Aquacultural Engineering 70 (2016) 56–62

Contents lists available at ScienceDirect

Aquacultural Engineering
journal homepage: www.elsevier.com/locate/aqua-online

Deterministic and stochastic models for analysis of partial harvesting

strategies and improvement of intensive commercial production of
whiteleg shrimp (Litopenaeus vannamei)
Alicia Estrada-Pérez a , Javier M.J. Ruiz-Velazco a,b , Alfredo Hernández-Llamas c,∗ ,
Iram Zavala-Leal a,b , Leonardo Martínez-Cárdenas a,b
a
Programa de Posgrado en Ciencias Biológico Agropecuarias (CBAP), Universidad Autónoma de Nayarit, Cd. de La Cultura Amado Nervo s/n, Tepic 63255,
NAY, Mexico
b
Escuela Nacional de Ingeniería Pesquera, Universidad Autónoma de Nayarit, Bahía de Matanchén Km 12, Carretera a los Cocos, San Blas 63740, NAY,
Mexico
c
Centro de Investigaciones Biológicas del Noroeste (CIBNOR), Av. Instituto Politécnico Nacional, 195 Col. Playa Palo de Sta. Rita Sur, La Paz 23096, BCS,
Mexico

a r t i c l e i n f o a b s t r a c t

Article history: Partial harvesting is an alternative for managing cash flow of aquaculture farms. We use deterministic and
Received 7 July 2015 stochastic models to analyze zootechnical, water quality and management factors influencing intensive
Received in revised form production of shrimp when incorporating partial harvesting strategies. Data from a commercial farm in
24 September 2015
the State of Nayarit, Mexico were used for modeling. The main factors affecting shrimp production and
Accepted 30 November 2015
its variability were: final weight and growth rate of shrimp, water temperature, pond size, length of daily
Available online 4 December 2015
aeration, and the time when the first partial harvest is conducted. Using the largest pond size (4.0 ha),
minimum length of aeration (7.5 h), and first harvesting at 8.5 weeks resulted in a minimum total harvest
Keywords:
Partial harvesting
of 2690 kg ha−1 (partial and final harvests of 643, 269, 1075, and 703 kg ha−1 ). Using the smallest pond size
Shrimp (1.0 ha), maximum length of aeration (7.9 h), and first harvesting at 11.5 weeks resulted in a maximum
Modeling production total harvest of 3524 kg ha−1 (partial and final harvests of 1111, 234, 997, and 1182 kg ha−1 ). The increase
in shrimp production from improved management was 31%. The stochastic model showed that increasing
farm size from 1 to 40 ha diminished the variability of shrimp production by 84.0%, meaning a reduction
of 2.2% per hectare as size increased. Sensitivity analysis indicated that, overall, final weight of shrimp
and length of aeration are the most important factors determining production. The models can be used
to determine, in future research, the optimum harvesting strategy, using a bioeconomic approach.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction
Cash flow problems lead to more aquaculture business failures
than any other problem (Engle, 2010). Partial harvesting is a strat-
egy for managing cash flow of aquaculture farms. Selling off part
of the inventory reduces stocking densities on the farm, resulting
in faster growth of the remaining fish and greater turnover of the
crop. Revenue from selling off a portion of the crop and the higher
turnover of the crop often improves cash flow and reduces cash
deficits (Engle, 2010).
∗ Corresponding author. Tel.: +52 612 123 8416; fax: +52 612 125 3625.
E-mail address: ahllamas04@cibnor.mx (A. Hernández-Llamas).
Despite the relevance that this practice entails from an eco-
nomic perspective, the number of studies analyzing the advantages
and management of partial harvesting is relatively scarce. Forsberg
(1999), using a bioeconomic approach, determined that it is more
profitable to size-grade salmon prior to harvest compared to har-
vesting a batch of fish with similar size distribution to that of the
standing stock. Brummett (2002) compared three typical partial
harvesting systems and an unharvested control for tilapia in terms
of gross yield and observed that significantly higher yields were
obtained in ponds that were partly harvested by hook and line. Yu
and Leung (2006) used impulsive control theory to develop a par-
tial harvesting model capable of addressing discrete and another
partial harvesting strategies for shrimp. Yu et al. (2009) developed
a model of partial harvesting of shrimp, using the network-flow
approach.

