Aquaculture 254 (2006) 701 – 713

www.elsevier.com/locate/aqua-online

The effect of temperature on gonad, embryonic development and

survival rate of juvenile seahorses, Hippocampus kuda Bleeker
Qiang Lin, Junyi Lu ⁎, Yongli Gao, Li Shen, Jin Cai, Junning Luo
Institute of Aquatic Economic Animals and Guangdong Provincial Key Laboratory for Aquatic Economic Animals,
Zhongshan (Sun Yat-Sen) University, Guangzhou 510275, The People' s Republic of China
State Key Laboratory for Biocontrol, Zhongshan (Sun Yat-Sen)University, Guangzhou 510275, The People' s Republic of China
Received 31 May 2005; received in revised form 3 November 2005; accepted 3 November 2005

Abstract

This investigtion examined the effect of varying temperatures on the gonad development and the reproduction efficiency of
Hippocampus kuda Bleeker. We demonstrated that gonad development, gonadosomatic index (GSI), female and male fecundity,
fertilization rate, hatching rate, and survival rate of juvenile seahorses varied significantly at different temperatures (18, 20, 22, 24, 26,
28, 30 and 32 °C). The periods of gonad development also differed significantly at different temperatures (F7, 32 = 154.065, P b 0.01).
The optimal temperature was 28 °C based on the shortest developmental duration to stage V (mean ± standard deviation) (85.2 ± 6.37
days). The GSI peaked at 28 °C (16.60 ± 0.43%) and it displayed significantly higher than those of other seven trials (P b 0.01). The
relationship between the GSI and the temperatures can be expressed by the following formula: Y = − 1.0737t2 + 8.0768t + 1.013
(r2 = 0.9894, n = 30, P b 0.01). In contrast, the treatments could not develop successfully at 18 and 20 °C, ending before stage II and
stage III, respectively. The range from 26 to 28 °C was suggested as the optimal temperature for fecundity and spawning of H. kuda
because of the large number of eggs in the ovaries. The relationship between fecundity number and temperature can also be
formulated: Y = −30.536t2 + 209.24t + 237.8 (r2 = 0.886, n = 30, P b 0.01). There were dramatic differences for fertilization and
hatching rate among different treatments (F5, 24 = 53.675; F5, 24 = 101.897, P b 0.01). Compared with control seahorses, the results
indicated that the condition indices such as the GSI, fecundity, spawning, fertilization, and hatching during the early development
could affect in part the survival rate of the newborn juveniles (which was also an indirect effect of temperature). However, there were
no marked differences at temperatures from 24 to 28 °C with the similar high survival rate. Based on this, the temperature range from
26 to 28 °C was recommended for gonad development and artificial reproduction of H. kuda.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Hippocampus kuda Bleeker; Temperature; Gonadosomatic index (GSI); Fertilization rate; Hatching rate; Survival rate

1. Introduction demand from the Chinese traditional medicine trade,

aquariums and the souvenir trade on a global scale
Seahorses and other syngnathid fish are especially (Marichamy et al., 1993; Vincent and January, 1995;
popular in some Asian countries due to the strong Lourie et al., 1999). However, because of the overex-
ploitation of seahorses, the Southeast Asian seahorse
⁎ Corresponding author. Tel.: +86 20 84115521; fax: +86 20 wildstocks have declined rapidly (Woods, 2000a; Job et
84036637. al., 2002). In order to meet the demand of the
E-mail address: ls61@zsu.edu.cn (J. Lu). commercial trade, seahorse aquaculture is considered
0044-8486/$ - see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquaculture.2005.11.005
702 Q. Lin et al. / Aquaculture 254 (2006) 701–713

to be a great potential industry in many countries factor (Fulton's K) and HSI are generally used as
(Vincent, 1996; Hilomen-Garcia, 1999), and many indicators of condition in oceanic fish (Lambert and
species are now successfully reared in culture (Forteeth, Dutil, 1997; Grant and Brown, 1999; Comeau et al.,
1997; Giwojna and Giwojna, 1999; Lu et al., 2001). 2000) and can also be used as a measure of condition of
Unfortuntely, some seahorse farms have proved techni- the seahorses, but these indices are not widely used in all
cally challenging with a few success (Vincent, 1996). conditions in seahorses (Woods, 2003a; Foster and
Seahorse aquaculture has great potential to integrate Vincent, 2004). Besides these, the male's brood pouch
both conservation and sustainable development goals plays a great role in eggs' development and fertilization
and nowadays it has been the subject of recent attention after egg transfer (Linton and Soloff, 1964; Vincent and
both by aquaculturists and in the policy domain (Payne January, 1995; Masonjones, 1997; Carcupino et al.,
and Rippingale, 2000; CITES, 2001). 1997, 2002). The relationships between the gonad
Since seahorse culture has received widespread development, spawning number, hatching rate, fertil-
attention, lots of research works have been conducted ization rate and the nutrition of food are extensively
to determine the basic conditions for rearing and studied in the other fish (Izquierdo et al., 2001;
breeding seahorses. For researchers, the survival rate of Watanabe et al., 1984; Fernandez-Palacios et al., 1997;
juvenile seahorses in the first few months was a great Wooster and Bowser, 2000). In contrast, the scientific
handicap (Scarratt, 1995; Forteeth, 1996), however, the studies on gonad development and reproductive effi-
reason for this was still poorly understood. The effect ciency are relatively lacking in seahorses.
of fatty acid composition on survival and growth of H. kuda is widely distributed throughout the tropical
Hippocampus spp. juveniles has been studied (Chang Indo-Pacific region, it usually lives in a broad range of
and Southgate, 2001). Woods (2000a,b, 2003a,b) shallow inshore beds (Lourie et al., 1999; Marichamy et
focuses on the growth and survival of juvenile al., 1993). Because H. kuda has great value in both the
seahorses Hippocampus abdominalis and has pub- Chinese traditional medicines and marine aquarium
lished many scientific studies, which mainly analyzed trades, it has been regarded as a heavily traded seahorse
the impacts of temperature, stocking density, gender species (Job et al., 2002). Fortunately, culturing of the
segregation and broodstock diet. The effect of different oceanic seahorse H. kuda has been reported, and a large
feeds on grown and survival has been examined in scale artificial breeding of H. kuda has also been
Hippocampus subelongatus (Payne and Rippingale, conducted in some countries (Lu et al., 2001), and the
2000), Hippocampus capensis (Lockyear et al., 1997) effect of salinity on the juveniles seahorses, H. kuda, has
and Hippocampus erectus (Correa et al., 1989). Job et been studied (Hilomen-Garcia et al., 2003). However,
al. (2002) had studied the relationship between growth there are relatively few scientific reports on H. kuda
and rearing conditions of H. kuda and demonstrated that published openly. Nowadays, the urgent need is to
the different stages of seahorse required different diet, search for an appropriate method to improve conditions
which was valuable. The first underwater study of wild of artificial breeding and reproductive efficiency in H.
seahorse, Hippocampus whitei reproduction and greet- kuda and develop sound management practice.
ing and courtship behavior has been helpful to artificial To date, there are no correlative scientific reports to
seahorse aquaculture (Vincent and Sadler, 1995). demonstrate the impact of water temperature on gonad
Certain elements of reproductive ecology are com- development and reproduction in H. kuda. The aim of
mon to all seahorse species (Selman et al., 1991; Perante this present study is to identify the effect of water
et al., 2002). Wong and Benzie (2003) have studied the temperature on gonad development, reproductive effi-
effects of environmental parameters on juvenile growth ciency, and survival rate of juveniles in the seahorse H.
of White's seahorse, Hippocampus whitei and indicated kuda. This is the first systematic study of artificial
relationships between gonadosomatic index (GSI), reproductive strategy in H. kuda under large-scale
hepatosomatic index (HSI), reproduction and tempera- artificial culturing conditions.
ture, and food habits of the seahorse also could affect the
reproduction (Texeira and Musick, 2001). Generally, the 2. Materials and methods
temperature that may be optimal for reproduction should
approach the temperatures in the wild when the 2.1. Parent seahorse source
seahorses' artificial breed (Vincent and Sadler, 1995).
The reproductive biology of female big-bellied sea- Healthy parent seahorses (H. kuda), which were the
horses, Hippocampus abdominalis was also studied eighth generation of parent seahorses captured from the
(Poortenaar et al., 2004). To date, Fulton's condition South China ocean, were provided by Zhongshan
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 703

