TPWD 1974 F-6-R-21 #1587: A Comparative Evaluation of Methods for Aging Fishes in Lake Corpus Christi, Texas: Federal Aid Project F-6-R-21

.«ufi FINAL REPORT As Required By FEDERAL AID IN FISHERIES RESTORATION ACT TEXAS Federal Aid Project F-6-R-21 Region 2-C Fisheries Studies Objective 4: Experimental Aging Study Project Leader: Roger L. McCabe Clayton T. Garrison Executive Director Texas Parks and Wildlife Department Austin, Texas Lonnie J. Peters Robert J. Kemp Chief, Inland Fisheries Director, Fish and Wildlife March 31, 1974 A Comparative Evaluation of Methods for Aging Fishes in Lake Corpus Christi, Texas Roger L. McCabe Texas Parks and Wildlife Department ABSTRACT This study compares the usefulness of scales, vertebrae, otoliths, Opercular bones, and pectoral spines for aging studies in southern Texas. The agree— ability of age determinations obtained by two observers was determined for 12 Species of fishes from Lake Corpus Christi. Statistical analyses indicated scales and otoliths were the preferred structures for aging scaled fishes and vertebrae, Opercular bones, and pectoral spines were the best structures for aging catfishes. INTRODUCTION The value of age determination of fishes has been recognized by investigators since the late 1800's (Van Oosten, 1923; McConnell, 1951). The ability to age fish provides biologists with information such as: (l) determination of growth rates, (2) time of sexual maturity, (3) age at which a given Species reaches catchable size, (4) environmental unsuitabilities affecting growth, (4) fluctuation of normal age composition of fish p0pu1ations from year to year, (6) the suitability of subsequent stocking (Lagler, 1956). Techniques generally used for fish aging have been described by Van Oosten (1929), Menon (1950), Rounsefell and Everhart (1953), Lagler (1956), Chugunova (1963), Tesch (1968) and others. The main methods of determining age are analysis of length frequency data and the counting of seasonal rings or zones in various hard body parts of fishes. The first method provides fairly precise estimates of age composition of fish p0pu1ations when large samples are taken. The second method provides the age of individual fishes, but precision of readings depends on the experience of the reader, region of country fishes are collected from, size of focus, sample size, degree of scale absorption, recog- nition of false annuli, etc. Most aging studies-have been confined to northern states where seasonal tem- perature differences are distinct. Growth of fishes in these areas st0ps during the winter and starts in the spring when the water is warm. Usually a well formed annulus or year mark is formed at this time. In the southern part of the United States, the winter season is characterized by periods of summer-like weather which may cause false annuli formation. This problem has discouraged age and growth studies in the South. But, because of the importance of these studies in the fisheries management field, ‘ this study was undertaken to find aging methods that might be used successfully in Texas. Fish samples were collected from Lake Corpus Christi-—a 22,000—acre reservoir located at 280 5I N and 970 28I W or about 4 miles southwest of Mathis, Texas on the Nueces River. The lake was built in 1929 (volume = 300,000 acre- feet; drainage area = 16,656 square miles). It is a shallow lake that does not thermally stratify. METHODS AND MATERIALS Aging structures from 1,316 fishes (12 Species) were collected from 1968 to 1971. Three species of catfishes [blue catfish (Ictalurus furcatus), channel catfish (Ictalurus punctatus), and flathead catfish (Pylodictus olivaris)j and nine species of ctenoid scaled fishes [white bass (Morone chrysoEs), largemouth bass (Microgterus salmoides), warmouth (Lepomis gulosus), redear sunfish (Lepomis MicroloRhus), bluegill (Lepomis macrochirus), longear sunfish (Lepomis megalotis), white crappie (Pomoxis annularis), black crappie (Pomoxis nigromaculatus), and freshwater drum (Aplodinotus grunniensi]'were used in the study. Fishes were captured with experimental gill nets, beach and bag seines, a bottom trawl, and wire traps. Catches were worked fresh or placed in ice and worked within 24 hours after capture. The species name, date of capture, standard and total lengths (mm) and weight (gms) were recorded on individual specimen envelopes. Age estimates were derived from seasonal rings or zones on scales and/or other hard parts. Scales, vertebrae, otoliths and Opercular