http://dx.doi.org/10.1016/j.aquaeng.2015.11.003
0144-8609/© 2015 Elsevier B.V. All rights reserved.
 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62 57

Partial harvesting in shrimp farms in Mexico is a common 2.2. Deterministic model

practice and, according to FIRA (2009), at least three strategies
are used. One focuses on increasing yields by stocking at high
densities and partly harvesting 12–18 g shrimp. This practice thins
the shrimp stock in ponds, allowing final harvests of medium and
large shrimp. This strategy brings liquidity, but the price of early
harvested shrimp is low. A second strategy seeks large shrimp size
by using low stocking densities and carrying out one or two par-
tial harvests of 16–18 g shrimp, which leads to large shrimp (>30 g)
at final harvest. A third strategy is a combination of the previous
ones, obtaining intermediate yields. Monitoring shrimp prices is
emphasized under this last strategy.
Hernandez-Llamas and Zarain-Herzberg (2011) used a bioeco-
nomic model to analyze shrimp production raised in floating cages
in northwestern Mexico, and determined that partial harvesting
provided higher revenue compared to a one-time harvest. Apart
from that study, there are no antecedents of investigations analyz-
ing partial harvests of shrimp in Mexico.
The principal objective of this investigation was to analyze
the zootechnical, water quality and management factors influenc-
ing intensive shrimp production incorporating partial harvesting
strategies. For this, we developed deterministic and stochastic
models that were calibrated with primary data from an intensive
commercial shrimp farm operating in the state of Nayarit, Mexico.
The production models from this study are intended to be used
as a part of a bioeconomic model in future research for definition
of the management strategies that, not only maximize biological
production, but economic performance as well.
2. Materials and methods
2.1. Data survey
Data from a commercial intensive shrimp farm in the State of
Nayarit were used. There were 29 cases (ponds) operating during
summer at a stocking density of 30 post-larvae m−2 for 19 weeks,
and conducting three partial harvests. The database allowed us
analyzing the following variables for each pond: mean weight of
shrimp, survival, mean pond water temperature, mean dissolved
oxygen content, mean length of daily aeration (length of aeration,
hereafter), time when each of the three partial harvests was done,
and percentage of the stock that was taken at each harvest. The
mean, minimum and maximum values of these variables are shown
in Table 1, together with the dates and mean weight of shrimp at
each partial harvest.
A deterministic stock model was used to calculate shrimp
biomass (bt ), as a function of time (t):
bt = wt nt (1)
where wt is the mean weight of shrimp and nt is the number of
surviving shrimp at time t.
The growth curve proposed by Ruiz-Velazco et al. (2010) was
used to calculate wt :
 3
  1 − kt
wt = wi + wf − wi (2)
1 − kh
where wi is the initial weight, wf is final weight, k is the rate at
which wt changes from its initial value to its final value, t is the
number of time units for which wt is calculateed (such as 5 if wt
is calculateed for five weeks), and h is the number of time units at
final harvest.
To calculate nt , survival was conceptualized as a series of succes-
sive events involving, for every partial harvest, two phases. During
the first phase, survival is calculated until the first partial harvest,
thereafter, survival is calculated for a second phase until the next
partial harvest. This second phase becomes, in turn, the first phase
of the next event involving the next partial harvest. Survival is cal-
culated in this way until the final harvest. As a consequence, the
analysis included four phases over the course of production: the
initial phase previous to the first partial harvest, and three more
immediately after the first, second and third partial harvests. These
phases will be named phase 1, phase 2, phase 3, and phase 4, mean-
ing that they correspond to the periods previous to partial harvests
1, 2, and 3, and the final harvest.
To model these events, the diphasic model proposed by Ruiz-
Velazco et al. (2014) was used. According with the authors, the
general form of the survival equation (Gulland, 1969) used for each
phase is:
nt = N0 exp (−zt) (3)
where nt is the number of live shrimp at time t, N0 is the initial
number of shrimp, and z is the instantaneous rate of mortality.
If F1 (t) and F2 (t) are the algebraic expressions of the general
survival model corresponding to the two phases of survival, we
have:
F1 (t) = N0,1 exp (−z1 t) (4)

Table 1
Mean, minimum and maximum values of the variables in the database used for analysis, together with dates
and means weight of shrimp at each partial harvest.

