The Variability and Growth of the Scales of Brown Trout (Salmo trutta) in New Zealand.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 63, 1934, Page 497

The Variability and Growth of the Scales of Brown Trout (Salmo trutta) in New Zealand. By Arthur W. Parrott, Biologist to the New Zealand Freshwater Research Committee, Canterbury College, Christchurch. [Received by Editor, 13th April, 1933; issued separately, January, 1934.] Section I.—Introduction. Since the works of Hoffbauer were published in 1899, many investigators have found it possible to determine, presumably with a fair degree of accuracy, the age of many species of fish by an examination of their scales. The practice has become an important branch of fishery research, especially during the past fifteen years, in all countries that are carrying out extensive research into their freshwater and marine fisheries. The soundness of the scale method depends upon the validity of certain fundamental assumptions, which may be stated as follows: (1) That the scales remain constant in number, and retain their identity throughout the life of the fish. (2) That corresponding to a given increment in the length of the fish there is a corresponding increment in the scales, and that this relative rate of growth of the scales to that taking place in the fish remains constant from the time when the first circulus is laid down throughout the life of the fish. (3) That the bands of narrow and broad circuli are formed at approximately the same time every year (or that some other discoverable relationship exists between their formation and increment of time). Incidentally other questions are raised, but the validity of the scale method of computation is not affected by them. They are given by Van Oosten (1929) as follows:— (1) “Do the annuli represent periods of retarded or arrested growth of the scales?” (2) “Is the growth of the fish in length retarded or arrested at the time of formation of the annuli?” (3) “What factors are responsible for the arrest of or retardation of growth in the fish and scales?” As the life-history of different species, and even different individuals of the same species, varies considerably in the more important details, under different conditions, it is obviously essential that the questions outlined above should be carefully investigated before we can be certain of the validity of the methods we are employing. We are not justified in assuming that that which occurs

in one species will occur in another species, even under similar conditions, or for the same species under different conditions, as has previously been assumed in many cases. Before the application of the scale method for the elucidation of the age, growth, and life-history of a particular species is justified, certain preliminary investigations must be carried out. The study of the age and growth of fishes is of a statistical nature. We are not concerned actually with the age and growth of any particular individual. What does concern us is the average age and growth of certain restricted fish populations, or a particular section of such a population. It is of the greatest importance to give in quantitative expressions variations occurring in their average age and growth-rate over a fairly long period of time, tracing, as it were, certain definite sections of a fish population from birth to maturity, and noting any changes occurring, as measured from previous observations. As the methods are purely statistical in nature, the accuracy of the results depends upon the adequacy of the material upon which the results are based. These methods tend to assume the form of vital statistics. Data are presented, in the present paper, to test the validity of the scale method, as applied to brown trout in New Zealand. Section II.—Collection of The Material. The localities from which material has been obtained for use in the present study are as follows:— (a) North Island (1) Dannevirke (Tamaki River). (2) At various places along the Manawatu River. (b) South Island (3) Selwyn River, Canterbury. (4) Aparima River, Southland. Some of the material was obtained during 1928 and 1929, although approximately 85% was obtained during 1930 and 1931. The following list gives the number of fish in each of the samples: 402 fish from Dannevirke (Tamaki River) obtained during May and June, 1931. 35 specimens from various localities along the Manawatu River taken during 1930–31. 106 fish taken from the Lower Selwyn River at various dates during 1928, 1929, and 1930. 32 specimens from the Aparima River. 405 fry from the North Canterbury Acclimatisation Society's hatchery at Christchurch. 115 comprising five odd samples, of about 20 to 25 individuals in each. The total number of specimens examined in connection with the present study was approximately 1085. The fish taken during May and June from the Tamaki River, together with smaller samples from the Lower Selwyn River and

