1987) 
and the Hoenig seasonal von Bertalanffy model (Hoe- 
nig and Hanumara, 1982)—were fitted to the length- 
frequency data by using the electronic length frequency 
analysis (ELEFAN) system implemented in the length 
frequency distribution analysis (LFDA) software, vers. 
5.0 (Kirkwood 1 ). The best estimators of asymptotic 
length (L c J and the growth coefficient (K) were select¬ 
ed on the basis of maximum value of the score func¬ 
tion and the best fit of the curve to the observed data 
(Kirkwood 1 ). 
The ELEFAN method is based on fitting the von 
Bertalanffy model to length- frequency data (Pauly, 
1987). It works by restructuring the length-frequency 
data, emphasizing the peaks and troughs in the data 
set, and calculating score functions for growth curves 
generated for different combinations of von Bertalanffy 
growth parameters. Every time the growth curves 
passed through the peaks in the restructured data it 
accumulates points, resulting in a score function for 
each curve. The Hoenig function takes into account 
seasonality in growth, including a parameter that fits 
the beginning of the sinusoidal function ( Ts ), and the 
parameter C that expresses the relative amplitude of 
the seasonal oscillation in growth (Pauly et al., 1992). 
Results 
Length data of 1198 individuals of H. arguinensis from 
9 different localities inside the Ria Formosa were used 
for length-frequency analysis (Table 1). 
1 Kirkwood, G. 2001. Enhancement and support of comput¬ 
er aids for fisheries management. Final technical report, 65 
p. MRAG, Ltd., London. [Available from website.] 
For the nonseasonal growth model, estimates of 
K=1.99 and L x -40 cm were obtained with a score func¬ 
tion of 0.460. However, the growth curve does not ad¬ 
equately fit the data (Suppl. Fig.) (online only). 
Considering these results, we then fitted the Hoenig 
model for seasonal growth. This provided estimates of 
K=0.88 and L„= 69.9 cm, and a score function of 0.565. 
The average length was 23.1 cm, sizes ranged from 4 
to 48 cm, and approximately 98% of the individuals 
ranged from 11 to 35 cm (Fig. 2). The growth curve 
showed a higher score function than that of the non¬ 
seasonal growth model (Fig. 3). The value of C= 1 indi¬ 
cates the existence of a period of no growth during the 
year for the species. 
Discussion 
The and K for H. arguinensis were estimated by us¬ 
ing 2 different functions. The first function, the nonsea¬ 
sonal von Bertalanffy model, did not provide realistic 
results. However, when seasonality was included (with 
the Hoenig model), more reliable values were obtained, 
which confirmed the seasonality in the growth of H. ar¬ 
guinensis. The observed growth rate (iC=0.88) indicates 
that this species achieves asymptotic size quickly, even 
faster than tropical commercially valuable holothuri- 
ans, such as the lolly fish (K=0.11), Thelenota anan¬ 
as (K=0.20), Stichopus chloronotus (K=0.45), or Isos- 
tichopus fuscus (K=0.18) (Ebert, 1978; Conand, 1988; 
Herrero-Perezrul et al., 1999). Herrero-Perezrul et al. 
(1999) discussed the difficulties of comparing growth 
parameters derived from von Bertalanffy-like models of 
different sea cucumbers species because not all model 
assumptions are fulfilled. The 3 general assumptions 