bones were used in esti- mating the ages of scaled fishes while pectoral spines, vertebrae and opercular bones were used in estimating the ages of catfishes. Collecting, aging and analysis procedures used for each type of structure are given below. Scales: Approximately 20 scales were taken from the left side of the body below the anterior portion of the dorsal fin. Scales were removed by scraping them free with a scalpel. Next, they were placed between two layers of absorbent paper and stOred in a Specimen envelOpe. Before reading, scales were cleaned by washing in warm, soapy water. Impres— sions of the scales were then made on 1” x 3” xCA040” cellulose acetate slides with the aid of a heated hydraulic press similar to the apparatus described by Greenbank and O'Donnell (1948). Scale impressions were projected on a screen with a Model X-lOOO Ken-A-Vision micrOprojector equipped with a 32mm Macro-Tessor lens of 100X magnification. Scale measurements (mm) were taken along an antero-lateral axis from the focus to the outermost point of one corner of the anterior scale margin. True annuli were determined by the following criteria: (1) crossing over of incomplete circuli distally by a continuous circulus in the lateral field (Rounsefell and Everhart, 1953; Lagler, 1956; Carlander, 1961; Chugunova, 1963; Tesch, 1968), (2) a mark preceded by closely Spaced circuli proximally followed by widely spaced circuli distally (Rounsefell and Everhart, 1953; Lagler, 1956; Carlander, 1961), (3) a wide space between series of circuli (Carlander, 1961; Chugunova, 1963), (4) an evident mark continuing as a dark band across the posterior field separating older, blunt ctenii from newer, sharp ctenii (Miller, 1966) and (5) marks where radii bend or end at their juncture (Tesch, 1968). The first three criteria were relied upon most heavily, while the last two were considered less important for designating true annuli. Criteria for determining these annuli were: (1) less distinct and/or less continuous marks than true annuli (Rounsefell and Everhart, 1953; Chugunova, 1963), (2) marks in close proximity to true annuli (Rounsefell and Everhart, 1953), (3) zones of relatively widely spaced circuli proximal to the focus and closely Spaced circuli distally (Sprugel, 1954; Chugunova, 1963), (4) lack of extensive crossing over of circuli and limited extension of the marks across the posterolateral field (Sprugel, 1954; Laakso and COpe, 1956), (5) marks illogically occurring in an area representing rapid summer growth (Laakso and COpe, 1956; Chugunova, 1963), (6) discontinuous circuli distal to year marks (Regier, 1962), and (7) failure of marks to appear on all scales observed (Sprugel, 1954; Miller, 1966). Scale marks identified as false annuli did not meet all of these criteria, but usually more than one was associated with these marks. Vertebrae: Five anterior trunk vertebrae were removed from each fish with surgical scissors or metal snips. The vertebrae were placed in small wire baskets along with other bone structures (Operculars and spines) removed from the same fish. The structures were heated at a temperature slightly below boiling in a water vat for approximately 5 minutes to remove unwanted tissue and bones. Whole vertebrae were placed in petri dishes containing anise oil which made the marks more pronounced. Structures examined dry, in glycerol, xylol, iSOpropol alcohol, or water showed less distinct marks than those in anise oil. These parts were examined with the aid of a variable-power binocular microscope of 10X to 15X magnification. Reflected and transmitted light were used for illumination. Criteria for determining true annuli in vertebrae were narrow dark bands concen- tric with the rim of the centrum.that separated broader light colored zones (Lewis, 1948; Applegate and Smith, 1950), and dark bands accompanied by a ridge on the surface of the centrum. False annuli were identified as faint or incom- plete marks without a ridge on the centrum's surface. Otoliths: Of the three pairs of otoliths (ear stones) present in the inner ear of all teleost fishes, the two largest were used for making age estimates in this study. The otolith method was not utilized for catfishes because of the small otolith size and the difficulty encountered in their removal. Otoliths were collected by removing the lower jaw of each fish with surgical scissors or metal snips. This left the ventral side of the neurocranium floor unobstructed. By scraping the floor