Aparima River, form the basis of the relative rates of growth of seales and fish. This group of fish, comprising 931 individuals, is fairly representative as regards length, age, and sex. A good part of this material, together with other odd samples, was used for a study on the time of formation of the winter band. For data on the variability and distribution of scales in the different regions of the body, three specimens taken from the Hinds River (Ashburton) were specially suitable (see page 500). The data presented on the early development of the scale were obtained from an examination of fry reared at the North Canterbury Acclimatisation Society's hatchery at Christchurch. A number were set aside during November, 1930, and samples of twenty to thirty individuals were taken every fortnight up to the end of March, 1931. For larger specimens several samples were used that had been collected at odd times during 1928 and 1929. Section III.—Early Development of the scales. The early development of the scale was studied in a series of hatchery-reared brown trout fry. The material consisted of 155 specimens of Salmo trutta. They ranged in length from 2.0 cm. to 8.0 cm. The eggs from which they were hatched were obtained during June, 1930, from fish stripped in the Lower Selwyn River. They hatched during the spring of 1930, and fortnightly samples were taken during December, 1930, January, February, and March, 1932. The fry were preserved and treated for scale observations by the method given by Paget (1920). From specimens preserved in 5% formaldehyde, one-half (the right side) of the skin (epidermis and dermal layers) was removed in one piece and stained with haematoxylin. The results, as Paget also found, were very satisfactory, especially when the fish were placed in weak (30%) alcohol for a day or two previously to staining. In this way the epidermis was removed, thus giving a clear view of the dermal tissues in which the scales are developed. My results, which agree substantially with those of Paget (1920), may be summarised as follows:— The scale papillae were first observed along the lateral line, slightly posterior to the dorsal fin, when specimens were from 2.4 cm. to 2.8 cm. in length (see Fig. No. 1). From this area they gradually developed in an anterior and posterior direction along the lateral line, at the same time spreading dorsally and ventrally, until eventually the whole surface was covered. They tended to spread more quickly dorsally than ventrally. In several instances of fish ranging from 3.8 to 6.0 cm., fully formed scales with one to four circuli were found above the lateral line, and slightly posterior to the middle of the body, while those on the belly and immediately

posterior to the gill covers were either just forming or entirely absent. Further particulars are given in the following table (No. 1):— Fig. No. 1.—Diagram of trout, showing areas from which scales were taken. Area × indicates where scales first appear in young fish. Table No. 1. Av. length of fry (cm.) 2.6–3.0 3.1–3.5 3.6–4.0 4.1–4.5 4.6–5.0 5.1–5.5 Stage of scale development papillae platlets platlets fully formed fully formed fully formed No. of circuli 1 to 3 2 to 5 3 to 7 Diameter of scale 0.34 mm. 0.35 mm. 0.5 mm. Anterior radius 0.13 mm. 0.19 mm. 0.21 mm. Posterior radius 0.21 mm. 0.16 mm. 0.30 mm. It should be remembered that fry living under natural conditions would be subjected to greater variations of temperature and food, which, according to Cutler, Gray, and others, have a direct influence upon scale-growth. In the case of wild fry a relatively wider range in the time of formation and subsequent growth of the scales would be expected. The present data indicate that the length of the fish when the scales first appear is on an average approximately 3.5 cm. Section IV.—Variability of Scales From Different Regions of the Body. Esdaile (1912) has shown in the case of the Atlantic salmon (S. salar) that scales from different parts of the body vary in length and number of circuli laid down. An intensive and systematic examination of the scales from various regions of the body is, as Esdaile points out, an essential preliminary to accurate work. She says: “How can comparisons be rightly made of scales from different fish when the extent of variation on one fish is not known?” Esdaile's work still remains the most complete study that has been carried out on this subject, and for this reason her results are given here in full. Her data were obtained from the examination of three individual salmon, viz.:— (1) 101b. fish taken from the Wye in August, 1910. (2) 171b. fish taken from the Wye in February, 1911. (3) 261b. fish taken from the Wye in March, 1910.

Esdaile came to the following conclusions:— “(1) The number of annuli in each peronidium* Esdaile uses the terms annulus and peronidium. More recent usage replaces her annulus by circulus and her peronidium by annulus. This is indicated when quoting Esdaile by enclosing in parentheses the notation used in the present paper. increases from the head to the adipose fin, on both dorsal and ventral sides of the lateral line, and then decreases again towards the tail.” “(2) A great variation in the number of annuli (circuli) and in lengths of the scales taken from different parts of the same fish is clearly indicated. This was found on each of the three fish, but the results obtained seemed to be in no way correlated.” “(3) In a comparison of scales taken from positions at corresponding distances from the head on both the dorsal and ventral sides of the lateral line, it is seen that, as a general rule, the scales on the dorsal side have fewer annuli in each peronidium (annulus) than the scales from the ventral side.” “(4) I find it impossible to recognise any distinction between different types of annuli (circuli).” It was thought advisable, for the present purpose, to carry out an intensive examination of the scales of three specimens of brown trout, each specimen, if possible, to be exactly the same length and weight. Three suitable fish were obtained in June, 1931, from the Hinds River, Ashburton County. Each of these fish measured exactly 12 in. in length and 15 oz. in weight. The length of the fish was taken from the tip of the snout to the end of the middle ray of the caudal fin. Scales were taken from the right side of the body in transverse rows as shown in Fig. 1. The letters indicate the transverse rows from the anterior (A) to the posterior (F) ends of the fish. The numbers indicate the longitudinal rows from the dorsal (1) to the ventral (6) regions. Each of the areas from which scales were taken is therefore designated by its position on the transverse and longitudinal lines. Thus A 1 denotes that area (approximately ⅓ of an inch square) on the transverse row A and the longitudinal row 1. In each of the areas three scales are included, and the figures in the following tables are the average measurements or counts generally of three scales. It was found in some cases that all the three scales were imperfect, and in some cases a blank appears in the corresponding tables. It was found that when the three sets of tables corresponding to each of the three fish were combined, relatively few blanks remained. Esdaile's tables show very significant variations in the length and number of circuli in the scales taken from practically identical positions (within the same square inch) in the different specimens. This variation is due, in the main, to the large difference in the size of the fish from which she obtained her material. Even by taking corresponding scales from fish of identical length and weight, individual variations will occur in the scales. Esdaile found that the number of circuli and the size of the scales were correlated

