DF ] VELOPMENT STAGE , AGEING AND NATURAL MORTALITY OF ANCHOVY EGGS ( Encrasicholina SPECIES ) FROM BACAN

ln or{er to strrdy the possible application of the Daily Egg Production Method (DEPM) for estirnating nnr,hovy l0nrraxicholdnc spp.) biomass lt Bacan. Maluku, we collected a large se' ries ol'plaqkton snmples. All the onchovy eggs fronr these satuples were staged and then aged accdrcling to tlre criteria of Moser and Ahlstrom (1985). We recognised 9 stages of embryonic {evelolrmept fropr I h to l5 h. From these data we estimated that anchovy eggs at Bacan hatch betweep l5 a1d 24 h after spawning. Natural mortality (M of these eggs was estimated on threeoccasionsandaveragecl0.lSi0.04.h'. Thismeansthatonlyaboutt0%ofeggsspawned by anchovies at Bacan survive to hatch.


INTRODUCTION
The anchovies (Ertt'rasicholina species: Fam' ilv Engraulididae) are the most important of the small pelagic fishes used for live-bait in Maluku.These fish are caught at night by "bagans" to sell to the pole-and-line tuna fishing vessels.Three sper:ies o{'anchovy used as baitfish occur in Bacan waters: E. h.eterolobus, E. deuisi and E. pttrtctifer (in small numbers).Some aspects of the reproduc- tive biology of Enc:rusit'holina have been studied in crther geographic areas (Tiews el al., l97O;   Sumadhiharga.1.995: Milton et al.,1994t Wright,  1992).but there have been no studies in this area of Maluku Provinr:e.The present paper is part of a study whose aim was to assess the feasibility of estimating the biornass of anchovies using the Daily Egg Production Method (DEPM).Estimates of biomass are an important first step towards a sustainable management plan for the baitfish resources.In order to estinrate such biomass.we first needed infbrmation on the time from spawning to hatching, the embrvonic stage at which eggs hatch, and then finally egg mortality rtrtes.These values are used in the equation to estimate the biomass of fish and acr:urate data are important if the bio- mass is not to be underestimated.
One of the most important parameters needed fbr the DEPM is an at:curate estimate of the number of'eggs in the plankton.The high mortality Arnbon Asscssrnt.ntlnstitutr,ol'Agrir:ultulll 'lirhrrologv, Jl.Chr.Sopltnit, Rurnuh 'l'iga.Anrb,rrr 1t7233.Indont'sil. ('SlR0 Mtrrirrr'lltscirt'r'h.l'O l}rx l2O.CL'vt'lrrntl.Qltl.1l(ijl.Arrstr:rlil   that rrccurs in these early life stages is one of the main f'actors affecting estimates of biomass.The objectives of this part of the study were therefore to document the stages of embryonic development of anr:hovy eggs, and to estimate the age of each stage in order ultimately to determine natural mortality rates. Very little information was available on the embryonic development of tropical anchovies, but Moser & Ahlstrom (1985) identified eleven stages of development in the Northern anchovy Engraulis mordax, and showed that the rate of development was temperature-dependent.Lo (f 985) used these data to fit a temperature-dependent model of the rate of egg development and to estimate the age of eggs at each development stage.

Sampling
Random plankton tows were collected in Sep- tember, November 1996 and April, July 1997 from the main spawning grounds within the baitground in Bacan waters.Plankton samples were taken using two plankton nets (0.5 m2 and 0'25 m2 ) ol' 500-p mesh.Tows were taken vertically at vari' ous depths (5, l0 and 20 m) with the majority at 10 m.These plankton samples were preserved in lO'Zn formalin and taken to the laboratorv ftrr analysis.

