DURATION AND TIMING OF SPAWNING SEASONS IN MARINE TELEOSTS: A BIOGEOGRAPHICAL APPROACH

The duration and timing patterns of spawning seasons among marine teleosls tishes over broad geographic ranges from the polars to the equator has been reviewed. The review was based on 2OO studies In order to evaluale potenlial patterns of the length or timing of spawning seasons in relation to tatitude, habitats and taxonomic groupings patterns. This review confirmed that the timing and duration of spawning of marine leleost fishes is related to latitude. The spawning duration of fshes liting in the equatorial regi6n and the tfopical regions is generally longer than that ot species living in the subtropiial and polar regions.-With respect to-lheir habital, in general, the duration and timing patterns of spawning seasons of peiagic fish were'not different to demersal fishes,


INTRODUCTION
The duration and timing of spawning seasons vary substantially among marine teleost fishes (Taytor, 1990;Munro et a/., 1990). Breeding s€asons c_an_range from periods of just a few days (Bye, 1990;Beddow et a/., 1999) td alt year round (Stequert & Ramcharrun, 1996). Ainong the seasonal breeders, there is also considerable variation in the time of the year when breeding occurs _ (Sundararaj, 1981i Longhurst & paulyl ]!87) Some species spawn mainiy during summ6i (Sabates & Martin, 1993), white o ier spaivn during winter (Amara et a/,, 1994; (Fowter ei a/., 1999)i Ine processes underlying vbriation in the duration and timing of spawning seasons are poo.ly understood. To date there has been n6 comprehensive review of pitterns in the length or timing of spawning seasons, making it ditfijult to rormutate appropriate theories.
The timing and duration of spawning seasons appear to vary with latitude, both within and amono species (Taytor, 1990). lt is widely assumed tha-t the length of the spawning season increases toward the equator as temperature increases and becomes less variable (Munro et at, 1990). However, there have been no systematic compansons over a broad latitudinal range to confirm this. Most studies on the spawning seisons of marine fish have been restricted to-a single locality and elso al high tatitude locations. Tiis restricls our ability lo determine how single species alter lheir spawning patterns over poteniially broad gpographic ranges. However, over recent vears. information on the spawning patterns of tropicai species has been increasing. There have now been ' Ressarch C€nlertor Captur€ Fisherier, Muare Baru-Jakada sufficient studies on a range of species from different locations to detect major trends from the polar regions to equatorial environments.
There are several factors that are widelv assumed as potential causative factors of th6 timing of the reproductive season. photoperiod. temperature, rainfall and food, among other factors, are important in regulating reproductive cycles in teleost fishes (Baggerman, 1990;Taytor, 1990).
Many spocies live over a wide range'of latitudes and so encounter diff€rent temperature and photoperiod regimes at different locations. This can result in major differences in the timing and duratipn of reproduction within specios (Wootton, 1990). Reproductive seasonality has been weli described for high latitude species and is correlated with major seasonal changes in temperature and hours of daylight (Wootton, 1990). Temperature and photoperiod appear to influence both the timing and duration of the spawning season, with mosl temperate _ species having reslricted spawning periods (Munro et a/., 1990). However, stron! reproductive seasonality has also been observed al low latitudes in some freshwater and marine species (Bye, 1990;Taytor, 1990). This suggests that the assumption of longer spawning seasons in tne troptcs may not always apply. Before any conclusions can be drawn, however, a systematic comparison of spawning seasons of different taxa across a wide range of latitudes is necessary.