Variable Mean Minimum Maximum

Weight of shrimp (g) 18.6 14.9 22.9

Water temperature (◦ C) 32.6 31.5 33.3
Dissolve oxygen (mg L−1 ) 6.5 5.4 7.3
Pond size (ha) 2.4 1.0 4.0
Length of aeration (h) 7.7 7.5 7.9
Time of first partial harvest (weeks) 9.5 8.5 11.5
Time of second partial harvest (weeks) 13 12.5 13.5
Time of third partial harvest (week) 15.5 14.5 17.5
First partial harvest of standing stock (%) 23.43 19.0 28.0
Second partial harvest of standing stock (%) 16.84 7.0 27.0
Third partial harvest of standing stock (%) 22.09 15.0 29.0
Date of first partial harvest (day/month) 20/Aug. 10/Sept.
Date of second partial harvest (day/month) 17/Sept. 24/Sept.
Date of third partial harvest (day/month) 01/Oct. 22/Oct.
Weight of shrimp at first partial harvest (g) 10.36 8.2 13.1
Weight of shrimp at second partial harvest (g) 14 12.6 15.1
Weight of shrimp at third partial harvest (g) 16 14.2 18
58 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62
and
F2 (t) = N0,2 exp (−z2 t)
where N0,1 , N0,2 , z1 and z2 are the initial populations and the mor-
tality rates corresponding to the first and second phases.
The following curve is introduced to link both survival functions:
H (t) = F1 (t) + G (t) × (F2 (t) − F1 (t))
where H(t) is the transition function used to simulate shrimp
biomass collected at each partial harvest, and F1 (t), F2 (t) are, as
defined above (Eqs. (4) and (5)), and G(t) is:
1 values of the management variables.
G (t) =  ; S > 0
1 + exp (P − t) /S
where G(t) is the link function, P is the time corresponding to
the central point of the transition between the phases (midpoint
between the beginning and the end of the partial harvest), and S
is the shape factor of the transition curve. The link function G(t)
varies asymptotically between 0 and 1, and G(t) = 0.5 when t = P, so
that for t < P, the function resembles F1 (t) and for t > P, the curve is
similar to F2 (t).
2.3. Correlation and multiple regression analyses
A correlation analysis was conducted, using Statistica 6.0 (Stat-
Soft, Tulsa, OK), to determine possible relationships between the
parameters of the stock model and water quality and management
variables, setting significance at P < 0.05. Complementarily, a mul-
tiple regression analysis was conducted, using datasets of the 29
cases, to calculate parameters of the stock model as a function of
water quality and management variables. Accordingly:
Q = a0 + a1 T + a2 DO + a3 PS + a4 LA + a5 TP1 + a6 TP2 + a7 TP3
+ a8 PH1 + a9 PH2
where Q is any of the parameters in the stock model (except wi ,
n0 , and S), a0 , a1 ,. . ., a9 are regression coefficients, T is mean water
temperature measured during the cultivation period, DO is mean
dissolved oxygen measured during the cultivation period, PS is
pond size, LA is mean length of aeration used during the cultivation
period, TP1, TP2, and TP3 are the times when the partial harvests are
conducted, and PH1, and PH2 are the percentages of the standing
stock that was collected during the first and second partial harvests.
For each parameter of the stock model, the multiple regression
analysis was conducted testing only water quality and manage-
ment variables that previously resulted significant from correlation
analysis. The percentage of the standing stock that was collected
during the third partial harvests was not significant and was not
considered for regression analysis. The multiple linear regression
procedure in Stata 10 (StataCorp, College Station, TX) was used
with P < 0.1 to accept or reject independent variables. This regres-
sion procedure automatically handles collinearity according to the
method described in Rencher (2002).
2.4. Model fitting
The fitting of the stock model and the multiple regression equa-
tions to the total harvested biomass as registered (“observed”)
in the data base was tested as follows. Total harvested biomass
was calculated for each case using the regression equations and
the stock model and compared with the corresponding observed
total yield in the database. A linear regression analysis between
observed and calculated biomass was then performed, setting the
intercept at zero (Poole, 1974), and the Student’s t-test was applied
to determine whether the regression slope differed significantly
from 1 (Zar, 2010). After obtaining a not significant result, an equiv-
alence test was conducted to protect against falsely accepting the
(5) corresponding null hypothesis (Type-II statistical error; Hauck and
Anderson, 1986). For these tests, procedures in Stata 10 were used,
setting significance at P < 0.05 and tolerance error at 10%.
2.5. Management strategies
(6)
Alternative management strategies were defined using combi-
nations of management variables that minimized (worst strategy)
and maximized (best strategy) total harvested biomass. Intermedi-
ate strategies were also defined and analyzed using intermediate
(7)
2.6. Stochastic model
Stochastic elements were incorporated into the stock model
using the “envelope” method described in Vose (2002). For this,
the deterministic value of Q in Eq. (8) was modified:
Qs = Q + r (9)
where Qs is the stochastic value of the parameter and r is a resid-
ual value calculated from a normal distribution fitted to the residual
values resulting from the corresponding multiple linear regression.
Probability distributions were also fitted to values of the water
quality variables that were significant in the regression analysis.
Monte Carlo simulations were conducted, using the proba-
bility distributions of the residuals and water quality variables
as inputs. Output probability distributions of the total harvested
biomass were inferred from the simulations corresponding to the
alternative management strategies. The coefficient of variation
(CV = standard deviation/mean) was used to compare the variabil-
ity in production between management strategies. The effect of
farm size on the variability of production was also estimated for
(8) farms operating 1 and 40 ponds (1 ha each pond). @Risk software
was used for Monte Carlo simulations.
2.7. Sensitivity analysis
The relative importance that management variables have in
determining total shrimp production was calculated by sensitivity
analysis, using the deterministic model, where
 
