5.3士1.5"
NS
*
と一響匁N冒'昌W巨一如てnル旨冨亘且一WN四式0昌跨ぷ貫ミミ言令雙一堅券団鴫鴬識S昇嚇
器
DF N班N' NO︑司,
Swimming speed representing the swimming activity of three different photoperiods (24L,12L:12D and 24D) manipulated undert1Ⅱee diffeNm feedmg schedules(DNF, DF and トJF) were presented in Fig.3.1 respectively. The white and black bar represents the natural Photoperiod ofdaytime (07:00‑19:0の and nigh伽me a9:00‑070の. The figures with grey and White area representsthe time thatthe larvae were fed and unfed respectively. For DNF (Fig 3.1a), there were 2 Significantly u)<0.05) higher peaks of activity at 12:oo and 20:oo for a11 Photoperiods. A110ther peak of higher activity at 04:0o could be identi丘ed fbr 24L and
12L:12D in DNF treatmentifthe activity for 24 h cyde showed similar continuoustrend. For DF (F喰.3.1b), adivity auctuated with a decrease ftom 08:oo t0 18:oo for aⅡ Photoperiods The activity slightly increased 丘om 18:oo t0 20:oo but decreased again until 04:00. The range of activity for daytime and nigh札ime irrespective of photopedod were 220‑4.06 and
2.03‑3.62 Cm sec、1 respectively. For night feeding (Fig.3.1C), activity increased for 24L and
12L:12D but decreased for 24D 丘om 08:oo t0 18:00. The activity slightly increased 丘om 1&oo t020:oo for a11 Photopedods. However,the activity 員lrther increased for 24L, remained Constant for 12L:12D and decreased for 24D. The range ofactivity for daytime and nighttime
irrespective ofphotopedod were l.32‑3.42 and 3.14‑4.65 Cm sec'1 respectively
近大水研報 14号 (2014)
ArifahRa1如ahBintisyedMuha1血ad : Bagridcat6Shuyst伽11e抗U地S における仔魚飼育技術の改善
a) 5
4
^24L
3
2
→コー12L:12D
Day and Night Feeding
b)
、 269 ・ 4
^24D
3
b (24D) (24L)
2
^
DayFeedi11g
(24L) (12Lt12D)
C)
4:00
Cd
4
b (12L桧D) bc (24D)
3
(24り
2
bc (24L) bc (24D)
ー'ー'‑1
(12L 12D)
Fi容.3.1 Swimming speed fbrt11ree photopedods; 24L,12L12D and 24D manipulated under three different feeding schedules a) day and night feeding (DNF), b) day feeding (DF) and c) night feeding (NF). Different subscripts within the same line indicate significant difference
ψく0'05, one、way ANOVA). white and black bar represents natural photopedod ofdaytime
and nighttime respectively. Grey area representsthe time thatthe laNae were fed0
Ni帥tFeed註lg
&00
b(12L柁の
(24り d (24の
12:00
12L:12D) d (24D)
1&00 20:00 Hours
b
0:00
(一.oom 婁口)でご身⑳三ΞΞ冨め 0 50 5
3.4
Survivalwas significantly improved for DNF compared W北h DF but was insignificant With NF for 6,10 and 14 dA11. Lower survival for DF and NF (Table 3.D correlated with higher cannibalism at 14 dA11 due t010w feed intake similarly reported in African catfish (Hecht and pien肌r 1993; AI・Hafedh and Ali 2004), Japanese aounder paralichthys Ohvace1ιS (Dou et al.2000) and Atlantic cod Gadus "10rh解α(Folkvord 199D demonstrating that significantly higher feed intake in DNF in this study can reduce the cannibalism. The results proved that ba8rid catfish shows noctumal characteristics. However, feed should not Only be provided during the nighttime or only dudng the daytime as conducted in the Common aquaculture practice.1hisis because they showed higher activity orfヒeding demand both dudng the daytime and nighttime (Fig.3.D. From the results of swimmin号 Speed, although laNae were fed 6 times daily fbr DNF (Fig.3.1a), they stiⅡ Show increased activity at 12:00 (indicates daytime) and 20:00 (indicates nighttime).1n additiotL increased activity Was also obseNed during the daytime when the laNae were fed (grey area of Fig.3.1b), Where the range ofactivity was similarto the nighttime. This means that although the laNae Showed noctumal charactedsticS 工Vith increased activity at ni8ht (Fig.3.1C), provision offヒed during the daytime was not rejected by the laNae, hence increasing the activity. Feeding demand both during 血e daytime and nighttime was also reported by Alnomsakun et al a998の Where 入10加α intake in bagrid catfish lalvae was Mgh during the nighttime but was not signi丘Cant dudng the daytime. This could be related to the high metabolism and digestion Where previous study indicated that gastdc emptying time for 2・15 dA11ba8rid catfish larvae to be between 211r 30 min and 4 h lo min (Alnomsakun et al.1998b). Lalvae at 12 and 15 dA11 Showed less than twice faster gastdc emptying time compared to earlier life stages indicatin今 increased digestive ability (Atrlomsakun et al.1998b) therefore requiring higher