tissue away with a scalpel, the enlarged saccular area containing otoliths was located. The sacculus was Opened to expose the otoliths for removal. Otoliths were allowed to air dry before being stored in envelopes. Otoliths were observed whole, with the exception of those from freshwater drum. Due to the thickness of drum otoliths, fracturing was necessary to reveal all year marks. A small chisel and hammer were used to break the otoliths through the nucleus, perpendicular to the long axis. A variable-power binocular microscope (8X to 12X magnification) was used for examining otoliths. Whole otoliths were viewed on their distal (concave) surface and broken otoliths on their cross-sectioned surface. To aid in identification of marks, anise Oil was used as a clearing agent for otoliths, and the structures were placed on a black background and viewed with the aid of reflected light (Schott, 1968). A narrow opaque band, accompanied by a broad translucent band proximally was considered one year's growth. Age was determined by counting the number Of white bands present. Opercular bones: The left Opercular bone was used for age determination in all 12 species of fishes studied. The Opercular assembly was removed from each fish by cutting anteriorally from the dorsal attachment to the eye orbit, ventrally to the corner Of the mouth, and posteriorly along the branchiostegal rays. Removing the entire assembly insured getting the Opercular bone un- damaged. The opercular assemblies were cleaned in a hot water bath to allow the easy removal of unwanted bone and skin (Le Cren, 1947; and McConnell, 1951). Growth zones appeared as Opaque and translucent bands running parallel to the margin of the dorsal, posterior, and ventral portions of the Opercular bone. Broad Opaque zones were periods Of rapid summer growth and narrow translucent bands represented slow winter growth. 'Marks Showing a gradual change from Opacity to tranSparency and ending aerptly at the beginning of another Opaque zone were treated as false annuli. Faint or incomplete marks occurring only along the posterior margin were also judged as false annuli. The proximal side Of the whole Opercular structure was viewed for making age estimates. Annuli were most distinct when viewed with reflected light against a dark background. A variableupower, binocular microscope (7X to 10X magni- fication) was used for viewing Operculars. Also anise oil was used as a clearing agent. Pectoral spines: Pectoral spines were selected for aging purposes since they have been used successfully in studies of channel catfish (Sneed, 1950; Hall and Jenkins, 1952; Marzolf, 1955; Muncy, 1959; Ambrose and Brown, 1971). Spines have also been successfully used for aging blue catfish (Kelly, 1968), and Stur— geon (Probst and COOper, 1955; Chugunova, 1963). Only the left pectoral spine was used for aging in this study. The Spine was removed by cutting through the pectoral girdle, on each side of the articulation, with scissors or metal snips. Unwanted bones and flesh were removed by the hot water method.as already described. Spines were air dried and stored in envelopes. Pectoral Spines were mounted in a fast-drying, clear plastic before cross- sectioning to reveal growth marks. This plastic cast provided a larger, more easily handled Object during the sectioning process. It also prevented spines from splintering. A section of Spine, approximately 0.2—inch thick, was Obtained distal to the basal groove. Both surfaces of the section were polished with abrasive belts mounted on a lapidary wheel. This technique resulted in bone sections that were thin enough to allow light transmittance, yet durable enough for handling and storage. Spine sections were observed under reflected light through a variable-power binocular microscope. Anise oil was used as a clearing agent. Age was determined by counting translucent winter rings. False annuli were identified by their faint or incomplete appearance. Statistical Analyses: Four scale readers were used at the beginning Of the study, but project transfers allowed only two Of the original Observers to complete the investigation. Statistical analyses are based on the observations made by these two individuals. They were fisheries biologists who read and reread the various aging structures until two readings came out the same. These readings were made independently of each other. Aging data from the two Observers were compared by correlation and E-test statistics to determine their