 percentage of change in production 
AS =   (10)
percentage of change in the variable
where AS is the absolute value of sensitivity attributable to the vari-
able, and the percentage of change in shrimp production and the
management variable correspond to the changes calculated when
passing from the worst management strategy to the best manage-
ment strategy. A high value of AS indicate high importance of the
corresponding variable.
The importance that the parameters of the stock model and
water quality variables have on total harvested biomass was calcu-
lated by sensitivity analysis using the stochastic model. Regression
procedures in @Risk were used for this analysis, where a high abso-
lute value of a regression coefficient indicates high importance of
the corresponding variable.
3. Results
The growth and survival models showed flexibility to fit the
growth and survival data in the database (Fig. 1a and b). Correla-
tion analysis showed that the parameters of the stock model were
significantly correlated with water quality and management vari-
ables (Table 2). Final weight of shrimp was positively correlated
 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62 59

Table 3
Equations resulting from multiple regression analysis to calculate parameters
of the stock model as a function of water quality and management variables.

Equation P

wf = −88.6676 + 1.58827 T + 7.2117 LA 0.00233

k = 0.420721 + 0.001588 PS + 0.055649 LA 0.00005
z1 = 0.0554 + 0.0045 PS − 0.0056 LA 0.00562
z3 = 0.0230 − 0.0007 T + 0.0005 PS 0.00010
z4 = 0.0797 − 0.0063 TH1 0.00024

wf , final weight (g); k, growth coefficient; z1 , z3 and z4 , mortality rates corre-

sponding to the first, third and fourth phases of cultivation; T, water temperature
(◦ C); LA, length of aeration (h); PS, pond size (ha); TH1, time for the first partial
harvest (weeks).