either separately or together with the age of the fish from which they were taken. The notation used to designate the various scale measurements is shown diagramatically in Fig. 1. In the following account each of the characters examined will be considered separately. Fig. No. 2.—Graphs showing variation in length of anterior (A) and posterior (B) radii of the scales from different regions of the body. Longitudinal Diameter of the Scales. The longitudinal diameter corresponds to the anterior-posterior axis of the fish. In the tables that follow the scale measurements are given in actual micrometer divisions, eighteen of which equal one millimeter. Table No. 2. Average Length of Scales (18 micro-divisions = 1 mm). Anterior A B C D E F Aver. Posterior 1 42.3 48.3 49.5 55.5 54.0 49.5 49.8 2 49.0 48.5 50.0 56.5 54.5 53.2 51.9 3 54.0 57.0 60.0 59.2 59.0 56.5 57.6 Lateral Line. 4 46.0 48.5 60.2 52.7 61.0 51.0 53.3 5 40.0 42.5 47.0 49.5 54.5 61.5 49.2 6 31.0 42.0 53.0 66.0 59.0 55.5 51.1 Aver. 43.8 47.8 53.2 56.6 57.0 54.5 It is seen from the above that the scales from different regions of the body vary considerably in total length. The anterior scales are generally shorter than those posteriorly situated. Anteriorly,

the scales above the lateral line are larger than the corresponding ventral scales. The most satisfactory scales for age and growth determinations are those situated in the following areas:—C. 1, 2, and 3, D. 1, 2, and 3, and E. 1, 2, and 3. The following reasons are given for this statement:— (1) The scales in this region (save those actually on the lateral line) are the first formed in the young fish. (2) The circuli, and consequently the winter bands, are of a more definite character in the scales from this region than in the scales from other parts of the body. It should be remembered that the above statement holds only for fish in general, for the character of the scales from this area exhibits only relative differences from the scales from other parts of the body. Further, scales from the same area of the body, in two individual fish of exactly the same size (length and weight), may differ more than scales taken from different parts of the body of the same fish. Certain regions of the body are covered with scales which are without exception characteristic in shape and form of circuli. For instance, those situated at the base of the gill-covers, or at the base of the caudal fin, or again those on the belly may be cited. It has been found frequently that such scales may not show the first winter band. The lateral line scales are, of course, characteristic; they are with few exceptions useless for age determination. The abdominal scales, although varying considerably in form, are relatively broad compared with their length. In area they are the largest scales on the body, especially is this apparent in large fish. Length of Posterior Radius. The posterior portion of the scale is that which is exposed when the scale is in position in the scale pocket (see Plate 52) Table No. 3. Average Lengths of Posterior Radius of Scales from Various Regions of the Body. Anterior A B C D E F Aver. Posterior 1 18.7 22.5 22.0 25.5 23.5 (22.4) 2 22.0 20.2 24.0 25.0 24.5 (23.1) 3 23.5 27.5 28.0 24.5 25.0 (25.7) Lateral Line. 4 20.1 21.0 29.5 26.0 26.5 24.0 24.5 5 16.0 18.5 23.0 23.5 22.0 30.0 22.2 6 14.0 23.7 26.0 25.6 28.5 (23.6) Aver. 19.0 22.2 25.4 25.0 (24.3) (27.5) From the above tables it is seen that the posterior radius varies more greatly than the longitudinal diameter of the scale. The posterior portion is generally largest in those scales situated above the lateral line.

