D. Determination of Additional Aggregation Parameters

Methods for examining the numbers of fish in an aggregation and the area it covers have been discussed. There are many other important parameters of aggregations that can be collected or assessed using a variety of techniques. These include duration of specific aggregations, activities of individual fish within and between aggregation periods, precise timing of spawning, etc. These are covered in the following sections or elsewhere in the manual, as appropriate.

Determining the length of time an aggregation persists at a site is quite important. Fishermen are often able to provide good information relative to this question as their ability to catch fish depends on the fish being present. Fishing activity can, however, reduce the time an aggregation is present, since a large proportion of it can be fished out in a short time, hence reducing its persistence in the short term (and possibly its existence in the medium term). Making direct observation through diving or snorkeling of aggregation presence is useful, but caution must be used in doing this. It is possible, for example, that the aggregation might have moved slightly since the last observation and will be missed if only a limited area is checked. The surrounding area can be checked if the aggregation is not in its "normal" location, however this does not prove that it isn't present somewhere nearby that was not checked. Such problems are simply indicative of the uncertainty of dealing with many aspects of aggregation field studies.

Individual fish may not remain at the aggregation for its entire duration. There may be sex specific differences in the length of stay, with males generally persisting longer. Studies on individual residence at the aggregation usually require the ability to identify individual fish. Some species have individually distinctive markings, such as the Nassau grouper, which make it possible to identify specific individuals. The disadvantages of using such individual markings is that you generally have to approach the fish closely, something that may not always be possible, and when there are hundreds or more of a fish, finding a particular individual may be nearly impossible. Determining persistence from identifiable individuals then usually requires daily observations to be certain the fish is consistently at the aggregation site.

In assessing various aggregation parameters, it may be important to understand the manner in which transient aggregations (see Section I) build up and disperse. This can vary considerably between species, even within the same family, and might well influence the type of data taken or the way in which data are collected. Different examples are included in various sections of this manual but, in general, the three following categories of aggregation build-up are known (dispersal is usually soon after spawning and tends to include a large proportion of assembled fish). (1) Males reach the aggregation site earlier than females and establish territories. Females enter subsequently, sometimes in small groupings and stay a much shorter time at the aggregation than the females. (2) Large groups (sex composition unknown) of individuals migrate along specific routes to reach an aggregation site. (3) fish slowly build up in numbers until spawning at the aggregation site and may then quickly disappear. As we come to learn more of the details of aggregating species, additional categories may well be identified.

Assessing Aggregation Sex Ratios and Sampling Gonads

An evaluation of a spawning aggregation often involves the assessment of its sex ratio, most often the number of reproductive males and females. There are two components to such a task; identifying which individuals are reproductively active, and, of these, how many are males and how many are females. This would seem to be a relatively simple task, but, like so many other apparently easy exercises, several factors must be considered. The first is that not all fish at an aggregation are necessarily reproductively active. An aggregation of red hind, Epinephelus guttatus, in Puerto Rico involved a significant percentage (13-36% at different times) of fish without ripe gonads; most of these fish were smaller than the 100% size of sexual maturation. Therefore, it can not be assumed that all fish present in an aggregation are reproductively active.

How do we sample an aggregation adequately to establish the sex ratio? Questions that need to be asked are which days to sample, how many days, what time of day to sample, how to sample (i.e. what method will be used to determine numbers of males and females) and how many fish need to be collected? Does the sampling method used produce a representative sex ratio? For example, spearfishers intentionally select male tiger grouper, M. tigris, (to leave the females to reproduce) so their samples tend to be male-biased (Sadovy et al., 1994b), relative to what was observed by the divers. On the other hand, the sex ratio obtained by hook and line fishing over a red hind aggregation did not differ significantly from that obtained by biologists spearing fish underwater (Shapiro et al., 1993). Whatever the method of collection, every attempt must be made to determine whether the sample represents the true sex ratio and, for exploited aggregations, fishers could be interviewed to identify any possibility for bias. Ideally, sex ratios obtained from fishing methods should be validated by in situ observations, as long as sex can be assigned accurately by divers. Finally, never assume a size and sex relationship since size and sex frequency distributions often strongly overlap, and size in relation to sex can be markedly different in different social groups, especially in sex-changing species (e.g. Shapiro, 1981).