Fig. 1. Sample-fitted growth curve using Eq. (2) (a), and sample-fitted survival curve
using Eqs. (4)–(7) (b).
with water temperature and length of aeration, and inversely cor-
related with the time of the second and third partial harvests. The
growth coefficient was negatively correlated with dissolved oxygen
content and positively correlated with pond size, length of aeration,
the week of the first partial harvest, and the biomass collected in the
first harvest. The mortality rate during the first phase was positively
correlated with pond size and inversely correlated with length of
aeration. The mortality rate during the second phase was inversely
correlated with pond size and the biomass collected during the first
harvest. Mortality rate during the third phase was negatively corre-
lated with temperature and dissolved oxygen content, and directly
correlated with pond size and the weeks of the second and third
harvests. Mortality during the last phase was negatively correlated
with pond size and the weeks of the three partial harvests.
Fig. 2. Results from the equivalence test used for fitting the stock model and the
multiple regression equations.
Results from multiple regression analyses showed that, with the
exception of z2 , the parameters of the stock model were signifi-
cantly related to water quality and management variables (Table 3).
Direct or inverse relationships between the parameters and water
quality and management variables were established, coinciding
with the results from the correlation analysis, although not all of
the variables appearing as significant in the correlation test were
necessary to satisfactorily calculate parameter values using the
regression equations. The fitting of the stock model, using these
equations, proved satisfactory as indicated by the confidence inter-
val of the corresponding equivalence test (−0.048, 0.011; within
−0.05, 0.05; Fig. 2).
The stock model was used to calculate, as a function of time,
partial and final harvests yielding the minimum and maximum
total harvested biomass of shrimp. Minimum total biomass was
calculated using the largest pond size (4.0 ha), the minimum length

Table 2
Correlations between parameters of the stock model and water quality and management variables.

Parameter T DO PS LA TH1 TH2 TH3 PH1 PH2

wf 0.42 0.38 −0.49 −0.50

k −0.45 0.61 0.42 0.44 0.48
z1 0.38 −0.39
z2 −0.43 −0.38
z3 −0.41 −0.50 0.63 0.71 0.82 −0.39
z4 −0.38 −0.63 −0.55 −0.61

wf , final weight; k, growth coefficient; z1 , z2 , z3 , and z4 , mortality rates corresponding to successive

phases of cultivation; T, water temperature; DO, dissolved oxygen; PS, pond size; LA, length of aera-
tion; TH1, TH2, and TH3, time of first, second, and third partial harvests; PH1 and PH2, percentage of
standing stock collected during the first and second partial harvests.
60 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62

Fig. 3. Biomass calculated for the worst and best management strategies using the
stock model as a function of cultivation time.

of aeration (7.5 h), and having the first partial harvest at 8.5
weeks (the worst management strategy). Total biomass calcu-
lated was 2690 kg ha−1 , resulting from successive partial harvests
of 643, 269, and 1075 kg ha−1 and a final harvest of 703 kg ha−1
(Fig. 3a). Maximum total biomass was calculated using the smallest
ponds (1.0 ha), longest aeration (7.9 h), and having the first par-
tial harvests at 11.5 weeks (the best management strategy). Total
calculated biomass was 3524 kg ha−1 , resulting from successive
partial harvests of 1111, 234, and 997 kg ha−1 and a final harvest
of 1182 kg ha−1 (Fig. 3b). Based on these figures, the increase in
total harvested biomass derived from improved management was
31%. From results obtained with the worst and best management
strategies, intermediate strategies were defined (Table 4).
Sensitivity analysis, using Eq. (10) indicated that changing the
Fig. 4. Output probability distributions of shrimp production for a 1 ha pond. (a)
values of the management variables as described above, resulted in Worst management strategy; (b) best management strategy.
increases in production of 70.0 kg ha−1 (pond size), 235.0 kg ha−1
(length of aeration), and 385.0 kg ha−1 (week of first partial har-
vest), representing increases of 2.6, 8.7, and 14.3%, respectively.
Yet, the corresponding changes in the values of the variables were
75.0, 5.3, and 35.3%, resulting in absolute values of sensitivity of
0.03, 1.63, and 0.4, indicating that length of aeration is the most
important management variable.
Results from the Monte Carlo simulation showed that improving
management resulted in increased shrimp production, as reflected
in higher mean production obtained with the best management
strategy, but also in relatively lower variability of production, as
indicated by lower values of the coefficient of variation (Figs. 4a
and b). The simulation also showed that improving management
and increasing farm size diminished the variability of production,
as indicated by progressively lower values of the coefficient of
variation (Fig. 5). Improving management diminished variability
by 8.0–10.0%, while increasing farm size diminished variability by
84.0%.