Length of Anterior Radius. The anterior portion of the scale is that which is enclosed in the scale pocket. From this part age and growth determinations are made. Table No. 4. Average Lengths of Posterior Radius of Scales from Various Regions of the Body. Anterior A B C D E F Aver. Posterior 1 26.6 25.6 27.7 31.0 29.5 (28.1) 2 25.0 28.0 27.0 25.5 29.5 (27.0) 3 29.6 30.0 24.0 33.7 33.0 (30.1) Lateral Line. 4 26.8 28.0 32.0 31.0 35.0 28.0 30.1 5 28.5 24.7 26.7 27.0 33.0 32.0 28.7 6 17.2 23.5 27.0 39.0 29.0 (27.1) Aver. 25.6 26.6 27.4 31.2 (32.0) (29.7) The anterior portion of the scale does not vary to such an extent as other dimensions. Generally the anterior radius of the scale increases in length from the head towards the tail. The Number of Circuli. From the point of view of scale reading the character of and the constancy in the number of circuli in relation to the length of the anterior radius of the scale are of the greatest importance. There is little correlation between the number of circuli and the size of the scale in an individual fish. On the other hand, there is generally a close connection between the growth of the anterior radius and the production of circuli. This correlation will be shown later. Table No. 5. Average Number of Circuli. Anterior A B C D E F Aver. Posterior 1 44.0 52.0 46.0 44.0 63.0 31.5 46.8 2 50.0 56.0 48.0 60.0 45.0 50.0 51.5 3 54.0 56.7 47.0 47.0 51.0 49.0 50.8 Lateral Line. 4 46.0 57.0 49.0 61.0 56.0 51.0 53.3 5 44.0 41.0 51.0 57.0 66.0 49.0 51.3 6 31.0 30.0 46.0 54.0 58.0 48.0 44.5 Aver. 44.8 48.8 47.8 53.8 56.5 46.4 There is apparently an increase in the number of circuli per scale from the head to a point a little past the adipose fin, beyond which the number of circuli decreases. Those scales above the lateral line and anterior to the dorsal fin possess a larger number of circuli than those below the lateral line. The area of the body, where the scales possess the largest average number of circuli, is that from the posterior insertion of the dorsal fin to midway between the adipose and base of caudal fin, both above

and below the lateral line. The scales from this area are generally large and most suitable for age determinations. There is apparently no significant difference in a dorso-ventral direction. In one specimen of Salmo salar which Esdaile examined the number of annuli (circuli) varied from 76 to 121. She shows that this variation is not due to any one particular peronidium (annulus), but to variations in each of the peronidia (annuli). This variation she states is caused to some extent by definite local differences. “That is to say, speaking generally, the number of annuli (circuli) in each peronidium (annulus) increases from the head to the adipose fin, on both dorsal and ventral sides of the lateral line, and then decreases again towards the tail.” The present data from brown trout bear out fairly well the conclusions that Esdaile came to in the case of the salmon (S. salar). The results of the present investigation may be summarised as follows:— (1) The scales from different regions of the body in the case of brown trout (S. trutta) vary greatly in length and in the number of circuli they contain. (2) This variation is not entirely due to casual variations in the scales, but is caused by definite local differences. (3) It would appear that these differences in the scales are to a more or less degree correlated with the time of scale formation in the young fish. As has been previously shown, the scales first appear along the lateral line, immediately below and slightly anterior to the adipose fin. The scales spread dorsally and ventrally from this area, but the rate at which they spread is slower below the lateral line than above. At the same time they spread anteriorly and posteriorly, ultimately covering the whole surface of the body. In the case of a fry 4.5 cm. long, fully developed scales with one to three circuli were found in the areas C 2, 3, 4, and 5; D 1, 2, 3, 4, and 5; and E 1, 2, 3, 4, and 5; while only minute scale platelets were forming on the remaining areas of the body. (4) The number of circuli laid down in each scale increases from the head to the adipose fin and then decreases towards the tail. No significant increase or decrease was observed on the dorsal and ventral sides of the lateral line. This agrees with Esdaile's findings. (5) A similar distributional variation is shown to occur in the length of the posterior and anterior radii of the scales. (6) The average number of circuli laid down in each scale anterior to the dorsal fin is greater above than below the lateral line; posteriorly to and slightly beyond the adipose fin the reverse is indicated, namely, the ventral scales possess on an average a greater number of circuli. This was not observed by Esdaile in the case of the Atlantic salmon, for