feed intake in shorter inteNals. This coincided weⅡ With increased cannibalism as the larvae Discussion
近大水研報 14号 ( 20 1 4 )
aged in this study where si3ni負Cant effects of feeding schedules on the cannibalism were Prominent at lo and 14 dA11 but not at 6 dAI{. Therefore, optimum feeding schedule should Considerthe periods ofhigh feeding demand ofdifferentlife stages
Nthough feed intake was significantly higher in DNF compared to DF and NF, the AFCR and sGR were insigni負Cant among a11 feeding schedules. This indicates that better feed conversion was obtained for DF and NF compared to DNF despite their low feed intake In other words, consumption ofits conspecifics due to higher cannibalism in DF and NF may have contdbuted to its growth performances referred to as Type l calmibalism (Hecht and Appelbaum 1988; Katavic et al.1989; Baras et al.2003). Type l cannibalism was indicated by the consumption of its conspecificS 丘'om the postedor body part but discarding the head Off. LaNae that perfonn cannibalism due t010w feed intake had gained growth advantage Over the others resulting in increase of size variation with higher coe伍Cients of variation reported in the present study similarly observed in underfed juvenile A丘'ican catfish (1U・
Hafedh and Ali 2004).1n this study, NF resulted in significantly higher coe錨Cient of Variation compared to DF. similar results were reported in catflsh Heter0みrα11Chus l0πglylhs (Kerchuen and Legendre 199D. Being a noctumal species, bagrid catfish laNae displayed higher activity during the nighttime compared to daytime especia11y when they were fed (Fig 3.1C).壬五gher activity leads to h礁her tendency of feeding competition that may result in higher differences of individual feed intake, hence increasing the size variation. This is because, in a population especia11y for calmibalistic species, certain individual may be more active or bolder than the others, causing t11reat, therefore suppressing the fbraging activity of Iess active or shy individuals (Dowling and Godin 2002; Brown and Braithwaite 2004) DiHerences in individual feed intake were also reported in sole solea solea where individua11y held 負Sh showed differences of799/o feed intake which significantly affected the Size variation (Nlas、Munoz et al.20ID. However, dudng the daytime when laNal bagrid
AtifahRahmahB血tisyedMuha1血ad : Ba部idcat丘Sh珍St加πe"1U地S における仔魚飼育技術の改善
・ 271
Catash were fed in this study (Fig.3.2b), activity was lower than NF, therefore leading to reduced t11reat and lowertendency off註eding competition 丘om bolderindividuals. As a resU丘,
10叉Ner s!ze varlatlon 叉Vas obser,1ed ln thls stud訂 for l'、F conl,、ared t0 卜JF
Iuthough the overa11 Coe伍Cient of variation was below lo%(3.8‑5.39る) for a11 treatments indicative of non‑heterogeneity (Jobling and Baardvik 1994), the present study 工Vas conducted at laNal stage unlike the study by Jobling and Baardvik (1994) which was Conducted atjuvenile stage.1fthe culture ofbagrid cat丘Sh in the present study is continued, emergence of Type 11 Cannibalism with increasing size variation (Hecht and Appelbaum 1988; Baras et al.2003) may occur. Type 11 Cannibalism enables the sma11er 丘Sh to be Consumed completely and 61rther enhance the size variation and social dominance, thus Ieading to higher cannibalism and lower suNival similarly reponed in sea bass Diceπlrach1ιS 1αみraz l".(Katavik et al.1989). From this study, higher f註ed intake for DNF did not ensure e伍Cient feed conversion but showed signi丘Cant improvement in the total body length and 丘nalweight compared to DF and NF atthe end ofthe experiment.1his was also reported in A負'ican catfiS11 Where increasing feeding leve16'om 4 t0 6,8 and lo% of the body weight Signiacantly increased feed conversion ratio (feed weighvfish weighり but also significandy improved growth (AJ・Hafedh and Ali2004). Lower growth forDF andNF inthisstudy could Probably be due to tbe increased stresS 丘'om higher cannibalism and aggression behaviour (Hengsawat et al.1997; Hecht and uyS 1997; Nma.an・Rueda et al.2004; Toko et al.2007) as also obseNed in A丘ican catfish (Britz and pienaar 1992).1'王igher total activity ofDF and IqF compared to DNF in this study (Fig.3.D probably in search of food could also lead to 10wer growth similarly reported in banamundi ιαtes calcarlf'er (Barlow et al.1995). These are associated 工Nith increased energy consumption.