Fig. 5. Shrimp production calculated for alternative management strategies and

Table 4 farms of 1 and 40 ha. Values above and below the line correspond to the coefficient
Values of management variables used to defines alternative management strategies of variation for 1 and 40 ha, respectively.
(N).

Strategy PS (ha) AT (h day−1 ) TH1 (weeks) Results from sensitivity analysis (Table 5) showed that, regard-
N1 (worst) 4 7.5 8.5
less of the management strategies, the stochastic elements
N2 3 7.63 9.5 associated with growth parameters contributed more importantly
N3 2 7.76 10.5 to total variability of shrimp production, followed by water temper-
N4 (best) 1 7.9 11.5 ature. Mortality related to the successive phases of the cultivation
PS, pond size; AT, length of aeration; TH1, time of first partial harvest. process were less important.
 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62
Table 5
Results from sensitivity analysis of total shrimp production to stochastic elements
associated with growth, mortality, and water quality. The worst and best manage-
ment strategies are analyzed. RC is the regression coefficient indicating sensitivity,
where a high absolute value of the coefficient indicates high sensitivity.
Worst Best
Parameter RC Parameter
wf 0.834 wf
T 0.385 k −0.399
k −0.369 T 0.364
z4 −0.132 z4
z2 −0.086 z1
z1 −0.079 z2
z3 −0.009 z3
4. Discussion
The results indicate that intensive production of white leg
shrimp L. vannamei, using a partial harvesting strategy should be
carefully controlled, because total harvested biomass is strongly
dependent on management variables. An increase, as high as 31%,
could be achieved by improving management, with no need to
increase initial stocking density. Other things being equal, such
increment in production would be equivalent to increase the initial
stocking density from 30 PL m−2 to 40 PL m−2 .
The stock model, together with multiple regression equa-
tions, proved adequate to calculate total harvested biomass, as
was recorded in the database. The growth curve has been used
by several authors to model growth of cultivated shrimp (Ruiz-
Velazco et al., 2010; Hernandez-Llamas and Zarain-Herzberg, 2011;
Hernández-Llamas et al., 2013; González-Romero et al., 2014).
From the growth datasets, we observed that the curve was well fit-
ted, indicating that the curve is also adequate for modeling shrimp
growth under cultivation protocols that use partial harvests. The
diphasic model used to simulate reductions in shrimp populations
from partial harvests showed flexibility in fitting the pond data.
This curve was initially used to calculate mortality of shrimp popu-
lations associated with white spot disease, and fitted cases where
mortality occurred smoothly or abruptly (Ruiz-Velazco et al., 2014).
For our current study, the curve proved adequate for calculating
abrupt reductions in shrimp populations, as a consequence of par-
tial harvests.
The results of the correlation analysis were in general agree-
ment with what is normally expected regarding the influence of
water quality and management variables on intensive shrimp pro-
duction. Higher water temperature had a positive influence by
increasing the final weight of shrimp and reducing mortality dur-
ing the third phase. Low values of the growth coefficient indicate
a more rapid approach of shrimp to their final weight. Higher con-
tent of dissolved oxygen favored rapid growth, as indicated by
its inverse relationship with values of the growth coefficient. Dis-
solved oxygen also positively influenced survival during the third
phase. It is well known that small ponds are easier to control
and manage and that higher production per hectare are usu-
ally obtained (Hernández-Llamas and Villarreal-Colmenares, 1999;
Milstein et al., 2005; Magallon, 2006; González-Romero et al.,
2014). In our study, more rapid growth and higher survival during
the first and third phases occurred in smaller ponds.
The benefits of aeration in intensive production of shrimp have
been thoroughly documented. In our study, despite the narrow
range in length of aeration, we found that a small increase in length
had a positive and strong influence on the final size of shrimp and