Figure 22. Ratio of active male to maturing and active female red hind, Epinephelus guttatus for 16 sampling days in 1991. Number of fish caught each day appears above each column and asterisks denote that the sex ratio differs significantly (p<0.05, chi-square) from the preceding sample date. Full moon occurred on 30 January and 28 February (after Sadovy et al., 1994a, with kind permission of Kluwer Academic Publishers).

Figure 23. Catch trends for camouflage grouper, Epinephelus polyphekedion, taken at the Kehpara Island Marine Sanctuary February 1998-March 1999 showing the arrival of males, followed by females several days later. Spawning time(s) was determined by a combination of gonadosomatic and histological analyses (Spawning = S; solid square = males; open circles = females) (after Rhodes and Sadovy, 2002, with kind permission of Kluwer Academic Publishers).

Only sex ratios on the few days of spawning can be considered to be operational (i.e. reproductive) sex ratios. In the red hind, E. guttatus, modest samples of males and females collected over 16 days showed significant shifts in sex ratio from day to day both before and after spawning (Fig. 22) (Sadovy et al., 1994a). Note that sample sizes in this case were small on any one day of sampling which could have contributed to extremes of day to day variation since inclusion of just a few individuals of the rarer sex can substantially change the sex ratio.

Establishing a meaningful and representative sex ratio can be difficult not only because of sampling problems but also because sex ratios can change from day to day and also during the course of a single day. Operational sex ratio might require assessment at the time of spawning since, in a number of species, males precede females to the spawning site. For example, in E. polyphekadion, which spawns around full moon, males start to enter the aggregation site 10-12 days prior to the full moon, while females start to arrive about 4 days before full moon, stay a few days until spawning and then all fish leave the site (Fig. 23).

If data on fish sex is to be taken from commercially fished samples, it is important to determine whether the gonads are valued for food and when the fish are likely to be cleaned and gutted; this will dictate how you plan your sampling. Always ask for permission to touch another person’s catch. If not all fish can be sampled, then some method must be devised to obtain a representative subsample of all available fish coming off fishing boats or landing in the market place. It is important to remember that the more stages there are between the spawning site and sampling, the bigger the possibility that errors of sampling will be introduced. Check whether fishers have returned particular size classes of fish (either too big and therefore possibly ciguatoxic, or too small for a good market price or whether small fish may be retained separately in the boat for later home use) to the water. Illegal, undersized, fish may also have been returned to the water. If the fish are weighed in markets or coming off boats, be sure to note whether or not they have been gutted.

If there is an opportunity and a need to collect gonads from fish caught at aggregations, then simple guidelines can be followed to handle and preserve them. Carefully remove the gonad trying not to rupture the ovary which, if full of hydrated eggs, will become difficult to handle. Subsequent treatment will depend on the objective(s) of gonad sampling. If fixation is necessary (i.e. the gonads will not be worked on while still fresh), the gonad can be placed in 10% formalsaline (10% formalin solution) in ratio of about 10 parts solution to 1 part of gonad. If the gonad is whole and very large (e.g., Fig. 24) it will be important to penetrate it or carefully slice it at intervals along its length to allow the fixative to penetrate the tissues before those towards the center of the gonad deteriorate. Fixation can take a couple of months. The gonad should then be washed and transferred to 70% ethanol for storage. If sections of the gonad are to be retained (i.e. the whole gonad is not necessary), then these should be carefully sliced (for very ripe gonads it helps to place the whole gonad in fixative for a short time [hour or so] and then to slice after the gonad has hardened a little). Slices of gonad should be as thin as possible and not thicker than 5 mm if this is practical – this allows for sufficient penetration of fixative. Fixation of slices is much quicker than fixation of larger unsliced gonads and sliced gonads can be transferred to preservative after a couple of weeks. Make sure that the jars used for fixation are large enough and seal well since formalin is unpleasant and should not be inhaled or touched if possible (use gloves while dealing with formalin solution). Place labels inside the jars with the time, date and place of capture and the size and weight of fish if possible; labels should be strong paper or card with details clearly written in pencil. If no fixative is immediately available, take the fresh gonad or gonad sections and freeze them in a bag. When fixative becomes available, place the frozen gonad or section directly in the fixative before defrosting and continue as above. Sometimes, as has happened to one of us, you may have no fixative or no freezer! There may still be hope of preserving some material if you have access to vodka or even surgical alcohol (local chemists would have this). This is not a perfect solution by any means but can make the difference between a gonad that is useable for some purposes and not useable at all. There are many good papers with instructions for histological preparation and studies - the key is to decide what you need the gonads for and to make sure that the materials are as fresh as possible when fixed.