Marine teleost fishes living in different habitats (e.9. demersal and petagic fish) experience different abiotic and biotic conditions. Micro climates associated with habitats mav also influence reproductive strategies, including breeding at a time and place which is most conducive to the survival of their otfspring (Bye, 1990). lt is thought that pelagic fishes tend to have longer spawning periods than demersal fish. Fot example the dolphin fish, Coryhaena hippuru (Wu et al., 2001\ and the skipjack tuna, Katsuwonus pe,amis (Stequert & Ramcharrun, 1996)  The puroose of this review is to draw together information from disparate sources to examine variation in the timing and duration of spawning seasons in marine fishes. The review will evaluate potential patterns in relation to latitude, habitats and taxonomic groupings. The following specific ouestions will be addressed:  (Johnson et al, 1998) (Wyanski ef at, 2OOO) (Rasem et a/, 1997) (Marino ef (Albert, 1994) 20.00 N (Golani, 1994) 20.00 N (Golani, 1994) 34 s. a a e. Sadhotomo' 2OO5), and the halfmoon irouoer Ephinephetus ivulatus (22' S) which loawns foi 6 m6nths (July to December) (Mackie' 2b0o), both of them living in tropical waters' Not all species living at low latitudes spawn all vear around or have longer spawning periods than inoie tiving at high latitudes some low latitude io""i". ."'n have-a fairty short spawning dwation eio tne wftafe shark, Rhincodon lypus, (16 30" N) arii soanish mackerel' scombe/omus commerson' 6 (19' S) each spawning for 2 months dunng spnng (Mcpherson, 1993) To overcome this problem I have grouped studies based on six wider latitudinal ranges Th-ese groupings are: 1) the southern polar zone-(60 to 5o' st: Zt the northern polar zone (60 to 90" N); 3) G .6utdern temperate zone (30 to 60" s)i 4) the northern temperaie zone (30 to 60' N); 5) the north iiooi"ir ron,j (15 to 30' N): 6) the south tropical io[" irsto 30; s); and 7) the equatorial.zone ('15' S tof S" Hl ( Figure 1) There are smaller differences in temperiture throughout the year at low latitudes comoared to high latitude Stable environmental piorEtti"t are liiely to provide suitable conditions ioi'ootn stable growth and reproduction The equatorial zone (15'S to 15',N) has the smallest Jlifv inO annual temperature variation' theretore mar'ne teleosts tiving in this zone will rarely experience extreme temperature fl uctuations On average marine teleosts Iiving at lower tatituJes (<30;) have longer spawning periods than i]ei;;i"ii'"i hisn t"iitud"t (>30') (Fisure 2)   There was a trend of increasing spawning duration from polar region (60 to 90'N/S) to the equatorial region (15' N to 15' S), and this trend was significant overall (Figure 2, one way ANOVA, F=4,878, p=0.000). Spawning duration of marine teleosb living in the equatorial region, however, was not significantly longer than that of species living in lhe northern and southern tropical zones (15 to 30'N and 15 to 30'S) ( Table 2, Posl hoc comparison, p=0.570; p=0.23 1). This result is probably due to the small number of studies (15) in the 15' N to '15'S zone, with only Seven of these close to the equator (i.e. 5' N to 5'S).
Generally, the spawning durations of species living in low latitudes breed longer than species living in sub tropical regions. Up to 43 out of 73 species in the tropical zone have a spawning duration of 5 to 12 months (Table 1) Teleost fishes living both in the northern and southern polar regions (60 to 90 N/S) have significantly lower average of spawning durations than those at lower latitudes. Subtropical teleosts fishes (30 to 60 N/S) also show a significant difference in spawning duration compared to the tropical area (15" N to 15" S, 15 to 30" N and 15 to 30' S). Although the exact time of breeding depends on the species, breeding generally starts later at higher latitudes (Baggerman, 1990). This supports the general assumption that in the tropics, marine teleosts are more likely to have a longer spawning period.
The patterns of spawning duration look closely match the pattern in sea surface temperatures ( Figure 3). Water temperature reach a peak of 28' C in the trooics and decline below 0'C at the Dolar Spawning periods of marine teleosts at different latitudinal ranges (To conform the assumption of using one-way ANOVA test the data were log lotransformed).  The sea surface temperatures in the tropic are much more stable than pole regions. For example, the annual temperature variation is much greater at the poles (i.e. 20"C )and in the subtropics (i.e. 25'C ) than in tropics (i.e. 5'C) (Bigg, 1996).
Temperature regimes are also influenced by seasons showing very different patterns between the equatorial and southern northern hemispheres.