in survival during the first phase. Sensitivity analysis indicated that
length of aeration was the most important management factor,
increasing shrimp production to a rate as high as 587.0 kg ha−1 per
hour of additional aeration; an increase of one week before the
61
first partial harvest increased production by 128.0 kg ha−1 ; and a
decrease of 1 ha in pond size yielded an increase of 23.3 kg ha−1 .
Considering the high sensitivity of length of aeration, we verified
that it was not significantly correlated with potential confound-
ing management variables, and directly analyzed the relationship
between total harvested biomass recorded in the data base and
RC length of aeration, finding a significant positive result (P = 0.013).
0.791 Further research is needed to determine benefits from increasing
the length of aeration.
By thinning shrimp populations by partial harvesting, competi-
−0.226
−0.135
tion among remaining individuals diminish, which leads to faster
−0.011 growth (Yu and Leung, 2006). In our study, the timing of partial har-
−0.010 vests and the biomass collected influenced the dynamics of shrimp
production, not only affecting shrimp growth, but also survival. It
is reasonable to expect that early and intense partial harvesting
leads to improved growth and survival of the remaining shrimp.
This explains why, when the first partial harvest was conducted
early in the production cycle, it resulted in fast growth and high
survival during the third phase. When the first harvest was intense,
it resulted in higher survival during the subsequent phase. Early
second and third partial harvests had a clear benefit by increasing
the size of shrimp at final harvest. Benefits were also present by
increasing survival during the third phase. These results indicate
that, depending on the specific dynamics of shrimp population, the
consequences of early and intense partial harvesting are almost
immediately expressed, or delayed.
The results from multiple regression analysis do not reflect the
factors affecting the dynamics of shrimp populations as completely
as correlation analysis because regression models entail simpli-
fications to explain dynamics. Multiple factors, when analyzed
separately, may indicate significant correlations with the response
parameters, but from a statistical point of view, not all factors
are necessary to calculate responses. Significance <0.05 are often
recommended because this results in more independent variables
staying in the model and reducing the risk of omitting important
variables (Quinn and Keough, 2002). In our study, we set signifi-
cance at P < 0.10, attempting to retain important variables affecting
the parameters of the stock model. Some variables were excluded
by regression analyses, but we consider the resulting equations as
adequate in combination with the stock model, after they proved
satisfactory calculating observed shrimp biomass.
Variability in shrimp production diminished as farm size
increased. This is a consequence of extreme variability in some
ponds being compensated by opposite extreme variability in other
ponds. Reduction of variance for a large sample size is expected
to occur according to the central limit theorem of statistics (Vose,
2002). From this study, we estimated that for every hectare that
farm size increases, variability diminishes 2.2%, as indicated by val-
ues of the coefficient of variation. It is worth noting that, according
to the sensitivity analysis and regardless of the management strat-
egy and farm size, stochastic variability of shrimp production was
mainly influenced by variability in growth parameters rather than
mortality.
5. Conclusions
Deterministic and stochastic models were developed to calcu-
late and analyze the intensive production of white leg shrimp L.
vannamei under partial harvesting strategies. The results from these
models indicate that the main factors affecting shrimp production
and variability are: final weight and growth rate of shrimp, pond
water temperature, pond size, length of aeration, and when the
first partial harvest was conducted. We recommend testing the
effect of lengths of daily aeration, longer than those analyzed in
this investigation, on shrimp production.
62 A. Estrada-Pérez et al. / Aquacultural Engineering 70 (2016) 56–62
