The increments appear under the microscope as concentric rings which are alternately clear (continuous zones) and dark (discontinuous zones). Each pair forms a daily growth increment. Sub-zones often appear, i.e. small rings formed with variable periodicity and probably caused by the ingestion of food, environmental variations or stress (Pannella, 1980). Thick and well-marked sub-zones may make increment identification a complicated process. Generally speaking, sub-zones lose clarity or disappear when the focus is changed slightly. It is therefore recommended that you bring the preparation into a focus where all growth structures can be clearly read and then vary the focus slightly to differentiate sub-zone increments.
You need a good microscope and a strong source of light to read the increments: long focal length lenses will allow you to work with thicker preparations, up to 1 mm, making otolith preparation easier. Optical resolution is better when the condenser is positioned as close as possible to the preparation. Polarized light will strikingly increase the contrast between the increment zones.
The magnification for reading will depend on increment thickness, the general range being 400 – 1 000 x. Many species have very fine growth increments undetectable by the light microscope (Morales-Nin, 1989). These fine increments (Fig. 6) may appear as otolith zones without clearly defined growth structures (Mizendo, 1984) but under the SEM they can be seen to be made up of fine increments (± 0.5 μm) (Morales-Nin and Ralston, 1990).
In beginning the study of a new preparation, it is advisable to locate the radius which most clearly shows the increments within a few degrees of magnification, and then move the preparation until the nucleus remains in the centre of the visual field. Using the magnification required for reading, begin the count following the previously established radius. The structural characteristics of the otolith, checks, wider rings, etc. can act as markers, making it easier to localize up to where the count has been made.
It is very tedious and fatiguing and tiring to the eyes to count the increments in the otoliths of fish older than one year. Place a hair from a brush on the eye-piece to get a marker which will facilitate the count; a manual plankton recorder can be used to record the increments counted.
A number of semi-automatic systems have been developed to count increments using a modified image analysis apparatus. These systems can differentiate between the grey levels of the image and identify increments. A sub-routine of the system allows manual correction of identification errors and marking of any increments omitted.
Finally, calculate the total number of increments in the otolith and age the specimen.
Fig. 6. SEM photomicrograph showing fine increments (< 1 μm) deposed during unfavourable growth periods of Sardinella longiceps (scale 15 μm). Observe the cyclical grouping patterns of the increments.
Under the scanning electron microscope, increments appear as crests and sutures due to the different response of the zones to acid. The sub-zones are usually less marked due to the greater continuity of the microcrystals through them. SEM resolution is sufficient to read all increments, no matter how fine (Fig. 7).
The specific microscope characteristics will determine the working distance, the degree of inclination and the voltage to employ. Low voltages (15 KV) are recommended to avoid problems with specimen conductivity during the reading.
In measuring increments, the specimen surface must be positioned horizontally to the electronic face (a specimen positioned at a slant will produce distortions that will affect the measurements). Increments can be measured in photographs or on the microscope screen using acetate grid paper. The television signal of the SEM can be recorded with a video-recorder and later analysed on a TV screen.
Fig. 7. Fine increment depositions on an Engraulis ringens otolith (scale 27 μm).
Age interpretation is a three-part process:
1. Identification of growth structures:
Several otoliths must be examined to define the increments whose periodicity will be considered daily. An increment can be counted when you can follow it around most of the otolith. When a discontinuous or check zone is found on the edge, the increment is to be considered as in the process of formation and is not counted. Increments on the edge of the otolith, however, are often masked by the optical distortion produced by the angle of refraction between the edge and the medium, which makes it difficult to count marginal increments.
It is advisable to begin to read otoliths of younger specimens, and, with practice, gradually progress to the otoliths of older fish.
2. Definition of interpretation criteria:
In starting to study a new species, the interpretation must be confirmed by several scientists who repeat the otolith reading and discuss the criteria for interpretation (Gjosaeter et al., 1983). A few otoliths covering the range of sizes to be studied are selected, interpreted and reinterpreted. This process allows the best area for otolith interpretation to be identified, as well as checking reader consistency and consensus. Variability among readers can be determined by some statistical test such as the t test. Unless agreement is good, the criteria for interpretation must be revised. Where interpretation and discussion do not produce reader consensus, this may be due to poor-quality preparations. The comparative analysis of the findings will show the type of error involved (Gjosaeter et al., 1983).
3. Age reading from a stock sample:
Once the otolith structures have been identified and the interpretation criteria defined, the individuals in the stock sample can begin to be aged. For maximum objectivity neither the size nor other data about the otolith must be known. Each otolith, identified by a reference number, must be read at least twice and only readings which coincide can be considered valid. Any discrepancies will depend on the age of the fish: the more increments there are the greater the discrepancies.
Ralston and Miyamoto (1981; 1983) and Ralston (1976; 1985) presented a method for ageing based on otolith increment thickness. Since otolith size depends on the age of the fish, which is closely correlated with length, increment thickness will reflect bodily growth (Campana and Nielson, 1985; Gutiérrez and Morales-Nin, 1986).
If otolith formation were constant, increment thickness would be uniform throughout the otolith. After determining the otolith growth rate, the age would be determined by dividing otolith size by the rate of growth. But as fish and otolith growth are both variable, this method can only be applied to small segments of the otolith where growth is considered uniform.
The segments usually used are 500 μm, and should not exceed 2% of the radius of the otolith (Ralston and Miyamoto, 1983).