she states: “In a comparison of scales taken from positions at corresponding distances from the head, on both the dorsal and ventral sides of the lateral line, it is seen that as a general rule the scales on the dorsal side have fewer annuli in each peronidium (annuli) than the scales from the ventral side.” Fig. No. 3.—Graph showing variations in number of circuli on the anterior radii of the scales from different regions of the body. Section V.—Formation of Winter and Summer Zones in the Scales. Information on the points to which we now turn is of the greatest value, and should, as Thompson (1904) says, be properly worked out for each species of fish from which it is desired to obtain data by scale examination. It is quite evident that the nature of the winter bands, and the time of the year during which they are formed, varies in individual fish from the same locality, and to a much greater extent in individuals from different localities. The present material does not permit of an investigation being made into the variability of the time of formation of the winter band in fish from different localities. An interesting phenomenon is indicated by the present study, namely, that the North Island fish show relatively wide winter bands on their scales, while those of the South Island, and especially in Southland, show relatively narrow bands.

The data obtained by measuring the increasing distance and number of circuli of the peripheral band on the scale of fish taken each month during the course of a year are given in the following tables (Nos. 6 and 7, and Fig. 4). Fig. No. 4.—Showing formation of summer and winter bands. Fish with two completed bands of narrow circuli on their scales were taken, and the data given relate to the formation of the third band.

Table No. 6. Time of year during which the Bands of Broad and Narrow Circuli are laid down in the Scale. Jan. Feb. March. April. May. June. July. Aug. Sept. Oct. Nov. Dec. Broad Circuli (Summer Band). Peripheral band— No. of circuli in peripheral Band 19.00 18.00 19.25 18.20 4.00 12.95 14.50 18.14 Width of peripheral band 6.91 9.50 14.25 9.00 3.00 7.16 7.77 11.90 Narrow Circuli (Winter Band). No. of circuli 4.50 4.75 6.75 7.50 8.21 14.13 15.75 16.66 Length (in micrometer divisions) 2.67 2.34 2.87 6.11 4.11 7.50 8.30 8.83 Table No. 7. Number of Fish per cent. of each Monthly Sample Showing Broad or Narrow Peripheral Band. Jan. Feb. Mar. April. May. June. July. Aug. Sept. Oct. Nov. Dec. Broad 100.0 67.0 22.0 20.0 40.0 100.0 100.0 100.0 Narrow 33.0 78.0 80.0 100.0 100.0 100.0 100.0 60.0

The information obtained from the above tables may be summarised as follows:— (1) The band of narrow circuli is laid down in the scale generally during the autumn and winter months of the year. (2) The bands of broad circuli are formed during the spring and summer months. (3) The bands of narrow circuli may begin to form as early as February, or, on the other hand, summer growth (broad circuli) may persist until April. The majority, however, show the beginning of a winter band on their scales by March. (4) Early in September the summer band begins to develop, and by the end of October the majority show summer growth to be taking place. In some cases the winter check is formed by only two, three, or four circuli. In such cases there are grounds for believing either:— (1) That growth, at any rate of the scale, has for a period, probably June and July, completely ceased, resuming to produce summer circuli towards the end of August or early in September. This would be expected where winter conditions are exceptionally severe. Or (2) That summer growth has persisted well into the autumn, and very early growth occurred in the following spring. It is probable that the first condition may occur in the southern portion of New Zealand, while the second condition may be the rule in the more equable conditions prevailing in the northern districts. However, the data upon this point are of too meagre a nature to justify definite conclusions. The significance of this nevertheless should not be lost sight of in future investigations. Fig. No. 4, plotted from Table No. 6, shows the rate of production of circuli during the year. We may conclude that the terms summer and winter bands can be applied to the bands of broad and narrow circuli respectively as observed on the scales of brown trout in New Zealand. As a corollary to this we may state that only one band of narrow circuli and one band of relatively wide circuli are laid down during the year. Thus by the enumeration of these alternate bands of wide and narrow circuli the age of the fish in years may be ascertained. It is interesting to correlate the time of formation of the summer and winter zones in the scales with the mean monthly temperature of the water in which the fish have lived. At the present time the necessary data on water temperature are not available for a complete treatment of this subject. Such a study should be undertaken in a particular locality from which scale and temperature data could be collected side by side over a period of several years. In Table No. 6 are given the mean monthly temperatures for four localities, corresponding roughly in latitude with the localities from which the present scale material was obtained. These temperatures are shown graphically in Graph No. 2.