university-Yi dazhou aquatics culture company in L × 6.0 W × 1.8 H m3). 10–15 seahorses were randomly
Guangdong Province, which is the biggest center of sampled from every treatment for measuring the
culturing seahorses in China. standard body weight and body length. They were
placed in a separate nylon net from the other parent
2.2. Parent seahorse rearing seahorses from the same tanks so as not to recount them
in the following experiments. Standard length was
The parent seahorses were cultured in two nursery measured as the sum of the length from the tip of the tail
tanks for reproducing juvenile seahorses (Table 1). The to the mid-point of the cleithral ring and the length from
seawater used in the experiments was treated by double the tip of the snout to the mid-point of the cleithral ring
sand-filtration (particle diameter, 0.05–0.10 mm), bio- (Lourie et al., 1999). Weight measured was the wet
filtered and also treated with ultraviolet radiation. The weight, i.e., the parent seahorses were blotted briefly on
two nursery tanks were linked to a 22 ton seawater the filter paper and then weighed (Job et al., 2002).
recirculating system. Temperature, pH, ammonia, nitrite There were no significant differences in their standard
and nitrate levels were measured every two days in the body length, body weight, or condition factor (CF)
seawater system, and dissolved oxygen (DO) and among all the treatments at the start of the experiments
salinity were measured every three day. Temperature, (ANOVA, P N 0.05) (Table 1).
DO, salinity, and pH values were 29 ± 0.3 °C, 7.6 ± 0.2 At the beginning of the experiments, all the experi-
mg l− 1, 32 ± 0.5‰, and 8.5 ± 0.2, respectively, through- mental treatments were placed at the outdoor temperature
out the culturing period of the parent seahorses. (30 ± 1 °C), then the heating and cooling system was used
Ammonia, nitrite, and nitrate concentrations were all to adjust temperature to the experimental temperatures
less than 0.05 mg l− 1. The two nursery tanks were run as (18, 20, 22, 24, 26, 28, 30 and 32 °C, respectively) at a
a flow-through system with a flushing rate of 6 tons h− 1, speed of 2 °C h− 1. Salinity, pH, ammonia, nitrite, and
and a strainer over the overflow ensured that no parent nitrate levels were also kept as described above. The
seahorses or food items were flushed out of the tank. parent seahorses were fed only with fresh shrimp (Mysis
The parent seahorses were fed with new and fresh spp.) (10–15 mm) that were enriched with local adult
Mysis spp. (10–15 mm ) and frozen adult Artemia (10– Artemia and copepods (Lu et al., 2002). The seawater in
15 mm ) at a density of 2–5 ind L− 1 at the same time, the the eight aquatic tanks was also run as a flow-through
same site and approximately the same quantity every system, and faeces and uneaten food were cleaned up
day. Faeces and uneaten food were siphoned out of the about two hours after feeding every day.
nursery tank at nightfall every day to keep a stable and
pure environment. The nursery tanks were located 2.4. Experiment methods
outdoors under a translucent roof to keep the illumina-
tion intensity at 2500–3000 Lux. According to the developmental progress of the
parent seahorses, the time of gonad development, the
2.3. Experimental treatment gonad weight and the body weight at different stages,
the fecundity number and spawning number, fertiliza-
Eight experimental treatments (18, 20, 22, 24, 26, 28, tion rate, hatching rate and the survival rate of juveniles
30 and 32 °C, respectively) were conducted in these were recorded in the experiments. The parent seahorses
experiments, and each treatment consisted of 150 pair were sampled randomly from each treatment and then
parent seahorses, which were randomly sampled from immediately experimented upon.
the two nursery tanks. The parent seahorses of eight Based on the experimental design, the following
experimental treatments were cultured separately in experiments were conducted in different replicates
eight aquatic tanks, which were the same size (5.0 (Table 2).

Table 1
The samples, H. kuda Bleeker, were randomly selected and then divided them into eight treatments, and the one had been deleted if too small
Temperature(°C) 18 20 22 24 26 28 30 32
SW (g) 4.88 ± 1.36 5.16 ± 1.81 5.40 ± 1.65 5.10 ± 1.77 4.96 ± 1.64 5.32 ± 1.49 4.86 ± 1.43 4.99 ± 1.45
SL (cm) 7.99 ± 1.73 8.05 ± 1.88 8.10 ± 1.70 7.94 ± 2.03 7.87 ± 1.74 8.23 ± 1.37 7.83 ± 1.58 7.79 ± 1.77
Note: Data in table were arithmetic mean and sample standard deviation (mean ± S.D.).
SW: Standard weight (Lourie et al., 1999).
SL: Standard length (Job et al., 2002).
704 Q. Lin et al. / Aquaculture 254 (2006) 701–713

The experiment of survival rate in different temperature

different controls (Inds)

Note: Based on the experimental design, the experiments of testing Fecundity and Spawning of H. kuda Bleeker, were 5 replicates because of their low veracities, and it was difficult to give the
The sampling numbers
of juvenile seahorse in
2.4.1. Gonad development time and GSI
One (or two) seahorse from each treatment were
randomly sampled for microscopic examination to

* 3 Repl
* 3 Repl
* 3 Repl
* 3 Repl
* 3 Repl
* 3 Repl
display the gonad development every three days. In
this experiment, based on the morphologic characters of
sperm and eggs and the histological characters of ovary,
50
50
50
50
50
50
0
0
the whole period of gonad development was divided
into five stages, such as Stage I, Stage II, Stage III, Stage
different treatments (Inds)

IV, and Stage V (Liu and He, 1992). After reaching stage
The sampling numbers
of juvenile seahorse in

I of gonad development, five seahorses were randomly

selected from each treatment to measure body weight
and ovary or spermary weight, and then the GSI (Mean ±
Repl
Repl
Repl
Repl
Repl
Repl

S.D.%) of each was calculated. The same method was

50 * 3
50 * 3
50 * 3
50 * 3
50 * 3
50 * 3

conducted in following about three months until reach-

0
0

ing the stage Vof gonad development. At the same time,

according to the experimental design, three replicates
The experimental design of temperature's effect on gonad, embryonic development and survival of juveniles seahorses, H. kuda Bleeker

testing Hatching rate

were conducted by three different experimental teams to

The experiment of

provide a credible and precise result (Table 2).

Repl
Repl
Repl
Repl
Repl
Repl

2.4.2. Fecundity number and spawning number

5 M *3
5 M *3
5 M *3
5 M *3
5 M *3
5 M *3

Following the above experiment, after reaching stage

0
0

V five female seahorses were randomly sampled from

each treatment, the ovaries removed for microscopic
The experiment of
testing Spawning

examination, the number of eggs counted and recorded

(Mean ± S.D. %). When the greeting partners had
Repl
Repl
Repl
Repl
Repl
Repl

courted and finished the egg transfer at mating about

5 M *5
5 M *5
5 M *5
5 M *5
5 M *5
5 M *5

five days later, five to eight male seahorses with rotund

and swollen brood pouches were selected from each
0
0

treatment and the pouches were opened carefully to

The sampling numbers of seahorse (Inds)

schedule for sampling because different treatments had a different developmental time.
The experiment of

examine the fertilized eggs. The sum of fertilized eggs

testing Fecundity

and unfertilized eggs was the summation of spawning

*5 Repl
*5 Repl
*5 Repl
*5 Repl
*5 Repl
*5 Repl

number of each seahorse (Mean ± S.D. %). Based on the

experimental design, five replicates were conducted to
5F
5F
5F
5F
5F
5F

offer a precise result (Table 2).