Surpdsingly in this experime址, there were no signi丘Cant effects of different Photopedods on the survival, cannibalism and growth performances ofbagrid cat丘Sh larvae
近大水研報 14号 (2014)
although they showed nocturnal characteristics. This indicates that tMs fish have natural noctumal rhythm where the activity 工Vas not affected by the availability of light even during the nighttime. However, it could be assumed that dark condition increased the activity ofthe IaNae hence improving the feed intake and final weight as obseNed in this study although not significant.1he insignificance of photopedod in this study was probably because the IaNae were already accustomed to the light condition during the experiment which lasted for
13 days (2‑14 dAfl). The behaviour of light avoidance was only conducted for lo seC 丑'om the introduction of light 工入lhich showed sudden startled movement in chapter 2.1n addition, IaNae tend to aggregate dudng the light phase probably demonstrating their natural mechanism of light avoidance related to shelter seeking behaviour as also obseNed in 負ngerling channel catfish lctalur1ιSP記11Ctatus (Brown et al.1970) and A6'ican catfish (Britz and pienaar 1992). Regarding this, the aggregation during the daytime is considered natural to catfishes including bagrid cat負Sh, therefore not a任ecting the survival, cannibalism and growth performances in this study atthe larvalstage
As a condusion, bagrid catfish can be reared both in light and dark condition without Significant effects on the survival, cannibalism and growth performances but showed higher activity in dark condition at larval stage. However, care6.11 Consideration should be stressed On the pr010nged effects of light in the early juvenile stage which may negatively affect Survival and gr0眺h performances. Availability offeed b0血 during daytime and nighttime is necessary to improve suNival and growth performances but not total eHmination of Calmibalism similarly reviewed by Hecht and pienaar (1993) in various cannibalistic species Therefore, the light avoidance and aggregation during the daytime in laNal bagrid catfish may be related to the shelter seeking behaviour of noctumal characteristics. since the aggregation 工N'as obseNed at the b0犹om of the tank representing benthic behaviour, suitable StocMng density may also affect the distdbution and distance among the larvae that may
ArifahRa1血ahBintisyedMuha1血ad : Bagridcat丘Shuyst加πe"1UruS における仔魚飼育技術の改善
・ 273
tdgger cannibalism. Hence, the manipulation of stocMng density and shelter investigated in the next chapter
近大水研報 14万 ( 20 14 )
叉Vere
kifahRa1ⅡnahB血tisyedMuha1血ad : Bagridcat丘Shuyst紹πe抗υruS における仔魚飼育技術の改善
ManゆUlation ofstocking densiw and shelterto improve suNivaland growth Performances oflarvalba晉rid catnsh
4.1 Introduction
Avoidance of light and aggregation were obseNed during the daytime (chapter 2), but photopedod did not significantly affect survival, cannibalism, growth performances and Swimming speed of laNal bagrid catfish (chapter 3). However,in dark condition or dudng nighttime, more active swimming was observed (chapter 3). This may suggestthatthe light avoidance and aggfegation during the daytime is related to the shelter seeking behaviour of noctumal characteristics ofbagrid cat6Sh. This behaviour was in agreement with Khan et al (1988) who reported that bagrid catfish feed 丘'om dusk to dawn mostly atthe bottom parts of the water body. During daytime, they are less active and remain 加 hideouts. since aggregation was observed at the bottom of the tarlk, over stocked laNae concentrated at the