15N-'t55 0.024 There are three hypothetical temperatures regiFnes in these three areas regimes. The equatorial zone exhibits a stable cycle through a year, whereas the southern temperate zone will have a temperature peak at about 35 to 40'C in December (summeO and decreases to the lowest temperature of -5"C in July (winteO. Following the opposite pattern, The northern hemisphere exhibits an opposite pattern reaching the lowest temperature of -5'C in December (winter) followed by an increase to the

RELATED TO FOOD AVAILABILITY
Food availability may also affect the spawning duralion of teleosts fishes. Differences in the timing of spawning in the Cod (Gadus morhual are not related to temperature but to the timing of plankton production (Brander, '1994). Adequate levels of nutrition must be fulfilled in order to satisfy the animal's physiological condition and ability to complete the reproductive cycle (Nielsen, 1998). The seasonal fluctuations in food availability may determine the timing of reproductive development in some sDecies of teleosts (Collins & Anderson, 1999 Figure 3. Sea surface temperature in the three major oceans within 70" N to 70' S (Bigg' 1996) maximum value of 40'C by July (Summer) (Bigg, 1996) (Figure 4). soulhatn ternpefale zon€ equalor nonhern l6mperate zone Hooth Figure 4. Three hypothetical temperature patterns (equatorial and temperate zones) (Bigg, 1996). rlt< Y--t_a L ts t-I ,n for reproduction. Fish in poor conditton (i.e. low fat content) will exhibit delayed maturation (Rajasilta, 1992; Slofte et a/., 2000). A restriction on feeding during oocyte recruitment or vitellogenesis, or both, could result in poor condition and thus delay the onset of maturation (Kjesbu al a/., 1996), this could also result in fewer maturing females (Bromley 6f al, 2000) at the start of spawning season.
An increase in food abundance can result in lish storing the maximum energy possible whlch is accumulated as body fat. This stored energy is to be used for metabolic and reproductive requirements (Paul 6t a/., 1993). A fast increase in body energy/weight indicates that much of the season's energy acquisition takes place in a very short period of intense feeding e.g. 2Eo/o ol the Yellowfin sole, Pleuronectes aspar, is accumulated within one month during mid May to mid June, when plankton reaches a peak in abundance (Paul 6t al, 1993). The spring to autumn energy storage strategy is used by the Yellowfin sole and the Northern flat fish sDecies. Inter annual variations in energy storage can, therefore be related to variation in food abundancd (e.9. plankton), and could affect the rate of gamete production (Paul sf al, 1993).
Phytoplankton and zooplgnkton, the main food resources of many marine teleosts, are strongly influenced by season (Munro ef a,., 1990). Thus the seasonAlity of plankton abundance will atfect both growth end reproduction of planktivorous tish.
In the North Atlantic plankton reaches maximum abundance within March to May and decreases to a minimum during November to February ( Figure 5). Some fish species in the Northwest Atlantic spawn mainly during periods of high plankton abundance.
For example, in three species of butterfly fish Peprilus triachantus, Peprilus butti, Peprilus alepidotus the peak spawning period is March or April (Rotunno & Cowen, 1997). Similarly, some fish in the northeast Atlantic for examples, the Rosefish, Helicolenus dactylopterus, and the Roundnose Grenadier, Coryphaenoides rupestris' spawn mainly in June which is during the period of peak zooplankton abundance (Allain, 2001) Plankton abundance in tropical waters is relatively stable throughout the year ( Figure 5) (Moyle & Cech, 1996). Therefore planktivorous fish in this region would rarely experience periods of insufflcient food. As a result, teleost fishes will be expected to have stable growth and reproductionr The lack of seasonal variation in food availabiliti means that tropical teleosts should have long reproductive periods. In contrast plankton aiundance in the polar and subtropical regions ii more seasonal. This might affect the growth and reproduction of marine teleosts in this region.