0
0

2.4.3. Seahorse fertilization rate and hatching rate

The experiment
of testing GSI

Five rotund male seahorses with swollen brood

*3 Repl
*3 Repl
*3 Repl
*3 Repl
*3 Repl
*3 Repl

pouches were sampled and cultured in a separate white

nylon net tank (diameter, 0.05–0.10 mm; Net tank,
5F
5F
5F
5F
5F
5F

1.5 × 0.8 × 1.4 m3) from the tank at each treatment for
0
0

reproducing the juvenile seahorses. The durations of the

numbers of parent

male seahorses' pregnancies were not identical for the

seahorses (Pairs)

same treatment (Ranged from 15 days to 30 days),

Repl : Replicate of the experiment.
F : The Female of parent seahorse.
The starting

M : The Male of parent seahorse.

therefore, it was a relatively long duration before every

brood of juveniles was counted in succession (Mean ±
150
150
150
150
150
150
150
150

S.D. %), and separated from the parent seahorses until

all the males finished reproducing. Based on the
experimental design, three replicates were conducted
°C
°C
°C
°C
°C
°C
°C
°C

(Table 2). When the number of newborn seahorses was

Treatment 18
Treatment 20
Treatment 22
Treatment 24
Treatment 26
Treatment 28
Treatment 30
Treatment 32

* : Mutiply.
Treatments

gained, the mean number of juveniles of each male

Table 2

seahorse was given. The fecundity number, spawning

number and the number of fertilized eggs were all
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 705

92.8 ± 3.97ab
90.4 ± 4.03ab

89.4 ± 8.04ab
90.4 ± 3.50ab
95.0 ± 3.16b

86.8 ± 5.98a
achieved in the above experiments, too. Therefore, the

Stage V
fertilization rate and the hatching rate were offered
(Mean ± S.D.).

2.4.4. Juveniles survival

74.2 ± 3.12ab

Note: Data in table were arithmetic mean and sample standard deviation (mean ± S.D.), and the different superscripts in the same column show significant difference (P b 0.05).
78.6 ± 2.65b

78.8 ± 3.19b
66.8 ± 7.22a
67.2 ± 5.42a

81.0 ± 4.77c
50 juveniles were randomly sampled. They were

Stage IV
newborn and separated from the males on the morning
after birth, and cultured in yellow plastic aquatic tanks
(1.0 × 0.8 × 0.6 m3). The juveniles were not born at the
same time at the different temperatures (22, 24, 26, 28,

58.4 ± 1.02ab
58.6 ± 3.38ab

63.6 ± 4.92b
64.8 ± 6.43b
65.6 ± 6.34c

52.8 ± 8.84a
53.6 ± 6.18a
30 and 32 °C). That the gonads might affect the

Stage III
juveniles' survival rate during gonad development at

Developmental time of spermary (day)

different temperatures was also considered, therefore,
the juveniles in each treatment (22, 24, 26, 28, 30 and 32
°C) were separately sampled to examine the effect of

51.6 ± 5.57ab
53.2 ± 4.02ab
52.2 ± 4.75ab

54.8 ± 4.71ab
51.4 ± 5.43ab
59.2 ± 6.97b

48.9 ± 2.76a
c
70.8 ± 6.11
gonad development at different temperatures on surviv-

Stage II
al, which required not rearing the juveniles of different
treatments synchronously.
In order to find a contrast to compare with the above

41.4 ± 2.25abc
treatments, 6 controls were designed. 900 juveniles (at

37.4 ± 3.26ab
38.8 ± 3.25ab
37.4 ± 4.36ab

41.0 ± 3.74bc
d

43.8 ± 2.99c

35.4 ± 3.61a
55.2 ± 2.99
least 50 individuals × 6 controls × 3 replicates because of

Stage I
the unnatural death, e.g., juveniles' collectting and
transport) were randomly sampled from the outdoor
tanks in the morning (water temperature 28.7 °C), and
then were reared in indoor tanks (1.0 × 0.8 × 0.6 m3) for

92.6 ± 4.32ab
90.0 ± 5.40ab

87.2 ± 6.85ab
88.4 ± 7.17ab
94.8 ± 4.17c

85.2 ± 6.37a
Effect of temperature on the gonad development (day) of male and female seahorses, H. kuda Bleeker

3 h (water temperature 28 °C). 6 controls (22, 24, 26, 28,

Stage V

30 and 32 °C) and 3 replicates were set for each control,

and 50 juveniles in each replicate (Table 2).
Throughout the culturing period, the seawater quality 74.4 ± 3.26bc
67.2 ± 7.03ab
78.6 ± 2.24c

66.2 ± 6.24a
77.8 ± 5.01c
78.2 ± 6.01c
parameters and diets were the same in both experimental
Stage IV

and control tanks, and each aquatic tank was linked to

each treatment tank and run as a flow-through system
with a flushing rate of 5 l per minute. The juveniles were
mainly fed with wild-caught nauplius of copepod nauplii
58.0 ± 2.28abc
58.6 ± 3.61abc

54.4 ± 6.09ab
62.8 ± 5.56bc

( index of selectivity was 0.704) and the larvae of Moina

64.6 ± 5.53c

53.0 ± 8.81a

63.4 ± 5.00c
Stage III

mongolica ( index of selectivity was 0.5125) (Chen et

al., in press). During the 10 days, the dead juveniles of
the different treatments were recorded to determine the
Developmental time of ovary (day)

final survival rate of the juvenile seahorses.

52.4 ± 5.22ab
52.8 ± 3.87ab

55.2 ± 5.46ab
59.8 ± 6.55b
c

51.4 ± 4.88a
51.2 ± 3.97a

50.6 ± 5.89a
73.6 ± 5.46
Stage II

2.5. Data analyses

Standard length and weight were analyzed sepa-

rately using one-way ANOVA, while the comparisons
40.0 ± 1.67ab
39.8 ± 1.33ab
43.6 ± 2.73b
c

38.0 ± 2.28a
38.6 ± 2.73a
37.2 ± 3.31a
36.0 ± 3.63a
56.8 ± 5.19

of gonad developmental time, GSI, fecundity and

Stage I

spawning number, fertilization rate, hatching rate, and

the survival rate of the juveniles at different tempera-
tures were analyzed by one-way ANOVA using the
software SPSS.11.5 (Statistical Program for Social
Temperature

Sciences.11.5). The regression analysis was used to

Table 3

test the relationship between GSI and temperature.