aggregated area m地ht lead t0 11igher contact between individual. This may tdgger
Cannibalism and affect survival and growth performances. F0110wing these, unsuitable StocMng density and unavailability of shelter may tdgger cannibalism in laNal bagrid catfish and should be investigatedStocMng density has been one ofthe main concerns related to fish welfare in intensive
Culture system (Kiistiansen et al.2004; van de Nieuwegiessen et al.2009; uglem et al.2009)
However, its ea'ects vary depending on behaviour and physi010gical requirement of theCultured species (van de Nieuwegiessen et al.2009).1n general, higher density impair larval
SUNival and growth as reported in rainbow trout Sα11110 gαかdπeri (Boujard et al.2002), matrinxa BI),C011 Cephα11ιS (Gomes et al.2000), asP 4δPiuS αδPi1ιS, ide ιeuciscus idus, chubChapter 4
275 ・
ιe1ιCiscus cephalus (Kupren et al.20ID, grass carp cteπOphal),πgod011ide11α(sharma and Chahaba此i 1998) and sand bass parα1αみrια万1αCU1αtqf'asciams (Alvarez・G0地会lez et al and coln‑etltlon for food and space (1王ecnt ana
?001、. Th!s resu!ted 6'on11ncreased 0即上、゛000+,.^60
Pienaar 1993). However, higher density bene丘ted species like African cat{ish (Hossain et al 1998) and common carp c)prin1ιS calpio (Hecht and pienaar 1993). For such species,10W density may trigger terdtorial and cannibalistic behaviour. Therefore, increasing the density resulted in diminishing of social dominance and reduction of cannibalism (Kaiser et al 1995a,b; Almazan・Rueda et al.2004), hence improving survival(Hecht and pienaar 1993) Effects of stocMng density is not only restricted between species but also within the same Species related to different life stages.1ncreased density in African catflsh larvae decreased growth performance and increased cannibalism (Hecht and Appelbaum 198& Hossain et al
1998). However, higher density injuvenile A6'ican cat丘Sh increased the 8rowth performance and decreased the aggression (Almazan Rueda 2004; Hecht and uyS 1997; Kaiser et al
199Sa,b).1n larval common barbe1βUrhoi h1ιrhoi, growth was not a任ected by density of20‑
20o laNae が(zarski et al.20ID.1n bagrid catflsh, a t11reshold capacity ofjuvenile stage (mean weight 20.45 g) was reported to be o.29 釦d o.38 負Sh l'1 Where reduction of growth
Was observed beyond the optimum density (Khan 1994). However, the optimal density for IaNalstage, which has high cannibalism‑related m0犹ality, remains unknown
In addition to optimizing stocMng density, provision of shelter may also reduce Cannibalism. As shelter increase available surface area and realge, it could increase the laNal StocMng t11reshold and maximize production. shelter is 0丑en used for aquaculture in benthic 丘Sh such as burbotιotα10ta (叉Nocher et al.201り and cNstacea11S including cray丘Sh, cherαχ deSか1ιCior (verhoefand Austin 1999) and westem rock lobsterPαπUlir1ιS cy8πUS σohnston et al.2006).1t is also used for cU1加ring species displaying noctumal and cannibalistic behaviour like African catfish (Hossain et al.1998). However, the use ofshelter was reported
近大水研報 14号 (2014)