In cases where fish are to be sampled alive, they can be measured and weighed on board the boat. In some cases the fish have been brought up from depths that require the swim bladder to be deflated with a needle. If you are working with a live fish fishery the fishermen will be very experienced in deflating these fish so you will not have to do this yourself. If the fish are to be returned to the water it is best to first see if the fish can swim back down on their own. Sometimes a little rest and a flick of the tail is all that the fish require; carefully sliding the fish into the water head first helps get them moving in the right direction. If the fish can not swim down on their own they must be deflated. This should be done with a sterile 18-gauge hypodermic needle. There is a very good spot to puncture the fish just behind and in-line with the pectoral fin. Since some species have elongated pectoral fins it is a good idea to dissect a dead specimen to identify the spot. You are looking for a location at the anterior dorsal region of the swim bladder, where the swim bladder forms a pocket in the muscle. The swim bladder adheres to the muscle at this location so the needle will pass through muscle, directly into the swim bladder without entering the portion of the peritoneal cavity that contains viscera. Find a marker on the outside of the fish that will guide you into this part of the swim bladder. Once you’ve identified the spot simply remove a scale (or slip the needle under the scale) and push the needle down into the fish (at a 90 degree angle to the body) until you hear gas escaping. Many people like to do this in the water so that they see bubbles escaping from the needle; do not do this since seawater can travel into the swimbladder and cause a bacterial or fungal infection.

Live fish can be sexed by gently squeezing the gonads of the specimen to express eggs or milt. Pressure should be applied (with thumb on one side and fingers on the other) to the dorsal- most region of the gut cavity. Start by gently squeezing the gut cavity about 2/3 of the way towards the head and then slide your fingers back to the vent as if trying to gently squeeze paste out of a tube. This is best done with the fish lying on its back. You will often see feces come out

Figure 24. (Left) Female gonads from a large Nassau grouper(held in a pair of man’s hands) indicate the size that some gonads can achieve in that species. In some species ripe testes can get almost as big. (Right) Close up view of Nassau grouper ovary with hydrated eggs visible through the surface integument. The hydrated eggs are about 1 mm in diameter and appear dark (actually translucent) in this view. Scale approximate.

of the anus; wipe them away and keep looking for the pore just posterior to the anus. From this pore you may see milt, or, sometimes, hydrated eggs. Males are very easy to sex in this manner since they are running ripe (with milky sperm) at all times during the spawning season. Females may only exude hydrated eggs (Fig. 24) during a very short period of the day (an hour or two before spawning, often, but not always, at dusk). Hydrated eggs are usually quite runny and clear, while milt is milky white. An experienced “fish squeezer” can eventually reliably (but maybe not 100%, depending on the species) sex females simply by the appearance of the urogenital pore. However, prior to hydration, the vent may not allow eggs to pass even if ripe and sex can not readily be confirmed. In such cases, it may be possible to gently cannulate the ripe ovary, a technique that requires training and appropriate materials. Note that people assisting on projects may react in unexpected ways to handling fish or to taking samples. In one place, for example, when fishers were part of the sampling team, they did not want to touch the ripe males (the sperm) and we had to introduce gloves to continue sampling.