°C

18
20
22
24
26
28
30
32

When necessary, Tukey's multiple comparison test was

706 Q. Lin et al. / Aquaculture 254 (2006) 701–713

100

Time of ganodal development (day)

90 Treatment 22˚C
80 Treatment 28˚C
70

60

50

40

30
I II III IV V
Developmental stage

Fig. 1. Comparison of durations between developmental stages at the temperatures of 22 and 28 °C in male seahorses, H. kuda Bleeker. The durations
between stages varied significantly at the same temperature (P b 0.05), e.g., At 22 °C the durations between stage I and stage II, stage II and stage III,
stage III and stage IV, stage IVand stage V were 12.80 ± 3.81b, 4.00 ± 1.35c, 22.60 ± 2.51a, and 16.20 ± 4.63b days, respectively (P b 0.05), and a similar
development trend happened at 28 °C.

performed to analyze the variances of the data among develop successfully because of the low temperatures,
treatments (Zar, 1996). and they only reached stage II (Male: 70.8 ± 6.11 days;
Female: 73.6 ± 5.46 days) and stage III (Male: 65.6 ±
3. Results 6.34 days; Female: 64.6 ± 5.53 days), respectively.
Data for the other four stages (stage I, stage II, stage
3.1. The effect of temperature on gonad development of III, and stage IV) at different temperatures displayed
seahorse Hippocampus kuda similar trends as stage V, they also showed sig-
nificant difference (e.g., stage III: F5, 24 = 7.513, P b
The durations of gonad development at different 0.01) (Table 3). In addition, the duration of gonad
stages varied significantly at temperatures of 18, 20, 22, development at the same temperature among different
24, 26, 28, 30 and 32 °C (F7, 32 = 154.065, P b 0.01). The stages also displayed significant difference (e.g., 22 °C:
results showed that the durations to reach stage V were F3, 16 = 15.561, P b 0.01), for example, durations at
relatively shorter at the range from 26 to 30 °C (Table 3), different stage varied at 22 and 28 °C (Fig. 1).
and 28 °C was the optimum temperature for gonad
development because it was when the shortest develop- 3.2. The effect of temperature on Gonadosomatic index
ment (Male: 86.8 ± 5.98 days; Female: 85.2 ± 6.37 days) GSI
occurred. However, when the temperature declined to
24 °C or increased to 32 °C the duration to stage V Water temperature was a very important parameter
increased. The specimens at 18 and 20 °C could not for the GSI of the seahorses with the artificial cultured

Table 4
The gonadosomatic index (GSI) in gonad developmental stages of seahorse, H. kuda Bleeker, in different temperatures
Temperature Stage II Stage III Stage IV Stage V
(°C)
SW (g) GSI (%) SW (g) GSI (%) SW (g) GSI (%) SW (g) GSI (%)
18 18.28 ± 4.92a 1.00 ± 0.30e
20 19.30 ± 5.67a 1.30 ± 0.49d 21.58 ± 4.83a 1.67 ± 0.64e 22.33 ± 4.45a 26.40 ± 3.79b
22 18.30 ± 5.11a 1.82 ± 0.86c 21.43 ± 3.76a 2.68 ± 1.25d 24.07 ± 3.54a 4.56 ± 0.60d 26.62 ± 1.54ab 8.04 ± 0.26e
24 19.75 ± 5.07a 3.05 ± 0.12b 22.47 ± 4.02a 3.49 ± 0.11c 24.57 ± 3.04a 5.39 ± 0.84c 26.74 ± 3.70a 13.02 ± 0.79d
26 20.85 ± 4.48a 3.26 ± 0.11b 23.50 ± 3.27a 4.54 ± 0.41b 25.17 ± 2.81a 7.75 ± 0.69bc 27.12 ± 2.79a 15.05 ± 0.80ab
28 19.87 ± 2.50a 5.01 ± 0.86a 23.37 ± 2.92a 6.80 ± 0.76a 25.53 ± 1.45a 8.56 ± 0.52a 26.93 ± 2.32a 16.60 ± 0.43a
30 20.37 ± 3.35a 5.11 ± 0.98a 24.60 ± 2.41a 6.89 ± 0.75a 25.67 ± 0.90a 8.91 ± 0.79a 25.80 ± 2.52b 14.51 ± 1.13bc
32 20.80 ± 3.55a 1.68 ± 0.25c 23.20 ± 1.37a 2.53 ± 0.37d 25.73 ± 0.61a 3.10 ± 0.19e 25.20 ± 3.41b 10.76 ± 1.11e
Note: Data were shown in mean ± S.D. The different superscripts in the same column show significant difference (P b 0.01).
SW: Standard total wet weight of seahorse (g).
GSI: Gonadosomitic index, GSI = 100 (GW/W).
GW and W represent gonad and somatic weights, respectively.
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 707

130 18
temperatures (F3, 16 = 15.561, P b 0.01) (Table 4). The
16
Gonad developmental time (day)

120 relationship between GSI (%) and duration of gonad

14 development was given in the Fig. 2, which also
110
12 indicated clearly that 28 °C was the optimum
temperature for gonad development. In contrast,

GSI (%)
100 10
when the temperature exceeded or was less than 28
90 8
°C, duration of gonad development lengthened and
6 GSI inversely diminished (Fig. 2). The relationship
80
4 among durations of gonad development, GSI, and
70 Gonad developmental time temperature can be expressed by the following
2
GSI
60 0
multinomial formulas:
22 24 26 28 30 32
Temperature (˚C) Y ¼ −1:0737t 2 þ 8:0768t þ 1:013ðr2 ¼ 0:9894;
n ¼ 30; Pb0:01Þ
Fig. 2. Comparison of the Gonad developmental time (day) and the
GSI (%) of H. kuda Bleeker, at different temperatures. 18 and 20 °C
were not marked in the figure because the gonad ceased to develop at Y: GSI%; t: water temperature
both these temperatures. Data in figure were Arithmetic mean and
Sample standard deviation (mean ± S.D.) (P b 0.01). Y ¼ 6321t 2 −5:9394t þ 100:9 ðr2 ¼ 0:8779;
n ¼ 30; Pb0:01Þ
condition. The results indicated that body weights of
male seahorses cultured at different temperatures were Y: durations of gonad development; t: water
not significantly different at different stages (F6, 14 = temperature.
2.429, P = 0.080), but water temperature mainly
influenced the gonad weights, which made a dramatic 3.3. The effect of temperature on fecundity and
variance of GSI of male seahorses at different spawning numbers
temperatures (F6, 14 = 49.872, P b 0.01). The GSI of
stage V displayed relatively higher at the temperatures The difference of fecundity number and spawning
of 24, 26, 28 and 30 °C (Table 4). The investigation number was particularly evident at different tempera-
demonstrated that 28 °C was the optimum temperature tures (F5, 12 = 29.777, P b 0.01; F5, 12 = 76.578, P b 0.01).
because of the largest GSI (16.60 ± 0.43% at stage V). The results indicated that 26–28 °C was the optimum
Among all the experimental treatments, treatments at range of temperatures for fecundity (580 ± 36.88 inds,
the temperatures of 22–32 °C could develop to the 608 ± 30.59 inds, respectively) and spawning (548 ±
stage V successfully, in contrast, treatments at 18 and 46.65 inds, 556 ± 28.71 inds, respectively) of male
20 °C did not reach stage V because the temperature seahorses H. kuda (Table 5).
was below the limited temperature of gonad develop-
ment, and gonad of treatment 18 °C and treatment 20 3.3.1. The relationship between temperature and
°C could only develop to stage II (1.00 ± 0.30%) and fecundity
stage III (1.67 ± 0.64%), respectively. GSI of the other The water temperature greatly influenced the GSI,
four stages (stage I, stage II, stage III, and stage IV) and the results also showed that the GSI was directly
also displayed the similar trends to stage V at different related to water temperature when temperatures ranged

Table 5
Comparsions of condition indices of H. kuda Bleeker at different temperatures, these indices mainly included Fecundity number (inds), Spawning
number (inds), Fertilization rate (%), Hatching rate (%), Survival rate (%) of treatments and controls of juveniles
Temperature (°C) 18 20 22 24 26 28 30 32
e e bc b a a a
Fecundity number 0.00 0.00 444 ± 20.59 490 ± 14.14 580 ± 36.88 608 ± 30.59 548 ± 37.09 372 ± 31.24d
Spawning number 0.00f 0.00f 348 ± 17.20d 448 ± 25.61c 548 ± 46.65ab 556 ± 28.71a 510 ± 14.14b 232 ± 31.23e
Fertilization rate (%) 0.00e 0.00e 70.17 ± 3.05c 75.71 ± 3.34bc 90.73 ± 3.62a 92.41 ± 3.57a 77.31 ± 2.40b 54.29 ± 6.25d
Hatching rate (%) 0.00f 0.00f 78.91 ± 5.45c 84.40 ± 2.74b 91.95 ± 2.45a 94.73 ± 2.22a 71.69 ± 2.18d 50.66 ± 2.93e
Survival rate (%) 0.00e 0.00e 82.98 ± 2.25cd 85.59 ± 1.72bc 88.03 ± 1.94bc 90.74 ± 1.40a 77.39 ± 2.76c 45.88 ± 4.88d
Survival rate of controls (%) 0.00f 0.00f 74.98 ± 3.49d 82.49 ± 3.51bc 87.37 ± 1.78ab 88.59 ± 1.37a 80.99 ± 1.93c 65.78 ± 3.93e
Note: Data were shown in mean ± S.D. The different superscripts in the same row show significant difference (P b 0.05).
708 Q. Lin et al. / Aquaculture 254 (2006) 701–713