Table No. 8 Mean, Maximum, and Minimum Monthly Temperature from the Various Localities. Locality. Jan. Feb. Mar. April. May. June. July. Aug. Sept. Oct. Nov. Dec. Mean Annual Range. Feilding (N.I.)— Min. 15.0 16.0 11.0 9.0 6.5 5.5 5.0 15.0 13.0 15.0 Max. 20.0 22.0 16.0 14.0 10.0 9.0 8.0 16.5 18.0 24.0 4.8° C. Mean. 17.7 18.7 13.9 11.2 8.0 7.2 6.7 16.0 16.1 18.3 Canterbury— Avon R. (S.I.): Min. 13.0 12.0 13.0 11.0 10.5 8.0 8.5 9.0 8.5 11.0 12.0 13.0 Max. 16.5 15.5 14.5 14.0 12.5 11.5 10.5 11.0 13.0 14.0 15.0 16.5 2.9° C. Mean. 14.5 14.2 13.8 13.0 11.5 9.9 9.7 10.2 11.2 12.3 13.6 14.5 Canterbury— Selwyn R. (S.L.): Min. 9.5 6.0 4.5 Max. 11.0 9.0 7.5 Mean. 9.8 7.6 6.7 Otago— Orona R. (S.L.): Min. 8.5 9.0 7.0 1.0 1.0 1.0 1.5 4.5 5.5 8.0 Max. 15.5 12.5 12.0 5.3° C. 5.3° C. 7.5 3.5 4.5 7.5 9.0 10.0 17.5 5.3° C. Mean. 11.1 10.6 9.9 3.2 1.5 3.0 4.7 6.7 8.2 10.7

Fig. No. 5. Bearing in mind the meagre nature of the data upon which the above table is based, attention may be drawn to the following apparently significant points:— (1) During the spring, summer, and autumn months the temperature at Feilding in the North Island is significantly higher than in Canterbury or Otago. (2) The Otago temperatures are considerably lower throughout the year than the Feilding and Canterbury records.

(3) The increase in temperature during the spring is more rapid in Feilding than in Otago and Canterbury. Bhatia (1932) has investigated the effects of temperature on the width of circuli in the scales of rainbow trout (Salmo irideus). He found that, irrespective of high or low temperatures, abundant nutrition produced broad circuli, and deficient nutrition produced invariably narrow circuli. He states: “The temperature seems to have no effect, except that at low temperatures the fish are less inclined to feed than at higher temperatures, owing, probably, to the lower rate of metabolism. In Nature this fact, coupled with the scarcity of food in winter, must, therefore, be regarded as the causative factor for the production of narrow ‘winter’ rings. Percival (1932) has shown, however, that there is abundant food available for trout in New Zealand throughout the year. In four collections made during July, 1930, in the lower Selwyn River, he found on an average 607 organisms per sq. foot. It would also appear from the work of Percival and Whitehead (1930), on the River Wharfe (England), that there does not occur in English streams a very significant decrease in abundance of aquatic organisms during the winter months. With regard to the opinion that trout feed little, or not at all, at low temperatures, Percival (1932, pp. 18) makes the following interesting statement: “That this is not so throughout the Dominion is evidenced by the fact that freshly spawned female fish taken in early August, from Lakes Taylor and Sheppard, were gorged with Corneocyclas (a small bivalve mollusc), Potamopyrgus, various caddis larvae, chiefly Pycnocentria, Gobiomorphus, and vegetable fragments. A male and female, newly spawned, were gorged with freshly deposited trout-eggs. The water temperature was 7 ¼° C. (45° F.). The indications were that, in the lakes under consideration, fish commenced to feed voraciously as soon as they had recovered from the act of spawning. Section VI.—Growth of the Scale in Relation to the Growth of the Fish. Upon the validity of the proposition, that an annual increment in the length (or some other dimension) of the scale maintains a constant ratio with the corresponding annual increment in body length throughout the life of the fish, depends the accuracy of the scale method. It is essential that data should be available to indicate what correlation exists between the growth of the scale and that of the fish in length. Van Oosten (1929) gives an excellent historical account of this subject. For the purpose of the present investigation, 510 brown-trout scales were measured. The majority of these fish were obtained from the Tamiki River, near Dannevirke, during the winter of 1931. From each fish a sample of scales was taken from the area (X) shown in Figure No. 1. Earlier in this paper the extent of variation exhibited in the scales from different regions of the body was

shown. Considering the variation which occurs in the size of and in the number of circuli in the scales from this area of the body, it was considered that the average of five normal scales from this area would give sufficiently accurate results for the present purpose. In the determinations of age and growth of the trout, we are concerned only with that part of the scale which is enclosed in the scale pocket—namely, the anterior radius. In the present paper only this measurement is dealt with. The results are given in Fig. No. 6 and Table No. 9. Fig. No. 6.—To show the length of the anterior radius of the scale in relation to the length of the fish.