Seasonality of Spawning

Gonads collected at regular and reasonably short intervals over a period of time of a year or more can often provide strong evidence for seasonality of spawning in both aggregating and non-aggregating fishes using GSI or histology (Fig. 25). However, it is important to collect regularly, frequently and in sufficient number to ensure a reliable indication of spawning season. Females will tend to provide the most detailed information on the spawning season. Monthly samples sizes of 25-30 or so mature-sized females are recommended as a general rule.

Determining What Time of Day Fish are Spawning

For many species (especially for those that spawn during daylight hours in shallow, accessible waters) direct observations may be made throughout the day including the period during which spawning is expected to occur (based on market samples, anecdotal evidence, etc). Where only a limited number of observation periods can be made, it is always possible that fish

Figure 25. Spawning seasonality of ocean surgeonfish, Acanthurus bahianus, from southwestern Puerto Rico based on gonadosomatic index (GSI). There is a clear peak to the spawning season during the northern hemisphere winter in this species. Data from Colin and Clavijo, unpublished MS. (PLC)

may be spawning at some time when we are not observing them, leading us to a false conclusion about the range of spawning times. We need to always keep this caveat in mind, and report our findings with suitable qualifiers.

It is more difficult to pinpoint the time of actual spawning if this occurs after dark, or in deep or inaccessible (i.e. low visibility or high current) conditions. This has been the case for many species of snappers and some groupers which we suspect may spawn after sunset, although definitive evidence is still lacking. In such cases, indirect means may have to be used to verify spawning time. Moreover, be aware that the presence of divers could influence fish behavior. Among those species that form transient aggregations, it is sobering to keep in mind that of all the groupers that aggregate to spawn only a handful of species have been observed actually releasing gametes, as far as the published literature is concerned.

If gonad materials can be obtained then either macroscopic or microscopic techniques may be used to determine time of spawning. Once the approximate time of spawning is determined, regular (daily or even hourly) samples can be taken from the aggregation site. If gonad and body weights are available, then the GSI (the percentage that the gonad contributes to total body weight) can be calculated (Sadovy et al., 1994a; Rhodes and Sadovy, 2002) (Figs. 26 and 27).

Figure 26. Mean gonadosomatic index (GSI) for ovaries taken from female Epinephelus polyphekadion in the week up to and including spawning. Sample sizes are given for each day. (redrawn from Rhodes and Sadovy, 2002, with kind permission of Kluwer Academic Publishers).

Figure 27. Mean gonadosomatic index (GSI) and standard deviation for ovaries taken from female red hind, Epinephelus guttatus, captured between 22 Jan and 28 Feb 1991 in Puerto Rico. Sample sizes are given for each day. Open circle indicates full moon and closed circle new moon (Sadovy et al., 1994a, with kind permission of Kluwer Academic Publishers).

Figure 27 plots the GSI relative to lunar phase and shows the clear relationship in E. guttatus between the approaching full moon and high GSI. In the days just before the full moon, the GSI drops quickly as fish spawn and release most hydrated (heavy) eggs.

Figure 28. Oocyte diameter measurements for female Epinephelus polyphekadion taken from the spawning aggregation area relative to the time of spawning (redrawn from Rhodes and Sadovy, 2002, with kind permission of Kluwer Academic Publishers).

Measurements of oocyte diameters (using unfixed material and a microscope with micrometer) can be used as a simple means to track oocyte maturation and identify, indirectly, when spawning occurs (Fig. 28). As eggs mature in the days leading up to spawning, they increase in diameter and when they are released, mean egg diameter will decline reflecting the smaller diameters of the oocytes remaining in the ovary. More precise estimates, needing subdaily (preferably hourly) sampling, are possible if oocytes can be measured for changes in oocyte diameter (of the largest oocytes present) over time (Fig. 29). If gonads can be prepared histologically, the presence of post-ovulatory follicles (the follicles surround the mature egg until egg release and collapse, remaining for a short time after spawning – they therefore represent a useful indicator of very recent spawning) can be used to detect recent (i.e. within the previous 1-2 days) spawning (Fig. 29b). As always, thought must be given, however, to ensuring that sufficient samples are taken to be scientifically meaningful.