700

Fecundity and Spawning number (Inds)

from 22 to 32 °C. When the temperature increased from
22 to 28 °C the fecundity number accordingly improved, 600
and the fecundity number (608 ± 30.59 inds) at 28 °C
was the highest, which was significantly different from 500
the experimental treatment 32°C (372 ± 31.24 inds)
(F5, 24 = 29.777, P b 0.01). The variational trendline of 400

GSI (%) at different temperatures displayed similar

300
trends to that of the fecundity number (Fig. 3). In
contrast, the experimental temperatures of treatment 18 200 Fecundity number
°C and treatment 20 °C were so low that gonads could Spawning number
not develop naturally, and both of them could not yet 100
22 24 26 28 30 32
produce germ cells in ovaries. The relationship between Temperature (˚C)
fecundity number and temperature can be expressed by
the multinomial formula: Fig. 4. The relationship between Fecundity number (inds), Spawning
number (inds), and Temperature (°C) in parent seahorses, H. kuda
Y ¼ −30:536t 2 þ 209:24t þ 237:8 ðr2 ¼ 0:886; Bleeker, which could be expressed by the multinomial formulas: Y =
− 30.536 t2 + 209.24t + 237.8 (r2 = 0.886, P b 0.05) (Y: fecundity
n ¼ 30; Pb0:05Þ number; t: water temperature); Y = − 44.179 t2 + 298.22 t + 66.6
(r2 = 0.9167, P b 0.05) (Y: spawning number; t: water temperature).
Y: fecundity number of female seahorse; t: water Data in figure were Arithmetic mean and Sample standard deviation
temperature. (mean ± S.D.) (P b 0.05).

3.3.2. The relationship between temperature and say, the spawning rate was high. However, when
spawning temperatures were out of this range from 24 to 30°C,
The spawning number of seahorses displayed the spawning rate was relatively lower, and the
significant variance at different temperatures (F5, 24 = spawning rates were (348 ± 17.20)% / (444 ± 20.59 )
76.578, P b 0.01) (Fig. 4). The results indicated that 26 and (232 ± 31.23) % / (372 ± 31.24) at temperatures of 22
and 28 °C were the optimum temperatures for spawning and 32 °C (Fig. 4).
of seahorses because of their higher spawning number
(548 ± 46.65 inds and 556 ± 28.71 inds). When temper- 3.4. The effect of temperature on fertilization rate and
ature decreased to 22 °C and increased to 32 °C the hatching rate
spawning number declined greatly, Fig. 4 showed when
temperatures ranged from 24 to 30°C the fecundity Under artificial culture conditions, the fertilization rate
number and spawning number were all high. That was to and hatching rate were significantly affected by the water
temperature. We continued our experiments to evaluate
18 700 the excellent conditions of fertilizing and hatching.
16 650
Fecundity Number (Inds)

14
3.4.1. The relationship between temperature and
600
12
fertilization
550 The fertilization rates at temperatures of 22, 24, 26,
GSI (%)

10
500 28, 30 and 32 °C strongly fluctuated (F5, 24 = 53.675,
8
450 P b 0.01). The result demonstrated that a particularly
6
400
high fertilization rate (N 90%) of 90.73 ± 3.62% and
4 GSI 92.41 ± 3.57% occurred at temperature of 26 and 28 °C,
2 350
Fecundity Number respectively. However, although eggs and sperm of
0 300 seahorses could be fertilized at a relatively lower
22 24 26 28 30 32
Temperature (˚C) temperature (b 24 °C) or a higher temperature (N 30
°C), the fertilization rates were very low (70.17 ± 3.05%
Fig. 3. The relationship between GSI (%) and Fecundity number at 22 °C; 54.29 ± 6.25% at 32 °C). In addition, compared
(Individuals) of parent seahorses, H. kuda Bleeker, at the different with the optimum temperatures (26 to 28 °C), the
temperatures. The formula of GSI was Y = − 1.0737t2 + 8.0768t + 1.013
(r 2 = 0.9894, P b 0.01) (Y: GSI%; t: Water temperature), and data in
fertilization in extreme temperatures displayed instabil-
figure were arithmetic mean and sample standard deviation (mean ± ities and the standard deviation (S.D.) was relatively
S.D.) (P b 0.01). higher (e g : 54.29 ± 6.25% at 32 °C) (Fig. 5, Table 5).
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 709

100
3.4.2. The relationship between temperature and
hatching 90
Water temperature significantly affected the fertiliza-
80
tion rate during artificial breeding and reproduction,

Survival rate (%)

with observed hatching rates of 94.73 ± 2.22%, 71.69 ± 70
2.18%, and 50.66 ± 2.93% at temperatures of 28, 30 and
32 °C, respectively (F5, 24 = 101.897, P b 0.01) (Table 5). 60

Compared with all experimental treatments, 28 °C was 50 Survival rate of controls

the optimal temperature for hatching because of its
highest hatching rate, and the next highest hatching rate 40 Survival rate of treatments
was 91.95 ± 2.45% at 26 °C.
30
During the hatching experiment, the flat empty brood 22 24 26 28 30 32
pouch of the male seahorses swelled after receiving eggs Temperature (˚C)
from the females, enlarged through pregnancy, and
Fig. 6. Comparison between the Survival rate (%) of treatments and
became very rotund in the days prior to giving birth. controls of juveniles, H. kuda Bleeker, at the different temperatures.
When the brood pouches of the males were opened for The relationship between both of them could be expressed by the
microscopic examination to observe the hatching following formulas: Y = − 2.92 t2 + 19.032 t + 57.768 (r2 = 0.9733,
progress, large amounts of light orange fertilized eggs P b 0.01) (Y: survival rate of controls of juveniles; t: water
were observed to be lined tightly in order under the light temperature); Y = − 4.1743 t2 + 23.295 t + 60.214 (r2 = 0.9191,
P b 0.01) (Y: survival rate of treatments of juveniles; t: water
microscope. In contrast, the unfertilized eggs were temperature). Data in figure were Arithmetic mean and Sample
getting small and gray and rotting little by little. standard deviation (mean ± S.D.) (P b 0.01).