degree of association; that is, the extent their Observations agree with one another. Of course, correlations were made accord- ing to Species and method used in aging. Analysis of variance testS‘weKBused to define Significant, statistical differences between aging techniques. RESULTS Scaled Fishes Mean ages Obtained by two Observers aging the same set of fish were Similar for a given aging method (scales, vertebrae, otoliths or Opercular bones), but mean age comparisons between methods Showed less agreement (Table 1). Results of the analysis of variance testing for differences in the mean ages between aging methods and Observers revealed a statistical difference between methods, but not between observers (Table 2). The resulting sums of squares analysis for the analysis variance indicated two non-significant groups of aging methods (Table 3). Vertebrae, otoliths, and Opercular bones formed one group; and they had a smaller coded age mean than scales. Note no statistical difference was found between Opercular bones and scales (the second group). In other words, this analysis does not clearly establish the differences in these methods Of aging fishes. Correlation coefficients obtained by comparing the estimated ages Of scaled fishes made by the two readers are given in Table 4. The corresponding z-trans- formations of the correlation coefficient are also provided. The analysis of variance of z-transformations (Table 5) and the resulting Simultaneous sums of squares test of the means of correlation coefficients indicate a Significant statistical difference between two groups of aging methods (Table 6). Overlap between the two groups does not permit a clear assignment of the group which vertebrae Should be assigned to. However, the correlation coefficients for otoliths (r = 0.916) and scales (r = 0.919) as compared to Operculars (r = 0.763) and vertebrae (0.821) indicate the first two are the most consistent aging techniques. Aging by either Of these two methods should give a correlation coefficient between 0.856 and 0.955 95 per cent of the time. TABLE 1. Mean age (: standard error of the mean) obtained by two Observers aging the same set of fish by four different methods, Lake Corpus Christi data 9 1968-1971. _.a_____a____I_______________a____a____________LL__H_L_______ . 1 1 Mean Mean Species Method n first observer Second Observer _______________________________________________L__________._________~___L__________a.__. Bluegill 1 81 1.26 i_0.21 1.27 i_0.20 2 81 1.09 i_0.18 1.41 i_0.l6 3 80 1.36 i_0.13 1.32 i_0.l4 4 78 0.86 i 0.17 1.01 i_0.18 Longear sunfish l 30 1.87-t_0.28 2.00 :_0.30 2 33 1.06 i_0.23 0.97 t 0.25 3 33 1.24-:_0.l9 1.36-:_0.23 4 ' 32 1.34 i_0.26 1.06 t_0.24 Redear sunfish 1 192 1.69 :_0.13 1.68 t_0.13 2 194 1.51 i 0.10 1.62-: 0.11 3 195 1.67 i_0.ll 1.67 t 0.10 4 185 1.49-:_0.12 1.50-: 0.11 White crappie 1 194 1.69-t 0.11 1.64 i_0.11 2 199 1.43 t_0.ll 1.47 i 0.12 3 _ 193 1.23 i_0.ll 1.31 i_0.12 4 I 189 1.59-i_0.12 1.69;: 0.13 Black crappie 1 114 2.28-i_0.l7 2.32 t 0.18 2 115 1.84 t_0.l7 1.78 i_0.20 3 115 1.77 i_0.18 1.70 + 0.20 4 108 1.98: 0.17 1.91+ 0.20 ______________________________._____.__..__——————-———-——-- um.» n... .....m-- ....-.-r ..., .0....m-..—.---.~.m.,w_mm._ . .._ _ .__..,...,_. .. .- ...-. “4.,__._..... a... ”a...” Mean Me an Species Methodl n first observer second observer _________________L____________m_fl_____________________________*___*__________________ Largemouth bass 1 135 1.57 i_0.20 1.53 i_0.20 2 139 0.97 -|_-_ 0.14 1.00 -I_; 0.14 3 134 l.00-i_0.14 1.01 i.0°14 4 124 1.06-t_0.16 1.18 i_0.16 White bass 1 162 1.90-:_0.14 1.89-i_0.l4 2 159 1.00-: 0.10 1.03 i_0.11 3 156 0.97 i 0.10 1.04 t 0.11 4 150 1.12 i_0.12 1.31 i_0.13 Warmouth 1 20 2.15 i_0.44 2.15 i_0.36 2 - 21 1.38-:_0.26 1.38 i_0.26 3 20 1.70_+_ 0.26 1.75 i 0.26 4 18 1.67 i_0.28 1.72 i_0.30 Freshwater drum 1 130 1.78 i 0.13 1.77 i_0.14 2 128 2.57 i_0.20 2.78 i_0.22 3 134 2.63 i_0.30 2.71 i_0.31 4 122 2.43 i_0.22 2.81 + 0.26 MW 1 Method: 1 = scales; 2 = vertebrae; 3 = otoliths; 4 = Opercular bones. TABLE 2. Results of analysis of variance of coded mean ages Obtained by each observer using four different methods of aging scaled fishes (data transformed by taking the square root of the coded mean ages). W Source of variation Sum of squares df MS F W Method 4.809 3 1.6030 5.497* Observer 0.815 1 0.8150 2.794 Interaction 0.531 3 0.1768 0.606 Error 18.665 64 0.2916 Total 24.820 71 ______—_________________._.______—————————-—-——-—-———— S 0.05