Laboratory Analysis
Eggs of anchovies were identified (using tht: descriptions in Delsman, 1931), sorted and re- muved {rom all plankton tows.F'or staging.each sample was placed in a Bogorov tray and examined under a low power microscope.Anchovy eggs f rom each sample were counted and staged accord- ing to their degree of embryonic development.We rer:ognisecl 9 develclpmental stages based on morphological changes that occur during embrvoge,n.esis (modified from Moser &Ahlstrom, 1985).Eegs were aged by relating the time of sampling to the distribution of'the modes of each development stage.The mean time between samples collected with the largest number of early (Stage l) and late stnge eggs (Stager 9) was used to estimate total hatching time.These results compared favourably with the total development time estimated with t,he temperature-dependent model of Lo (lg8b).As tr f'urther confirmation that the estimated devel- opntent time was realistic, we used the equation of ' Longhurst & Pnuly ( l9U7) to relate water tem- IFR Journal VtLIV No.2.1998 pcrature and egg diameter to estimate total hatch- ing time.All three approaches agreed that the to- tal time fbr eggs to hatch was approximately 16 h.lllving established that eggs took 16 h to hatch, wc. then interpolated using the model of ln ( 198S)   to estimate the age at each of the g development strrges.
Natural mortality of eggs was assumed to be relatively constant, enabling us to estimate mor.
tality from the exponential decay model (Sparre & Venema, 1992).We sampled continuously for 36 h on seven occasions during the study.This enabled us to identify the peak period of hatching and to try and follow a cohort of eggs through to hatching and thus estimate their mortality.Tows were made from a fixed position during the night (usually from an anchored bagan) and then ran- dornly in the spawning area during the daytime.A linear regression was fitted to the logarithm of the number of eggs in each stage against time (in hrs).The slope of this regression is an estimate of the egg mortality during that period.head can be identified but the strur--ture is not clearly defined head is clearly defined and vertebrae can be seen in the middle of the egg head, vertebrae and tail make up more thtrn half of the egg the tail begins to separate from the volk the tail is clearly seen and the yolk has begun to be resorbed R. Attclunurt'i, D.A. Millon, and A.N. Slsrullo

RESULTS AND DISCUSSION
Most plankton samples contained eggs in sev- eral stages of development.This means that an- chovies do not spawn their eggs at the same time.
The stages of embryonic development were sinrilar to those identified lt>r Engraulis ntorclax (Moser  & Ahlstrom, 1985).The stages and the estimated age are described in Table I and shown in Figurcs I to 3. The data suggest that anchovy eggs hatch alter at least l6 h.This result is also contirmed when the relationshill between mean egg diameter and water tempertrture is calculated (l,onghurst & Pauh'. 1987).The estimated mean time to hatch- ing using this equation was also lG h.
The 36-h plankton sampling did not show any clear pattern of redtrcing egg abundance with time on lbur of the seven da.vs.However, the plankton data on the other three davs (13 April, l4 and 16  Jull' 1997) could be used to estimate mortality for some of the older age classes.The mean estimated nrortalitv (il4) was 0.15 t 0.04.hr for those three days (F'igure 4).The mortality estimates for each clav varied f rom 0. 1 4 t 0.03 to,O.17t 0.01'-r.hr.This ec;uated to a me an rnortalitv ol'approximatel.y9O'2" over thc 16 h fronr sltlwning to hatching.These values are: much higher than those estimated {br more temperzrte slrccies.Koslow & Bulman (1997)  estimated egg mortalitv to be ab<-rut I'r'2,.d' {br Blue (i renad ier Mac r u r o n u s nouaeze I ondiae off south - ern Tasmania and Fletcher el o/. (199G) found values between 0.43 to 0.81.d'' for the sardine Srlrdiiro;.rs $(rgo-r in Western Australia. Anchovies have adapted to the high, and prob- ablv variable, mortality rates of eggs and larvae by spawning frequent, small batches of eggs (Antlamari & Milton, this issue).Both E. hetero- /obrus and E. deuisi have higher lifetime egg pro- duction in Bacan than in the Pacific (Solomon Is- lands) which mny be an adaptation to high egg mrlrtality.We tried to separate the anchovy eggs in the plankton into different species according to Delsman (1931) and almost all were of the E.
heterolobu.s form.Given that E. lrcterolobus is also the dominant species in adult catches and their higher egg production (Andamari & Milton, this  issue), our mortality estimates are probably specific to this species.the Indonesia-Australia Baitfish Research Project in eastern Indonesia funded by the Australian Centre for International Agricultural Research (P/ N 9424).

Figure 3 .Figure 4 .
Figure 3.The mid-stages of anchovy embryo development showing the early larvae'

Table 1 .
A description of the embryological development stages of anchovy eggs ancl the estimated time since spawning.