Adfah Ra111nahBintisyedMuhan血ad : Bagrid catfishuyst加 11e"1Urus
Unbene負Cial in culturing sleepy cod 0り'ele0加S liπeoams lnger' g ,
n。ctuma1 叩ecies with no improvementin suNival and groM ( er
1995)
。fshelter was also reported to complicate obseNation and harvesting pr゜cess (゜
t。 invest璃ate the manipulation of stocMng density an
Therefore, this study alms
Of laNal bagrid
the suNival, cannibalism and groMh performances
Shelter availability on
Catfish larvae to develop optima11arviculture condition
Materials and Methods Preparation onarvae
1、、r[ificial fertilization was conducted in septem er
C。11ab。rative Research in Aquaculture, universiti Maaysia Malaysia according to the methodsin section 3・2・1
4.2
4.2.1
における仔魚飼育技術の改善
4.2ユ Experimentaldesign
the effects of 負Ve stocMng
A tw。̲factodal experiment was designed to investlgate
LaNalsuNival
and 50 larvae l、1) each either with or without shelter
densities (10,20,30,40
measured in each combination 丘om 2 t0 14 dA111aNae (13
and gr0航h perf0血ances were
I n。n̲recirculating experimentaltanks (30.O × 15.O × 18・5 Cm) were pNpare
days). Thirty 7
f。r each st。CMng density、shelter combin飢ion). The shelter(24・゜ X X
(tdplicates
C。nstNcted with plastic mesh with the opening of l.5 × 1・5 Cm (Fig・ 4')' ac a
Cm)was
WaS 611ed 工Nith 5 1 dicNorinated tapwater and one shelter was p ace
forthe sheltered treatments
20H at the center of Sabah、Kinki university,
、 27フ・
a)
近大水研報 14号
Fi留.4.1 Plastic mesh used as shelter for sheltered treatments; a)丘ont view, b) top view
( 20 14 )
At 24 1L、、H,1arvae were transferred into their respective experimental tanks with given stocMng densities. Atthe same time, the mean initialweight (0.003士0.oo g) assumed to
be equalwas measured using pooled sample of loo laNae 丘om the acclimatization tank. This
Was due to the difficulties in measuring individual weight of24 hA111arvae since the were
too sma11. A total of lo larvae were anesthetized with V20 diluted Transmore containin a̲methyl quinoline as active ingredient (Nika Trading, Malaysia) and measured for total body Iength. First feeding was provided with rotifer (βrachi0π1ιSplicatiljs sp. complex) at 30 hAH (chapter 2) at 20 individuals ml、1 then with 冱rte1ア1iajシα11CisCαπα naupliuS 行・om 2 dA11 (20 individual ml、1) unti1 6 dA11. Artificial feed (otohime B2 Containing approximately 50%
Protein and lo% 1ipid) was fed from 2 t0 14 dAH (chapter 3). The 負Sh were fed daily at 08:00,12:00,16:00,20:00,00:oo and 04:oo to apparent satiation. weight of a丘途Cial feed
Was measured and total amount ofgiven feed for each tank was calculated atthe end oftheexpenment
Uneaten feed and faeces were removed and 20% water exchange was conducted twice a day at 06:30 and 1830. Dead laNae 丘'om each tank were gently siphoned before each feeding time and observed immediately under the light microscope (Edipse E600, Nikon,
b)
Atifah Ra1血ahB血tisyedMuh釘血lad : Bagid catfishuysiuS11e"1Urus
Japan) at 4X magnification. Any la1刃ae with bitten marks on its body were considered as being attacked and categorized as cannibalism. At 6,10 and 14 dA11, ten randomly sampled
IaNae were then anaesthetized with ν20 diluted Transmore to measure the individualtotal b。dy length. Atthe same time, surviving laNae 丘om each 仇nk were counted to measure t e