Table No. 9. Showing Mean Rate of Production of Circuli in the Scales of Brown Trout. Length of Fish in Inches Average number of Circuli in Scale. Length of Fish in Inches Average number of Circuli in Scale. 9.0–9.9 52 19.0–19.9 86 10.0–10.9 51 20.0–20.9 94 11.0–11.9 59 21.0–21.9 94 12.0–12.9 56 22.0–22.9 92 13.0–13.9 63 23.0–23.9 100 14.0–14.9 68 24.0–24.9 99 15.0–15.9 73 25.0–25.9 101 16.0–16.9 73 26.0–26.9 104 17.0–17.9 77 27.0–27.9 123 18.0–18.9 81 28.0–28.9 111 Finally, there remains to examine the third, and last, assumption upon which the validity of the scale theory is based—namely, “that the scales remain constant in number and retain their identity throughout the life of the fish.” It is well known that the number of scales situated along the lateral line in any one species of fish is practically constant for the individuals of that species. That is, the number of scales does not increase as the fish increases in length. We have also another line of evidence, in what is known as imperfect centred scales. Normally, a scale has a very small centre; if, however, a scale at some time or other is accidentally rubbed off, another scale will grow to take its place. In these new scales the centres are large; actually the size of the centre is proportional to the size of the scale which was displaced. Such imperfect scales are frequently found, and are obviously of no use for age and growth determination. Section VII.—Summary. (1) Scale papillae are first observed along the lateral line, slightly posterior to the dorsal fin. From this area they gradually develop in an anterior and posterior direction along the lateral line, at the same time spreading dorsally and ventrally, until eventually the whole surface of the body is covered. They tend to spread more rapidly dorsally than ventrally. (2) The average length of the young fish when scales first make their appearance is approximately 3.5 cm. (3) The scales from different regions of the body in the case of the brown trout (Salmo trutta) vary considerably in anterior-posterior diameter, and the number of circuli occurring along the anterior radius. This variation is not entirely due to casual variations in the scale, but is caused by definite local differences, which may to a certain degree be correlated with the time of scale formation in the young fish. (4) The number of circuli laid down in any one scale increases from the head to the adipose fin, and then decreases to the tail.

(5) A similar distributional variation is shown to occur in the length of the posterior and anterior radius of the scales. (6) The average number of circuli laid down in each scale, anterior to the dorsal fin, is greater above than below the lateral line, but posteriorly, and slightly beyond the adipose fin, the reverse is indicated—namely, the ventral scales possess on an average a greater number of circuli. (7) The nature of the winter bands, and the time of year during which they are formed, varies in different individual fish taken from the same locality. (8) The bands of narrow circuli may begin to form as early as February, or, on the other hand, summer growth (broad circuli) may persist until April. The majority, however, show the beginnings of a winter band on their scales by March. (9) The bands of narrow circuli, seen on the scales of brown trout in New Zealand, are laid down generally during the autumn and winter months of the year, while bands of relatively broad circuli are formed during the spring and summer months. (10) Early in September the summer band begins to develop, and by the end of October the majority show summer growth to be taking place. (11) There is a definite correlation between the growth of the fish and the growth of the scales. Acknowledgments. The present investigation was carried out under the auspices of the Fresh-water Research Committee of the New Zealand Acclimatisation Society's Association. The writer wishes to express his thanks to the members of the Committee, especially is he indebted to Professor E. Percival, of Canterbury College, and Mr M. H. Godby, of Christchurch, for valuable assistance and advice in connection with the preparation of this paper, and to Mr D. F. Hobbs, Christchurch, for specimens of trout fry. References. Bhatia, D., 1932. Factors Involved in the Production of Annual Zones on the Scales of Rainbow Trout (Salmo irideus). Journ. Exper. Biology, Vol. 9, No. 1, January, 1932. Creaser, Charles W., 1926. The Structure and Growth of the Scales of Fishes in Relation to the Interpretation of their Life History, with Special Reference to the Sunfish, Eupomotis gibbosus. Museum of Zoology, University of Michigan. Miscellaneous Publications No. 17, Dec. 15, 1926, pp. 1–82, 1 pl., 12 figs., Ann Arbor, Michigan. Esdaile, Phillipa C., 1912. Intensive Study of the Scales of Three Specimens of Salmo salar. Memoirs and Proceedings, Manchester Literary and Philosophical Society, 1911–1912, Vol. 56, pt. 1. Memoir III, 22 pp., 1 pl., 5 diagrs., 4 graphs, Manchester. Gray, J., and Setna, S. B., 1931. The Effect of Food Supply on the Scales of Salmo irideus. Journ. Exp. Biology, Vol. 8, pt. IV, No. 1, Jan., 1931, pp. 55–63.