During periods of high food abundance, marine teleosts in the subhopical and polar regaons grow and store energy as fat. During seasons with low food abundance, fish depend on stored energy for their reDroduction. A number of trends can be observed from this analysis. The first trend at high latitude, peak spawning periods are generally during spring and summer (i.e. April-July in the northern hemisphere, October to December in the southern hemisphere) ( Figure 6). However there are some exceptions, for examples, the Greenland halibul, Reinhardtius h,ppoglosso/des at 73" N spawns in December-January (Albert et al.,2OO1); the deepwater squalid shark, Etmopterus prr'hceps at 65.30" S spawns in June to July (Jacobsdittir, 2001). The second trend is that teleost fish tend to spawn from January to July (Winte Spring)  :..-=::----,<*: JFMAMJJASONO seasonal productivity in zooplankton and phytoplankton in different parts of the world (Bigg, 1996;Moyle & Cech, 1996) (Golani, 1994), Also the dusky gtoupet, Epinephelus marginatus, at 36' N increases it reproductive output five lo eightfold during summer (April to June) and decreases it drastically during winter (December to January) (Marino et a/., 2001) as the processes of gonadal development are considerably slowed down by low water temperatures (.8') (Scott, 1990), Meanwhile the spawning ot time of the marine plotosid, Cnidoglanis macrocephalus at 28" S, is related to water temperature with a higher tem petalure of 22" C accelerating the start of spawning by a month compared to the lower temperature of 18' C (Laurenson ef a/., 1993). Spawning in the bluefish Pomatomus safati'i (L) at around 41' N is limited to the warmest months, from July to September when the surface lemDerature is about 25'C (Sabat€s & Martin, 1993). Lutanus yiffus at around 30'N spawns for up to 8 months but reaches its peak within the summer time (Cuellar et a/., 1996). These examples provide some evidence to support the idea that marine teleosts are likely to begin spawn in the warmer period.
Photoperiod is also widely known as a factor potentially affecting the spawning time, but its effects vary according to species. For example, constant short photoperiod regimes advanced spawning whereas constant long photoperiod period regimes delayed it in the sea bass,  , 1999). In contrast, studies on salmonids show DIFFERENCES BETWEEN DEMERSAL AND the opposite pattern with a detiy in the start of PELAGIC FISHES lN THEIR PATTERNS OF .p",rnihg by constant short phoioperiod and an DURATTON OF SPAWNING SEASON advancjby ionstant long photoplribd Fakashima Habitat differences may also be associated with & yamada, 1994). This inOicatei that ihotoperiod ditferent environmental fqctors such as potentially affects the reproducfive timini. Ho\iever, temperature, salinity and water pressure as well as io validjte this assuinption more riproductive food availability Patterns o.b-served in demersal tish studies are required to examine the effect of which mostly spend their lifetime close-to the sea ptt-otop.rloo on "'*ider range of species.  -2006. 1-14 to the sea surface. Associated seasonal changes in food availability and water temperature can lead to changes in fish behaviour. Pelagic fish biologically are more likely able to respond to changes by moving from place to place, For example The skipjack tuna, katsuwonus pelamis, migrates from the south (20' S) to the north (20' N) in order to inhabit areas with conditions suitable for growth and reproduction (Hunter ef a/., 1986;Stequert & Ramcharrun, '1996). This result in this species having a relatively long reproductive period.
Skipjack tuna spawnjng occurs throughout the year in tropical waters and seasonally in subtro,ical waters in all major oceans (Nishikawa ef a/., 1985). They are oppurtunistic in their reproductive strategy and are thought to spawn throughout their distribution whenever water temperitures rise above 24oC (Schaefer, 2OO0).
. Analysis of spawning durations of pelagic and demersal fishes did not support the pradiction that pelagic fish have longer spawning durations ( Figure  7, Table 3), This may be due to the fact that nolt atl pelagic fishes are able to increase the length of their spawning season by moving to areas ihere conditions are favourable. lt may also be due to a lack of data for pelagic species.

CONCLUSION
This review confirmed that the timing and duration of spawning of marine teleost fishes is related to latitude. The spawning duration of lishes living in the equatorial region and the tropical regrons rs generally longer than that of species living in the subtropical and polar regions. In the equatorial peak spawning time is spread throughout the. year, whereas at high latitude peak spawning periods were generally during summer. In terms of the spawning duration of mafine teleosts with respect to their habitat found that pelagic fish were not different to demersal fishes. The main oroblem encountered was the paucity of studies on species centred on the equator. More reproductive studies on marine teleosts are still required for equatorial soecres.