Survival. The anesthetized larvae were a110wed to revive and returned to their respective ta s At the end ofthe experiment, a11 Sulviving larvae ftom each tank were pooled an
Weighed. SUNival (S) and cannibalism (C) were calculated as: S (96)=(final number of
IaNae/initial number of laNae)*10o and c (%)=(number of injured and disappeared Iarvae/initial number of laNae)*10o respectively. Growth pedormances; feed intake (FD,
apparent feed conversion ratio (AFCR), speci丘C groMh rate (SGR) and coe伍Cient of Variati。n for tota1 10dy lengh (CV) were determined using the formula: FI (g)= final feed Weight(g)‑ initial feed weight(g); AFCR = feed giV即(g)*weight 部in (g)、; SGR (血d、) (1n final weight ̲ 1n inmal weight)*days'1*10の and cv =(standard deviation*average total b。dy length、1)*10o respectively. Apparent feed conversion ratio (AFCR) was used instead of FCR in this study because uneaten feed were not measured due to the sma11 quantity an easily dissolved in the water. A11 results were presented for 14 dAH except for suNiva ,
Cannibalism and totalbody length which were presented for 6,10 and 14 dA11
における仔魚飼育技術の改善
4.2.3 Sねtisticalanalyses
Nl data were statistica11y analyzed using spss version 11.0. The effects of stocMng density and shelter and their interaction on suNival and growth pedormances were teste Using two̲way analysis ofvariance (ANOVA). Data with significant differences were 6・1rther C。mpared among 血e means using post HOC Tukey'S HSD test u)<0.05). The normality of
Variance was analyzed using shapiro・xvilk test
・ 279 ・
4.3
For two・way ANOVA analysis, shelter did not significandy affect the suNival U)>0.05). However, stocMng density signi丘Cantly affected survival at 6,10 and 14 dA11 ψく0.02上 Pく0.013;Pく0.004 respectively). post Hoc test fortwo・way ANOVA indicated that
the lowest density uo larvae rl) resulted in the lowest suNival for both sheltered and
Unsheltered treatments. survivalincreased as the density increased with higher value at 30and 40 lawae l、1but was S1礁htly lower at 20 and 50 lawae l'1 for a11 ages (6,10 and 14 dAH)
However, survival was more prominently reduced 丘om 6 t0 10 and 14 dA11 at 50 laNae l"1 (Table 4.リ
TWO・way ANOVA analysis showed that calmibalism was highest at the lowest
Stocking densiw (10 laNae l、1) for 6,10 and 14 dAH with insigni丘Cant e任ectS 丘om shelter at
a11 ages. post Hoc test for two・way ANOVA indicated that at 6 dA11, cannibalism was
insignificant u)>0.05) among aⅡ StocMng densities. However, the effect of stocMng density
became signi丘Cant at l0 ψく0.02D and 14 dA11 ψく0.0OD. At lo dAH, cannibalismdecreased at 20 laNae " onwards but increased again at 50 larvae l'1 At 14 dAH, Cannibalism was significantly lower for 20‑50 lalvae l'1 Compared t010 laNae l、1(Table 4. D
For two・way ANOVA analysis, stocMng density did not signiflcantly u)>0.05) affect
the total body length at 6 dA11 but significantly affectthe total body length as the larvae aged at lo u)<0.00の and 14 φく0.00の dA11. The total body length was si8ni負Cantly higher at 10wer density and decreased as the density increased for aⅡ ages. However, more prominent differences were observed between lo and 50 larvae が at lo and 14 dAH (Table 4.1)
For two・way ANOVA analysis, shelter also signi6Cantly improved feed intake u)<
0.006). post Hoc test fortwo・way ANOVA showed that feed in仇ke was significanuy higher
