Internet Archaeol. 3 Brinkhuizen Proposal

3. Proposal for a new approach

By reading growth rings from only unpaired or paired elements, the chance of over-representation of a given season will be limited. When we are dealing with paired elements it will be clear that the analysis is, in the first instance, restricted to bones from one side. In order to obtain more accurate results we have to keep some points in mind, namely:

The identification of the skeletal element and the species must be correct.
The question whether this species is suitable for seasonal dating must be answered.

As mentioned at the beginning of this paper, with many species age estimations are based on otolith readings. This has been done because the time involved in retrieving them from fish skulls is short and the results of age determination of many fish species are highly accurate. However, for some species it appears that the results of age reading their otoliths are not very reliable. Dekker (1986) demonstrated this for European eel (Anguilla anguilla) by investigation of the location of the last winter ring.

Thus, at the beginning of November 1984 - just before growth of the winter ring starts - more than 2000 individuals were measured in length and marked in vivo with tetracycline, then released in a pond in a nature reserve. Tetracycline becomes deposited in newly formed bone tissue, and may be detected by its fluorescence for several years after marking. In addition, many of the eels were marked with an external mark. Because of these marks the investigator was certain that he would be studying injected animals. In October 1985 more than 400 eels were recaptured and measured. From this it became clear that the mean growth of the total sample had been 33mm that year. Otoliths from 22 individuals were collected and firstly examined under ordinary illumination to locate the last winter ring. After the observer had decided (on subjective grounds) which ring it was, the illumination was changed to ultraviolet to check whether the winter ring had been correctly identified or not. From this it appeared that under ordinary light only half of the otoliths were read correctly. In addition to this, two eels that had certainly been marked both with tetracycline and the external marker did not show any fluorescent ring at all. It is concluded that if this feature is common for skeletal elements of eel, then the seasonal dating of excavated eel remains loses its validity.

3. An answer is also needed to the question of whether any element of a known species can be used for seasonal dating or if a specific element is necessary.

According to Casteel, scales are useful elements for growth analysis and consequently for seasonal dating. However, a number of objections can be raised against the use of scales alone. Firstly, the fact that the scales form the outer protective layer of the fish and therefore are subject to wear and damage. Furthermore, they are partly of ectodermal origin and have a growth pattern which is more or less independent of the growth pattern of the animal. When modern fish are aged by growth ring analysis, scales are selected carefully and, depending on the species, only scales from certain parts of the fish are used. This information is unknown for scales recovered archaeologically. Additionally, according to Wheeler (1977), excavated scales are often damaged to such an extent that their determination to species is tentative (this is the author's own experience). Thus, unless one is fortunate enough to be in an area where a good proportion of the fauna is made up by families represented by a single species - monotypic families - (Lagler et al. 1962: 118) fish scales are not very promising material. Therefore it is evident that age determination and seasonal dating of fish based solely on scales is highly uncertain.

Reading growth rings from other elements can also give unreliable results. For example, fin rays are often used in fishery research for age determination of freshwater fish. Growth rings are visible in the polished plane of section of the fin ray. Reliable results can be obtained for a number of cyprinids (Abramis brama bream, Cyprinus carpio carp, Leuciscus cephalus chub, Idus idus ide, Rutilus rutilus roach, Scardinius erythrophthalmus rudd, Perca fluviatilis perch, and Stizostedion lucioperca pike-perch). The fin rays used for perch, bream and the roach-like species are from the dorsal fin, for pike-perch from the anal fin, and for carp from the pectoral fins. However, the age determinations obtained with fin rays from pike are not reliable (Deelder and Willemse 1973).

From this example it can be concluded that, even though a particular element of a species is appropriate for age determination of individuals of that species, it does not follow that the corresponding element of another species is appropriate for age determination. The same is true for seasonal dating. Thus, by studying the literature and examining modern material, one has to find out which skeletal elements are the most appropriate ones for use in age determination for each individual species.

4. If these problems can be solved, then age determination may be carried out. However, one must not only look at the outermost ring, but rather examine all rings for regular growth, beginning in the centre. If growth has been regular, it should be possible to see that the rings become closer to each other towards the outer edge of the element. As the individual becomes older, appositional growth diminishes (i.e. there is a gradual reduction in increment growth from year to year). If the pattern is irregular, the bone cannot be used for seasonal dating, as it indicates that changes in fish growth have taken place in the preceding years. This may be due to changes in the biotic and/or abiotic environment, or factors such as an illness, a wound or a heavy parasite burden may have influenced growth.

If a regular growth pattern can be observed, then the outermost ring should be examined. This may be either a winter ring or a summer ring. If the ring appears to be a summer one, and we estimate from the increment width of this ring the time when the individual died more precisely, then it is clear that we have made a basic assumption, namely that in the season of the year in which the individual was caught, the conditions for fish growth were exactly the same as those in the preceding years. From biological and climatological studies, however, we know that this is not the case. In the temperate regions the climatological conditions may differ considerably in corresponding seasons in successive years. This inevitably has its effect on the growth of fish. According to Van Utrecht and Schenkkan (1972: 308) it cannot be expected that the rate of growth and its variations are exactly the same each summer in successive years. This phenomenon will not have been very different in the past. Another point worth mentioning is the fact that many fish species show a differential growth rate between the sexes (Lagler et al. 1962: 175; Pitcher and Hart 1982: 137). This means that during the growth season, individuals of one sex will show greater incremental growth than individuals of the other sex, a difference which will be less pronounced in immature fish.

4. Conclusions

From what has been said above, the question arises whether it is legitimate to divide the increment width of the outermost ring into two, three or more parts for estimating the time of death of the individual more precisely within the growth period established? Is this accuracy disputable? In my opinion, results of seasonality estimations which indicate whether fish were caught during the period of slow growth (winter half-year) or during the period of rapid growth (summer half-year) are reliable. The accuracy of estimations any more precise than these two periods must be considered as very tentative.