3.5. The effect of condition indices of early development difference between experimental treatments and con-
on the survival rate trols, and this difference was also shown to be
significant but opposite at the temperature of 22 °C
The early development of parent seahorses could (Fig. 6 and Table 5). Therefore, the above finding
affect the survival rate of newborn juveniles, and this proved that early development could affect in part the
effect was indirect from temperature, which had already survival rate of the juveniles. The study also displayed
affected the gonad development of parent seahorses that the survival rates differed greatly at temperatures of
(above experiments). The result showed that the survival 22, 24, 26, 28, 30 and 32 °C (F5, 12 = 26.767, P b 0.01)
rate was similar between experimental treatments and (Fig. 6). The survival rates were relatively higher and
controls at temperatures from 24 to 28 °C, and all did not display significant difference when temperature
exceeded 80% (Fig. 6). However, when the temperature ranged from 22 to 28 °C (F3, 8 = 0.492, P = 0.997). The
was 30 to 32 °C, the survival rate displayed significant optimum temperature range was from 26 to 28 °C, and
the survival rates were 88.03 ± 1.94% and 90.74 ± 1.40%
100 in experimental treatments, and 87.37 ± 1.78% and
90 88.59 ± 1.37% in controls.
Fertilization rate (%)

80
4. Discussion
70

60 Temperature is one of the primary parameters that

50 can greatly affect the growth and survival of the
seahorse (James and Woods, 2001; Wong and Benzie,
40
2003). In this present study, the results demonstrate that
30 temperature can significantly influence the gonad
22 24 26 28 30 32
Temperature (˚C) development, reproductive efficiency, and survival of
the juveniles in seahorse H. kuda. When salinity,
Fig. 5. The relationship between Fertilization rate (%) and temperature dissolved oxygen (DO), and pH values are 32 ± 0.5‰,
in parent seahorses, H. kuda Bleeker, which could be expressed by the 7.6 ± 0.2mg.l− 1, and 8.5 ± 0.2, respectively, and ammo-
following formula: Y = − 4.7014 t2 + 30.827 t + 40.182 (r2 = 0.9104,
P b 0.01) (Y: fertilization rate; t: water temperature). Data in figure were
nia, nitrite, and nitrate concentrations are all less than
Arithmetic mean and Sample standard deviation (mean ± S.D.) 0.05 mg/l, the 28 °C is optimum for gonad development,
(P b 0.01). fecundity and spawning, fertilization, hatching, and
710 Q. Lin et al. / Aquaculture 254 (2006) 701–713

survival of the juveniles. In comparison, the examina- GSI differ significantly at different temperatures
tions over 107 days of the juveniles in H. whitei (P b 0.01), and the duration between stage IV and
displayed that several body condition indices fell with stage V is larger than that between other stages in the
increasing temperature and the results also indicated that same temperature (P b 0.05) (e.g., 28 °C). Coinciden-
the optimal temperature for reproduction was 20 °C tally, the GSI also varies greatly in this duration (Table
because the gonadosomatic index (GSI) peaked at 20 °C 1). Besides temperature, the nutritional requirements of
(Wong and Benzie, 2003). The reason for the great seahorse can also affect the GSI, and this finding is
difference between our results and the latter study may widely displayed in other fish (Watanabe et al., 1984;
be related to the natural ranges of these seahorse species Pickova et al., 1997; Santiago and Reyes, 1993).
in the wild. H. Whitei is a sub-tropical species from the However, more research is needed to understand
east coast of Australia whereas H. kuda is a tropical between food nutrition and condition indices in
species with a wide distribution throughout the seahorse H. kuda.
IndoPacific (Foster and Vincent, 2004). However, both When the seahorse develops to stage V (Table 3) in
of them illuminate that there must be an optimal all the treatments, the male's brood pouch swells and the
temperature for gonad development, fertilization effi- female's ovary also matures one after another. Few days
ciency, and survival rate of the juveniles in seahorse, and later, the female will find a chance to finish transferring
this temperature is consistent with the metabolic rates of the eggs to the male's brood punch, which is an
the organisms (Masonjones, 2001). interesting behavior (Vincent and Sadler, 1995; Mason-
Duration of gonad development in this experiment jores and Sara, 2000). During all the developing time,
differs significantly at temperatures ranging from 18 °C the water quality parameters and diets are the same
to 32 °C (F7, 32 = 154.065, P b 0.01). Water temperature except for the water temperature. The data and the
plays an important role in gonad development and relative analysis of this exeperiment all display that the
results in the failure of development at lower tempera- processes of gonad maturation are different in diverse
tures (18 and 20 °C), which only reach stage II and stage treatments, i.e., temperature, cannot only influence the
III, respectively. According to the present observation, development of the brood pouch, but also the eggs'
high temperature (30 and 32 °C) can also prolong the development. Our study points out that fecundity
developmental time of gonads, which is similar to that number and spawning number vary significantly in
of abalone Haliotis diversicolor Reeve (Lu et al., 2004), different temperatures (F5, 12 = 29.777, P b 0.01; F5, 12 =
turtle Chelydra serpentina (Rhen and Lang, 1999), and 76.578, P b 0.01), and the spawning rate is 62.07% at
crayfish Pacifastacus leniusculus (Perez et al., 2003). 32 °C, which is very low in contrast with other
Fig. 2 shows a wide variation in the gonad development treatments.
time, which relates to the species' diversity, for example, When the female transfers its eggs into the male's
the species' body size, feeding condition, behavior, and brood pouch, the brood pouch begins to play an
disease resistance, and above internal factors can affect important role in the subsequent reproductive behaviour
partly the development time. Besides above inner and ecology (Linton and Soloff, 1964; Gronell, 1984;
factors, the environmental condition should also be Vincent and Sadler, 1995; Jones and Avise, 1997; Jones
considered. Data in Table 3 can clearly prove that et al., 1998; Lunn and Hall, 1998; Kvarnemo et al.,
developmental speed is faster from 26 to 28°C because 2000; Foster and Vincent, 2004). In contrast, water
of shorter developmental times. And generally, the temperature also can greatly affect the reproductive
shorter time of gonad development, the shorter time the efficiency, and observation indicates that the fertilization
individuals are exposed to adverse environmental rates are significantly different at temperatures from 22
factors such as parasitic fungi or predation by copepods to 32 °C (F5, 24 = 53.675, P b 0.01). It indicates that 28
in the large-scale artificial breeding conditions (Lu et al., °C is the optimal temperature for fertilization because of
2004). the highest fertilization rate (92.41 ± 3.57%), and the
Metabolic rate is correlative with water temperature lower or higher temperatures lead to a poor fertilization
(Masonjones, 2001), so seahorses cultured at different rate (Fig. 5). In addition, throughout this experiment, the
temperatures may have diverse metabolic rates, as lower (22, 24 °C) and higher (30, 32 °C) temperatures
reported for other fish (Porter, 2001; Brodeur et al., also prolong the duration of hatching, which is beyond
2001). The variance in condition indices such as GSI in our assumptions (It's an evaluative error or else we'll
seahorses may be due to the diverse metabolic rates at design an experiment to test the duration of pregnancy
different temperatures, which is also proved in our of the male). Perante et al. (2002) indicated that the
present study. The observation demonstrates that the duration before the male released the young depended
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 711

100
on species and water temperature, generally this
duration ranged from 10 days to 6 weeks. This report 90
agrees in part with our observations. Generally, the

Hatching rate (%)