The production models developed in this study can be used to
determine, in future research and using bio-economic analysis, the
best management strategy. For such analysis, it must be taken into
account seasonality of shrimp prices, and investment and produc-
tion costs associated with alternative management strategies of
pond size, length of culture period, length of aeration, and the time
when the first partial harvest is conducted.
Acknowledgements
The authors are grateful to the shrimp farmers in the State of
Nayarit for their cooperation. Ira Fogel of CIBNOR provided valuable
editorial services. A.E.P. is a recipient of a student fellowship from
Consejo Nacional de Ciencia y Tecnología of Mexico.
References
Brummett, R.E., 2002. Comparison of African tilapia partial harvesting systems.
Aquaculture 214, 103–114.
Engle, C., 2010. Aquaculture Economics and Financing. Willey-Blackwell, Ames, IA.
FIRA, 2009. Situación Actual y Perspectivas del Camarón en México. Banco de
Mexico, 2009 (3).
Forsberg, O.I., 1999. Optimal harvesting of farmed Atlantic salmon at two cohort
management strategies and different harvest operation restrictions. Aquacult.
Econ. Manage. 3, 143–158.
González-Romero, M.A., Hernández-Llamas, A., Ruiz-Velazco, J.M.J.,
Plascencia-Cuevas, T.N., Nieto-Navarro, J.T., 2014. Stochastic bio-economic
optimization of pond size for intensive commercial production of whiteleg
shrimp Litopenaeus vannamei. Aquaculture 433, 496–503.
Gulland, J.A., 1969. Manual of Methods for Fish Stock Assessment: Part 1, Fish
Population Analysis. Food and Agriculture Organization of the United Nations,
Rome.
Hauck, W.W., Anderson, S., 1986. A proposal for interpreting and reporting
negative studies. Stat. Med. 5, 203–209.
Hernández-Llamas, A., Villarreal-Colmenares, H., 1999. TEMA: a software reference
to shrimp Litopenaeus vannamei farming practices. Aquacult. Econ. Manage. 3
(3), 267–280.
Hernandez-Llamas, A., Zarain-Herzberg, M., 2011. Bioeconomic modeling and risk
analysis of raising shrimp Litopenaeus vannamei in floating cages in
northwestern Mexico: assessment of hurricane hazard, stochastic variability of
shrimp and feed prices, and zootechnical parameters. Aquaculture 314,
261–268.
Hernández-Llamas, A., Ruiz-Velazco, J.M.J., Gomez-Muñoz, V.M., 2013. Economic
risk associated with white spot disease and stochastic variability in economic,
zootechnical and water quality parameters for intensive production of
Litopenaeus vannamei. Rev. Aquacult. 5, 121–131.
Magallon, F.J., 2006. Desarrollo y aplicación de una metodología, para evaluar la
variabilidad de la capacidad de carga de la acuicultura de camarón, en la región
del Golfo de California. Universidad de La Habana, La Habana, Cuba, Doctoral
Thesis.
Milstein, A., Islam, M.S., Wahab, M.A., Kamal, A.H.M., 2005. Characterization of
water quality in shrimp ponds of different sizes and with different
management regimes using multivariate statistical analysis. Aquacult. Int. 13,
501–518.
Poole, R.W., 1974. An Introduction to Quantitative Ecology. McGraw-Hill, New
York, NY.
Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for
Biologists. Cambridge University Press, London.
Rencher, A.A., 2002. Methods of Multivariate Analysis. John Wiley & Sons, USA.
Ruiz-Velazco, J.M.J., Hernández-Llamas, A., Gómez-Muñoz, V.M., Magallón, F.J.,
2010. Dynamics of intensive production of shrimp Litopenaeus vannamei
affected by white spot disease. Aquaculture 300, 113–119.
Ruiz-Velazco, J.M.J., Hernández-Llamas, A., Gómez-Muñoz, V.M., 2014. A
continuous diphasic model for prediction of survival of cultivated populations
when affected by disease: the case of shrimp and white spot disease. Aquacult.
Res., http://dx.doi.org/10.1111/are. 12454.
Vose, D., 2002. Risk Analysis: A Quantitative Guide. John Wiley & Sons, Chichester,
West Sussex, UK.
Yu, R., Leung, P.S., 2006. Optimal partial harvesting strategies for a single
production cycle in aquaculture operations. Mar. Resour. Econ. 21 (3),
301–315.
Yu, R., Leung, P., Bienfang, P., 2009. Modeling partial harvesting in intensive shrimp
culture: a network-flow approach. Eur. J. Oper. Res. 193, 262–271.
Zar, J.H., 2010. Biostatistical Analysis. Prentice Hall, Upper Saddle River, NJ.