at lo laNae r and decreased as the density increased with insignificant value fbr 30‑50 Iarvae r . During feeding, the laNae in the sheltered treatments forage actively t11roughoutResults
近大水研報 14 万' ( 20 14 )
ArifahRalmlahBintisyedMuhammad : Ba即idcat負Shuyst加11e形川加における仔魚飼育技術の改善
the tank (vertical distribution) and fed dose to the mesh of the shelter. on the other hand, IaNae in the unsheltered tanks stayed at the b0札om and rarely forage in the water column (horizontal distribution). specifiC 号rowth rate and final weight were not significantly aHected by shelter Uフ>0.05) but 工入lere signi丘Candy affected by stocking density Uフ<0.ool andpく0.001
respectively). specific groMh rate and final weight were signi丘Cantly higher at lo larvae l、1 Compared t020‑50 larvae l、1. No significant effects ofshe此er and stocMng density on AFCR
and coe伍Cient ofvariation was obseNed u)>0.05)(Table 4.2). There was also no significant interaction between shelter and stocMng density in aⅡ Parameters (TableS 4.1 and 42)
・ 281
Table 4.1 Survival, cannibalism and totalbody length ofdifferent stocMng density and shelter for 6,10 and 14 dAH
Parameters Sheltered
10 20 30 40 50 Unsheltered
10 20 30 40 50
TWO・wayANOVA
Shelter
Stocking density 110
↑20 130 140 t50
Interaction
6 dA11
93.3士1.4 95.8士2.6 97.5士0.8 97.7士13
96.5士3.1
SUMval(%)
10 dA11
90.8士1.4 93.8士4,5 953士1.3 96.0士0.フ 95.8士3.8
91フ士5.8 97.1士1.9 96.3土13 97.1士2.5 977士1.8
14 dAI{
86.7土2.9 92.5士4.5 942士2.2 95.8士0.3 94.5士2.6
89.2士3.8 942士3.1 953士1.8 95.4士2.3 93.0土1.5
Values are mean士 SD (π=3)
*Pく0.05;tTukey'S HSD test(TWO・way ANOVA). val゛esfm Tukey'S HSD test(TWO・way ANOVA) are moon士 SD (π=の
Values in the column with different subscripts indicate significant difference u)<0.05) NS, non・significant
6dAH
NS
*
C廸libalism (%)
92.5士1.1b 96.5士0.9北
972士0.8゜
97.4士0.4゜
97.1土0.8゜
NS
2.5士0.0 1フ士0.フ 1.1士0.5 1.1士0.4 2.0士13
86フ士5̲2 92.1土47 93.1士2.フ 94.6士3.6 89.9士1.3
10dAH
NS
*
90.0士1.1b 94.0士03北
953士0.0' 95.7士0.4'
94.4士2.0北
NS
フ.5士2.5 4.2土1.9 3.0土0.5 2.3士0.3 33士2.1
2.5士2.5 2.1士07 1.4士0.5 1.5士13 3.2士13
14 dAH
NS
*
12.5土2.5 フ.1士3.8 5.0士1.フ 3.3士0.4 4.8士2.1
86.6土0.0' 92.3士0.3北
93.6土0.8゜
95.2士0.8'
922士3.3北
NS
6.7士3.8 4.2士2.6 3.3士17 3.3士2,4 63士1.4
6 dA11
NS NS 2.5士0.0
1.9士0.3 13士02 13士03
2.6士0.8 NS
T0仇lbody length (mm)
12.6士0.3 12.0士0.4 H.8士0.2 12.3士02 12,1士0.5
H.7士3.8 6.7士3.6 5.8士17 4.6士3.5 9.0士1.3
10 dAH
NS
*
17.9士03 177士0.5 16.9士0.5 16.9士0.6 16.9士02
フ.1士0.6"
4.2士0.0北 3.2士0.2b 2.8土0.7b 4.8土2.1'b
NS
12.3士0.5 12.1士0.1 12.0士02 12.0士0.1 H.8士0.0
14 dAH
NS
*
25.1士0.4 24.8士0.6 229士0.8 23.2士02 23.1士0.8
12.1士0.6'
6.9士0.3b 5.4士0.6b 4.0士0.9b 6.9士3.ob
NS
173士0.フ 16.6士07 16.3士0.4 16.5士0.4 15.8士02
NS
*
12.5士0,2'
12.1士0.1北 H9士0.1b 122士0.2北 12.0士0.2北
NS
23.8士04 22.6士1.4 22.0士0.5 21.9士1.0 21.7士07
*
17.6士0.4'
172士0.8'b 16.6士0.4b゜
167土0.3b゜
16.4士0.8゜
NS
*
*
24.5士0.9'
23.7士1.6北 22.5士0.6b゜
22.6士0.9b゜
22.4士1.0゜
NS
*
配汁牙聖業一'︑ d(N0一ど
, N伽N '
Table 4.2 Feed intake, apparent feed conversion ratio, specific growth rate, anal weight and coefficient ofvariation of di任erent feeding Schedule under di丘erent photoperiod at 14 dAH、
Par飢leters Sheltered
10 20 30 40 50 Unsheltered
10 20 30 40 50
TWO・wayANOVA
She丘er
StocMng density 110
↑20 130 140 t50 Interaction
Feed intake
(g)
0.058士0.005 0.044士0.002 0.035士0.004 0.033士0'003 0.035士0.003
AFCR
0.059士0.002 0.042士0.006 0.030士0.005 0.027士0.001 0.022士0.009
Values are mean士 SD (π=3).