80
fluctuation in the duration of the male's pregnancy is
about 25 days in H. kuda (Foster and Vincent, 2004), 70
which is consistent with the report of the wild 60
Hippocampus whitei (Vincent and Sadler, 1995). This
study indicates that water temperature can strongly 50
affect the hatching rate, especially when temperature is 40
high, as the hatching rate declines dramatically. The
reason why this phenomenon happens may be explained 30
22 24 26 28 30 32
from the point of view of ecological physiology, and Temperature (˚C)
high temperatures lead to a higher metabolic rate just
like other fish (Il-Nam et al., 1995; Brodeur et al., 2001), Fig. 7. The relationship between Hatching rate (%) of juveniles, H.
kuda Bleeker, and temperature (°C), which could be expressed by the
and the drop in condition indices may be due to a greater
following formula: Y = − 4.5529t2 + 26.824t + 53.89 (r2 = 0.9377,
metabolic rate, with less fat reserves created at higher P b 0.01) (Y: hatching rate; t: water temperature). Data in figure were
temperatures (Discussed in Wong and Benzie, 2003). Arithmetic mean and Sample standard deviation (mean ± S.D.)
During large-scale artificial breeding and reproduction (P b 0.01).
in our seahorse center (Guangdong Province), the
hatching rate is customarily very low during the segregation, inbreeding, salinity, and stocking density
summer, which is partly attributed to the high water may also partly affect the developmental indices and
temperature (often N 30 °C). reproduction, so there will be more research required
To date, survival rate of the juvenile seahorse has into improving the gonad development and reproduc-
been widely studied by researchers (Scarratt, 1995; tive strategy for H. kuda, especially at the problem of
Forteeth, 1996; Payne et al., 1998; Payne and seahorse generation and genetic effects.
Rippingale, 2000; Adams et al., 2001; Woods, In conclusion, this investigation has demonstrated
2000a, 2003a,b; Chang and Southgate, 2001). In this that water temperature can significantly affect the gonad
investigation, the survival rate is also studied. development and reproductive efficiency in artificial
However, compared with the above research, the breeding conditions. The optimal temperature for the
goal of designing this experiment is to examine the condition indices of development is 26 to 28 °C and this
effect of the early development at each temperature on proved feasible in large-scale artificial breeding and
the survival rate. Considering above goal, it is not reproduction of H. kuda. We achieved mass production
sound to set too many days (N10 days ) to study, of juveniles of H. kuda (about 2.50 million individuals)
because in the hatchery the juveniles die mainly from in 2001, however, only a few (about 0.23 million) could
the external factors, such as the water quality be reared to commercial size because of the low survival
parameters, diets and breeding conditions will signif- rate. Therefore more research on juvenile survival is
icantly affect the survival rate of juvenile seahorses in needed.
8–10 days later after hatching, especially in transition
dietary period (10 days post hatch) (Lu and Chen, in Acknowledgements
press). Therefore, 10 days was chosen as the sampling
time in this experiment. The finding indicates that We thank Professor Xiao-lin Chen (Xiamen Univer-
temperatures ranging from 24 to 28 °C are the sity. China) for his commment on the manuscript. We
optimum temperatures for the survival of juveniles also thank Dr.Chang-xin Ma (Tsinghua University.
of H. kuda, and the highest survival rate is 90.74 China) and Professor Ting-bao Yang (Sun Yat-Sen
± 1.40% at 28 °C, which is beyond our expectation. University) for amending the manuscript. At last, we
Before this investigation, we forecasted that the would like to thank the staff of the Zhongshan university-
indices such as gonad development, fertilization and Yi dazhou aquatics culture company for their unselfish
hatch at different treatments could influence in part the help in near 4 years. This study was funded by the
survival rate of the newborn juveniles. However, the National High Technology Research and Development
compared result proves that the difference is small Program of China (863 Program) (2002AA624010)
(Fig. 7). Therefore, besides water temperature, other and Key Science and Technology Program of
parameters such as nutrition composition, gender Guangdong Province of China (2003A3050402) and
712 Q. Lin et al. / Aquaculture 254 (2006) 701–713

Science and Technology Program of Guangzhou city Il-Nam, K., Yong-Jin, C., Joon-Yeong, K., 1995. The pattern of
(2003Z3-C7321). oxygen consumption in six species of marine fish. Bull. Korean
Fish. Soc. 28, 373–381.
Izquierdo, M.S., Fernandez-Palacios, Tacon, A.G.J., 2001. Effect of
References broodstock nutrition on reprodutive performance of fish. Aqua-
culture 197, 25–42.
Adams, M.B., Powell, M.D., Purser, G.J., 2001. Effect of acute and James, P., Woods, C.M.C., 2001. Rearing seahorses: does temperature
chronic ammonia and growth of juvenile big bellied seahorse. J. matter? Aquac. Update 28, 9–10.
Fish Biol. 58, 848–860. Job, S.D., Do, H.H., Hall, H.J., 2002. Culturing the oceanic seahorse,
Brodeur, J.C., Dixon, D.G., Mckinley, R.S., 2001. Assessment of Hippocampus Kuda. Aquaculture 214, 333–341.
cardiac output as a predictor of metabolic rate in rainbow trout. J. Jones, A.G., Avise, J.C., 1997. Microsatellite analysis of materuity
Fish Biol. 58, 439–452. and mating system in the Gulf pipefish Syngnathus scovelli, a
Carcupino, M., Baldacci, A., Mazzini, M., Franzoi, D., 1997. species with male pregnancy and sex-role reversal. Mol. Ecol. 6,
Morphological organization of the male brood pounch epithelium 203–213.
of Syngnathus abaster Risso (Teleostea: Syngnathidae): before, Jones, A.G., Kvarnemo, C., Moore, G.I., Simmons, L.W., Avise, J.C.,
during and after egg incubation. Tissue Cell 29, 21–30. 1998. Microsatellite evidence for monogamy and sex-biased
Carcupino, M., Baldacci, A., Mazzini, M., Franzoi, D., 2002. recombination in the Western Australian seahorse, Hippocampus
Functional significance of the male brood pouch in the angustus. Mol. Ecol. 7, 1497–1505.
reproductive strategies of pipefishes and seahorses:a morpholog- Kvarnemo, C., Moore, G.I., Jones, A.G., Nelson, W.S., Avise, J.C.,
ical and ultrastructural comparative study on three anatomically 2000. Monogamous pair bonds and mate switching in the Western
different pouches. J. Fish Biol. 61, 1465–1480. Australian seahorse, Hippocampus subelongatus. J. Evol. Biol. 13,
Chang, M., Southgate, S.E., 2001. Effects of varying dietary fatty acid 882–888.
composition on growth and survival of seahorse, Hippocampus Lambert, Y., Dutil, J.-D., 1997. Condition and energy reserves of
spp., juveniles. Aquar. Sci. Conserv. 3, 205–214. Atlantic cod (Gadus morhua) during the collapse of the northern
Chen, Q.X., Lu, J.Y., Sheng, J.Q., in press. Effect of diet on the Gulf of St. Lawrence stock. Can. J. Fish Aquat. Sci. 54,
fecundity of Pseudodiaptomus annandalei Sewell, 1919. J. Trop. 2388–2400.
Oceanogr. Linton, J.R., Soloff, B.L., 1964. The physiology of the brood pouch of
CITES, 2001. Notification no.2001/034. Notification to the Parties Con- the male seahorse, Hippocampus erectus. Bull. Mar. Sci. Gulf
cerning: Seahorse and Other Members of the Family Syngnathidae. Caribb. 14, 45–61.
Comeau, L.A., Campana, S.E., Hanson, J.M., Chouinard, G.A., 2000. Liu, J.K., He, B.W., 1992. Cultivation of the Chinese Freshwater
Seasonal changes of thyriod hormones in field-collected Atlantic Fishes, Third edition. Sci. Press, pp. 65–72 (In Chinese).
cod in relation to condition indices, water temperature and Lockyear, J., Kaiser, H., Hecht, T., 1997. Studies on the captive
photoperiod. J. Fish Biol. 57, 571–588. breeding of the Knysna seahorse, Hippocampus capensis. Aquat.
Correa, M., Chung, K.S., Manrique, R., 1989. Cultive experimental Sci. Conserv. 1, 129–136.
del caballito de mar, Hippocapus erectus. Bol. Inst. Oceanogr. Lourie, S.A., Vincent, A.C.J., Hall, H.J., 1999. Seahorse: an
Venez. 28 (1–2), 191–196. Identification Guide to the World's Species and Their Conserva-
Fernandez-Palacios, H., Izquierdo, M.S., Robaina, L., Valencia, Salha, tion. Project Seahorse, London. 214 pp.
M., Montero, D., 1997. The effect of dietary protein and lipid from Lu, J.Y., Wu, J.Y., Yang, D.W., 2001. Growth rate of Hippocampus
squid and fish meals on egg quality of broodstock for gilhead kuda Bleeker under intensive culture. J. Fish. China 8 (1), 59–63
seabream (Sparus auratus). Aquaculture 148, 233–246. (In Chinese).
Forteeth, N., 1996. Seahorses, Hippocampus abdominalis in culture. Lu, J.Y., Li, B.J., Sun, Y.Y., 2002. The ingestion,growth and ecological
Aust. Aquac. 9, 83–84. conversion efficiency of Hippocampus kuda under the intensive
Forteeth, N., 1997. The large bellied seahorse, Hippocampus rearing conditions. J. Fish. China 26 (1), 61–66 (In Chinese).
abdominalis, a candidate for aquaculture. Aust. Aquac. 11, 52–54. Lu, J.Y., Lin, Q., Sun, Y.Y., Sheng, J.Q., Chen, Q.X., 2004. Effect of
Foster, S.J., Vincent, A.C.J., 2004. Life history and ecology of temperature on the early development of Haliotis Diversicolor
seahorses: implications for conservation and management. J. Fish Reeve. J. Shellfish Res. 23, 963–966.
Biol. 65, 1–61. Lunn, K.E, Hall, H.J., 1998. Breeding and management of seahorse in
Giwojna, P., Giwojna, B., 1999. Seahorse breeding secret: Part I, Ten aquria. Briefing Documents for the First International Aquarium
common mistakes and how to avoid them. Freshw. Aquac. 8–31 workshop on seahorse Husbandry, Management and Conservation.
(January). Project seahorse, Chicago. 48 pp.
Grant, S.M., Brown, J.A., 1999. Variation in condition of coastal Marichamy, R., Liption, A.P., Ganapathy, A., Ramalingam, J.R., 1993.
Newfoundland 0-group Atlantic cod (Gadus morhua): field and Large scale exploitation of seahorse (Hippocampus kuda) along
laboratory studies using simple condition indices. Mar. Biol. 133, the palk Bay coast of Tamil Nadu. Central Mar. Fish. Res. Inf. Ser.
611–620. 119, 17–21.
Gronell, A.M., 1984. Courtship, spawning and social organisation of Masonjones, H.D., 1997, Sexual selection in the dwarf seahorse,
the pipefish, Corythoichthys intestinalis (Pisces: Syngnathidae) Hippocampus zosterae (Syngnathidae): An investigation into the
with notes on two congeneric species. Z. Tierpsychol. 65, 1–24. mechanisms determining the degree of male vs.female intrasexual
Hilomen-Garcia, G., 1999. AQD's marine ornamental fish project. competition and intersexual choice, PhD thesis, Tufts University,
SEAFDEL Asian Aquac. 21 (2), 31–38. U.S.A.
Hilomen-Garcia, G.V., Delos Reyes, R., Garcia, C.M.H., 2003. Masonjones, H.D., 2001. The effect of social context and reproductive
Tolerance of seahorse Hippocampus kuda (Bleeker) juveniles to status on the metabolic rates of dwarf seahorse (Hippocampus
various salinities. J. Appl. Ichthyol. 19, 94–98. zosterae). Comp. Biochem. Physiol., Part A 129, 541–555.
 Q. Lin et al. / Aquaculture 254 (2006) 701–713 713