The method involves the following steps:
Preparing the otolith for viewing with the light microscope.
Measuring the radius of maximum otolith growth starting from the focal point (R), where the otoliths are to be aged and measured.
Calculating the ratio between the otolith radius (R) and fish length (FL).
In each zone where the increments are visible: a) count the number of increments (t), b) measure the size of the zone, c) measure the distance (r) between the otolith nucleus and the mid-point of this zone.
The data for each otolith are summed up at 500 μm intervals from the radius r of each segment.
The linear relationship between the sets of data pairs dr/dt and r is calculated logarithmically for each otolith. The equation:
In(dr/dt) = c-br + a
where r is the length in millimetres, t days (number of increments), c, b, are regression constants, and a is a variable.
Fish age is determined by the integration of:
dt = c'ebrdr, c'= e-b
where fish age (T) is: T = c'/b(ebrr).
The Delta method (Seber, 1973) is used to find the integration.
The method has been critically evaluated through ageing by annual rings, Monte Carlo simulations, length frequency analyses and studies of spawning periods related to birthday determined by back calculation (Ralston and Williams, 1989).
Counting otolith increments in adult fish is extremely tedious and difficult and so this method is particularly useful for a semi-automatic application using an image analyser recently developed by Ralston. The method's major limitation is that in some species increment depositions during periods of slow growth are extremely fine (Fig. 7) and are therefore undetectable by light microscope. The zones with this type of increment will appear to lack clear growth structures and will not be included in the age calculations. Age will be underestimated if only the thicker increments are used, and this will give exaggeratedly high growth rates (Morales-Nin, 1989; Morales-Nin and Ralston, 1990).
Many authors have reported the presence of seasonal growth rings (annulae) in tropical fish otoliths (Brothers, 1979; 1982; Sainbury and Whitelaw, 1984; Samuel et al., 1985) although ring deposition in some species is irregular (Mathews, 1974). Once the annual ring periodicity has been determined, age reading is rather simple.
Otolith rings in tropical species are rather less clearly defined than in cold-water fish: the predominance of transparent zones requires methods which can bring out the contrast between the opaque and the hyaline rings. For interpretation, the otoliths can be immersed in a dense clarifying liquid (clove oil, liquid paraffin, glycerine) in a dark container. Reflected light and a binocular microscope are used for the reading. Slow growth rings will appear dark through the dark transparent background of the container, whereas fast or opaque growth rings will appear light under reflected light. Move the light focus and alter the positioning of the otoliths to help differentiate the growth rings: avoid fixed mounting media which will not let you touch the otoliths.
Thick, opaque otoliths can be interpreted after they have been left for some time in a clarifying liquid such as water or glycerine. The otoliths can be positioned in numbered ice-cube trays containing the liquid selected and left to clarify. The time depends on the species but 12–24 hours is enough for most. Another technique is to heighten the contrast between the growth rings by burning the otoliths (Christensen, 1964). The slow, hyaline growth rings which contain more protein (Casselman, 1974) acquire a darker, caramelized tone when burned. The otolith is burned on a metal sheet under a low flame (Bunsen burner) or in an oven at 100°C. Burning time depends on otolith size and flame heat. The process must be carefully monitored to avoid burning the otolith and thereby losing it. The otoliths may be stained with a protein-affinity dye.
If necessary, otoliths can be polished or sliced to read the macrostructural growth pattern (Fig. 8). Generally speaking, fish which reach a great age and grow slowly can be aged more precisely in sections than by reading the whole otolith. Reading whole otoliths for young specimens and sections for larger fish can facilitate the study and produce good results.
Ring periodicity can be read by following the progression of the rings formed on the edge of the otolith throughout the year. A graph of the monthly percentage of otoliths with opaque and hyaline borders will show the ring deposition period. In the case of annual formation, one maximum per year will be found for each type of ring (Fig. 9).
Once the daily periodicity of the increments has been determined for a species, the number of annulae increment components can be used to determine how frequently these growth structures form. When the number of increments in one opaque ring and one hyaline ring does not differ significantly from 365, annulae formation can be considered to be annual. This method is applicable even when daily increment periodicity has not been determined to obtain an approximation of annulae periodicity. In this case, however, the results obtained must be validated by other methods.
Fig. 8. Sagitta of Merluccius capensis showing seasonal growth rings. The slow (hyaline) growth rings appear darker (arrows) under incidental lighting. In the upper part of the figure there is a cross-section of the otolith. Note the numerous false (non-seasonal) rings which form in this species.
Fig. 9. Yearly variation of the percentage of otoliths with a rapid (opaque) growth ring on the edge.
After measuring the distance from the nucleus to each ring, ring distances can be plotted for each age group. This will give a unimodal distribution in approximately the same places for each of the growth rings (Fig. 10) when ring formation is regular and corresponds to a seasonal growth pattern common to the stock (Manooch III, 1987; Taubert and Coble, 1977).
After counting the rings and determining their annual periodicity, the age has been established. A fish's age is the interval between birth up to a given point in time, usually capture. The age class or annual group is determined from the birthday. The birthday of a given fish is unknown and therefore arbitrary birthdays are used for an entire stock based on maximum spawning or other decisive factors in recruitment. The standard birthday in the northern hemisphere is 1 January and in the southern 1 July. Other dates can be used as convenient, however.