*Pく0.05;1Tukey'S HSD test(TWO、way ANOVA). values for Tukey'S HSD test(TWO・way ANOVA) are mean士 SD (h=の.
Values in the column with di丘erent subscripts indicate sig11i6Cant difference q)<0.05) NS, non、significant; AFCR apparent feed conversion ratio; SGK speci丘C growth nte
0.27士0.04 030士0.03 027士0.00 0.28士0.02 028士0.03
SGR
(ツ0)
*
0.058士0.001^
0.043士0.oolb
0.032士0.004゜
0.030士0.001゜
0.028士0.009゜
NS
035土0.15 0.36土0.07 0.22士0.03 023士0.01 0.20士0.08
*
32.9士1.5 30.1土1.0 29.0士0.8 28.4士0.4 28.8士0.2
Final weight (g)
NS NS 031士0.06 033士0.04 0.25士0.04 0.25土0.04 0.24士0'06
NS
31.7士3.1 28,4士2.1 29.3士0.8 28.5士0.4 28.0士0.6
0.22士0.04 0.15士0.02 0.13士0.01 0.12士0.00 0.B士0.00
Coe命ficient ofvariation
(%)
NS
0.19士0.08 0.12士0.03 0.14士0.01 0.12士0.01 0.12士0.01
32.3士0.9"
29.2士1.2b 29.2士0.2b 28.5士0.1b 28.4士0.6b
NS
*
82士22 4.5士0.フ 6.1士2.2 57士0.3 5'9士2.2
NS
021士0.02'
0.14士0.02b 0.B士0.olb 0.12士0.oob 0.12士0.olb
NS
*
8.1土2.6 6.8士42 4.9士0.8 4フ士2.フ フ.4士0.フ
NS NS 8.1士0.1 5.6士1.6 5.5士0.9 52士07 6.6士1.1
NS
習一寄ず匁仙茸N=W巨一m吾又昌N冒温一W円四区旦芽ずミ一冬ミ貫言π雙斗小中谷団叫響﹂'︑S建嚇
N︹
4.4 Discussion
A t11reshold ofimproved survival was obtained at moderate stocMng densities ofmore
than 20 but less than 50 laNae l、1in this study. Khan (1994) reported similar trend of Optimum sulvival fbr juvenile bagrid cat丘Sh of 20.45 g stocked at 029 and o.38 flsh "
(moderate density) compared to o.H,0.20 (10w de那丘y),0.47 and o.56 丘Sh l、'(high de船ity)
However, stocMng density was relatively low compared to the present study. This could be related to different fish size and developmental stages since larger fish occupies larger area Per unit space. At higher density beyond its threshold, cannibalism resumes resulted in
S1喰htly lower sulvival at 50 laNae l、1in this study. This could be due to the increasing social
Contact and insufficient space as the density increased. However, higher t11reshold was Obselved in A丘'ican catfish. stocMng density of 50‑250 laNae r showed high suNival of more than 809る(Haylor 1992).1五gher density encouraged shoaling behaviour (Hecht and UyS 1997) and an increased in activity (Almazan Rueda 2004). This in turn hinders the signs Of agonistic and tenitorial behaviour, hence lower cannibalism and hi3her survival (Appelbaum and Kamler 2000; Almazan Rueda 2004; van de Nieuwegiessen et al.2009) Aggregation observed in bagrid catfish (chapter 2) may also suggest the preferences of Shoaling which improved survivalin this study. F0110wing this, it is possible that agonistic or
tenitodal behaviour was the cause of lower suNival at lo larvae r which coincided with
higher cannibalism. Low survival was also obseNed in A6'ican catfish flngerlings at lower
densiw ranging between 5‑10 丘Sh l"田Ossain et al.1998) and 5‑20 負Sh l、'(Hecht and
Appelbaum 1987). Atlower density, the urge to maintain a terdtory was higher compared to the urge to f己ed, hence increasing the cannibalism (Hecht arld uyS 1997). contradictory to this study, feed intake was sti11the highest atthe lowest density
Higher feed intake at lower density (10 laNae l'1) resulted in higher growth
Pedormances (total body length, SGR and final weighり in contrast with survival.1n
近大水研報 14号 (2014)