Hoffbauer, C., 1898. Die Altersbestimmung des Karpfens an seiner Schuppe. Allgemeine Fischerei-Zeitung, Jahrgang, XXIII, Nr. 19, Oct. 1, 1898, Art. 111, pp. 341–343, 2 figs. München. — 1900. Idem. Ibid., Jahrgang XXV, Nr. 8, April 15, 1900, Art. V, pp. 135–139; Nr. 9, May 1, 1900, pp. 150–156. München. — 1901. Weitere Beiträge zur Bestimmung des Alters- und Wachstumsverlaufes an der Struktur der Fischschuppe. Jahresbericht der teichwirth-schaftliche Versuch-Stations zu Trachtenberg, 1901, pp. 50, 3 pls. Breslau. — 1904. Zur Alters- und Wachstumsei kennung der Fische nach der Schuppe. Allgemeine Fischerei-Zeitung, Jahrgang 29, 1904, pp. 242–244. München. — 1905. Weitere Beiträge zur Alters- und Wachstumsbestimmung der Fische, spez. des Karpfens. Zeitschrift für Fischerei, 12. Band (1905), pp. 111–142, 5 figs. Berlin. — 1906. Untersuchungsergebnisse über Alters- und Wachstumserkennung nach der Schuppe. Stenographisches Protokoll über Verhandlungen des Internationalen Fischerei-Kongresses, Wien, 1905 (1906), pp. 131–134. Wien. Hutton, J. Arthur, 1910. Salmon Scale Examination and its Practical Utility, with Notes on the Wye Salmon Fisheries and the Photography of Scales, pp. 56, XXXIX pls., 1910, London. Lee, Rosa M. (Mrs T. L. Williams), 1912. An Investigation into the Methods of Growth Determination in Fishes. Publications de Circonstance No. 63, November, 1912, Conseil Permanent International pour 1′ Exploration de la Mer., 35 pp. Copenhague. — 1920. A Review of the Methods of Age and Growth Determination in Fishes by Means of Scales. Ministry of Agriculture and Fisheries, Fishery Investigations. Series II, Vol. IV, No. 2 (1920), pp. 1–32. 8 diagrs., 1 pl. London. Paget, Geoffrey W., 1920. Report on the Scales of Some Teleostean Fish, with Special Reference to their Method of Growth. Ministry of Agriculture and Fisheries, Fishery Investigations. Series II, Vol. IV, No. 2 (1920), No. 3, pp. 1–24, 4 pls., 3 text figs. London. Percival, E., and Whitehead, H., 1930. Biological Survey of the River Wharfe, II Report on the Invertebrate Fauna. Journ. Ecology, Vol. 18, No. 2, August, 1930. Cambridge. — 1932. On the Depreciation of Trout-fishing in the Oreti (or New River), Southland. Fisheries Bulletin No. 5., N.Z. Marine Dept. Thomson, J. Stuart, 1902. The Periodic Growth of Scales in Gadidae and Pleuronectidae as an Index of Age. Journ. Marine Biological Ass. of the United Kingdom. New series, Vol. 6, 1902, No. 3, pp. 373–375, 1 pl, Plymouth. — 1904. The Periodic Growth of Scales in Gadidae as an Index of Age. Ibid., Vol. 7, 1904, No. 1, pp. 1–109, 8 pls. Plymouth. Van Ooston, John, 1923. The Whitefishes (Coregonus clupeaformis). A Study of the Scales of the Whitefishes of Known Ages. Zoologica, Vol. II, No. 17 (June 19, 1923), pp. 380–412, figs. 137–144, Tables I-VII. New York. — 1929. Life History of the Lake Herring (Leucichthys artedi Le Sueur) of Lake Huron as Revealed by Method. United States Dept. of Commerce. Bureau of Fisheries Doc. No. 1053.

Scale of brown trout.