If fish are to be returned alive to aggregations, see above for handling ripe fish and detection of hydrated eggs. Regular, hourly or so, sampling of aggregating fishes will be necessary to pinpoint time of spawning. The presence of running sperm is not necessarily an indication of imminent spawning in males, although the release of clear, hydrated, eggs does signal imminent spawning in females.

DNA Samples

Molecular genetics can now be done using a small sample of tissue from which DNA is extracted. Tissues can be preserved in ethanol (the higher concentration the better, 95% being preferable) or other buffers in small o-ring sealed vials of 1-2 ml volume. Tissue samples can be taken from freshly fish caught by fishermen, and just about any tissue will do. Samples can also be collected by fin clip of live fish, cutting a piece of the membrane between fin spines (usually the dorsal), which does not hurt the fish and allows it to be returned alive to the water. It is also possible to take tissue samples by spear, using a modified biopsy type needle to collect a plug of tissue when the spear is shot into a fish. Fin clips have also been taken from large groupers such as jewfish (Goliath grouper) (E. itajara), that were docile enough to let a diver grab their dorsal fin membrane and quickly slice a small piece of tissue from the membrane. The fish hardly react to the taking of such a tissue sample.

DNA samples, if they can be analysed at a fine enough resolution, may be useful for looking at the population structure of aggregating fish and in some case might provide some insights into where an aggregating population is coming from. If samples are taken over a broad geographic region, the population structure of a species might be evident and allow some statements about the genetic interchange via planktonic larvae among a species. Care is needed in interpreting genetic information, however, since the absence of any apparent population structuring following analysis of DNA material, while it may indeed indicate that no structuring is present, could also be reflecting inability to detect such differences due to the resolution of the techniques used.

Gut Contents

The gut contents of aggregating fish may be of interest if it is important to determine whether or not fish of both sexes are feeding during the aggregation. In many cases aggregating fish take baited lines or enter traps which are baited with live fish or other live or dead matter; in others, however, aggregated males may not feed and can be difficult to sample by hook and line (Sadovy, pers. obs.). In most cases speared fish from areas where fishermen and traps are not present are to be preferred for examining natural gut contents. Hungry fish are the ones that tend to take baited lines or go into baited traps, and even if a fish is speared, it is always possible that the fish speared may have gone into a trap, taken bait and exited or might have taken bait from a hook without being caught. Where traps are unbaited, gut contents may say something about the state of the aggregated fish, but again should be interpreted with caution, as other fish, better fed, may not enter the trap.

Sampling Otoliths

Otoliths are often collected from aggregating fish. They can potentially provide information on age and growth or may be used for fine microchemical analyses. Various otoliths have been reported to have "spawning marks" indicative of when the fish has spawned previously. While such marks have generally not been positively verified, it might prove useful to compare otoliths from fish captured at an aggregation with a number of individuals, tagged at the aggregation and collected later, for evidence of spawning marks. The collection, preparation and analysis of otoliths is not further within the scope of this manual and the reader is urged to consult appropriate references for these items. However, it should be noted that many reef fish can not easily be aged using otoliths so these should be checked for growth marks before a full- scale sampling programme is launched which can be time-consuming and costly.

Figure 29. Oocyte diameter frequencies taken (A) in the morning (SL=508 mm) and (B) in the afternoon (SL=551 mm) of 22 December 1988 during the spawning of the Nassau grouper, Epinephelus striatus, from the Bahamas from two reproductively active and aggregating females. The appearance of a mode of larger oocytes in 29B reflects hydration reflecting imminent spawning (Sadovy and Eklund, 1999 – NOAA Tech. Rpt. NMFS 146).