Masonjores, H.D., Sara, M., 2000. Differences in potential reproduc- Texeira, R.L., Musick, J.A., 2001. Reproduction and food habits of the
tive rates of male and female seahorses relates to courtship roles. lined seahorse, Hippocampus erectus (Teleostei: Syngnathidae) of
Anim. Behav. 59, 11–20. Chesapeake Bay, Virginia. Rev. Bras. Biol. 61 (1), 79–90.
Payne, M.F., Rippingale, R.J., 2000. Rearing West Australian Vincent, A.C.J., 1995. Exploitation of seahorses and pipefishes.
Seahorse, Hippocampus subelongatus, juveniles on copepod NAGA. ICLARM 18–19 (January).
navplii and enriched Artemia. Aquaculture 188, 353–361. Vincent, A.C.J., 1996. The International Trade in Seahorse. TRAFFIC
Payne, M.F., Rippingale, R.J., Longmore, R.B., 1998. Growth and International, Cambridge. 4 pp.
survival juvenile pipefish (Stigmatopora argus) fed live copepods Vincent, A.C.J., Sadler, L.M., 1995. Faithful pair bonds in wild
with high and low HUFA content. Aquaculture 167, 237–245. seahorses, Hippocampus whitei. Anim. Behav. 50, 1557–1569.
Perante, N.C., Pajaro, M.G., Meeuwig, J.J., Vincent, A.C.J., 2002. Watanabe, T., Arakawa, T., Kitajima, C., Fujia, S., 1984. Effect of
Biology of a seahorse species, Hippocampus comes in the central nutritional quality of broodstock diets on reproduction of red
Philippines. J. Fish Biol. 60, 821–837. seabream. Nippon Suisan Gakkaishi 50 (3), 495–501.
Perez, J.R., Celada, J.D., Gonzalez, J., Carral, J.M., Saez-Royuela, M., Wong, J.M., Benzie, J.A.H., 2003. The effects of temperature, Artemia
Fernandez, R., 2003. Duration of egg storage at different enrichment, stocking density and light on the growth of juvenile
temperatures in the astacid caryfish Pacifastacus leniusculus: seahorses, Hippocampus whitei (Bleeker, 1855), from Australia.
critical embryonic phase. Aquaculture 219, 347–354. Aquaculture 228, 107–121.
Pickova, J., Dutta, P.C., Larsson, P.O., Kiessling, A., 1997. Early Woods, C.M.C., 2000a. Improving initial survival in cultured
embryonic cleavage pattern,hatching success and egg-lipid fatty Seahorse, Hippocampus abdominalis leeson, 1827 (Teleostei:
acid composition:comparison between two cod stocks. Can. J. Syngathidae). Aquaculture 190, 377–388.
Fish. Aquat. Sci. 54, 2410–2416. Woods, C.M.C., 2000b. Preliminary observations on breeding and
Poortenaar, C.W., Woods, C.M.C., James, P.J., Giambartolomei, F.M., rearing the seahorse, Hippocampus abdominalis (Teleostei:
Lokman, P.M., 2004. Reproductive biology of female big-bellied Syngnathidae) in captivity. N.Z. J. Mar. Freshw. Res. 34, 475–485.
seahorses. J. Fish Biol. 64, 717–725. Woods, C.M.C., 2003a. Effect of stocking density and gender
Porter, S.M., 2001. Effects of size and light on respiration and activity segregation in the seahorse, Hippocampus abdominalis. Aquacul-
of walleye Pollock (Theragra chalcogramma) larvar. J. Exp. Mar. ture 218, 167–176.
Biol. Ecol. 256, 253–265. Woods, C.M.C., 2003b. Growth and survival of juvenile seahorse,
Rhen, H., Lang, J.W., 1999. Temperature during embryonic and Hippocampus abdominalis reared on live, frozen and artificial
juvenile development influences growth in hatching snapping foods. Aquaculture 220, 287–298.
turtlesm, Chelydra serpentina. J. Therm. Biol. 24, 33–41. Wooster, G.A., Bowser, P.R., 2000. Remediation of Cayuga Syndrome
Santiago, C.B., Reyes, O.S., 1993. Effect of dietary lipid source on in land locked Atlantic salmon Salmo salar using egg and sac-fry
reproductive performance and tissue lipid levels of Nile tilapia bath treatments of thiamin-hydrochloride. World Aquac. Soc. 31,
Oreochromis niloticus (Linnacus) broodstock. J. Appl. Ichthyol. 9, 149–157.
33–40. Zar, J.H., 1996. Biostatistical Analysis. Prentice Hall, Upper Saddle
Scarratt, A.M., 1995. Techniques for raising lined seahorses River, New Jersey, U.S.A. 481 pp.
(Hippocapus erectus). Aquar. Front. 3, 24–29.
Selman, K., Wallace, R.A., Player, D., 1991. Ovary of the seahorse,
Hippocampus erectus. J. Morphol. 209, 285–304.