TPWD 1968 F-6-R-15 #1199: Job Completion Report: The K Factor Index, KI; A Qualitative Measure of Fish Populations, Federal Aid Project No. F-6-15, Fishery Investigations - Region 5-B, Job No.

--- Page 1 --- JOB COMPLETION REPORT As required by FEDERAL AID IN FISHERIES RESTORATION ACT TEXAS Federal Aid Project No. F-6-15 FISHERY INVESTIGATIONS - REGION 5-B Job No. B-26 (Seg. 4) The K Factor Index, KI; A Qualitative Measure of Fish Populations Project Leader: John C. Barron J. R. Singleton Executive Director Parks and Wildlife Department Marion Toole Eugene A. Walker D-J Coordinator Director, Wildlife Services June 11, 1968 --- Page 3 --- JOB COMPLETION REPORT State of Texas Project No. F-6-15 Name: Fishery Investigations - Region 5-B Job No. B-26 Title: K Factor Index _ __ Objective: To develop a method by which fish populations can be qualitatively measured through the use of K factors. Procedures: Originally the K Factor Index was to be a collection of average K factors, corrected to eliminate certain inequities, from which field personnel could measure water productivity from netting samples with the use of desk calcu- lators. As the index evolved, 1 became convinced that this approach was impractical. I believe that the best procedure is to store regional length- weight values on computer tape and have field personnel send their data to the Statistical and Data Processing Section in Austin. T could figure KI (automatically up-dating the regional values during the process), print the results, and return them to the originator. During this segment I have secured permission to work toward this goal. ere the machine Since ADP methods are to be emploved, we can by-pass the K factor calcu- lation entirely and work simply with lengths and weights. By definition, the K Factor Index is: KI = Sum P (tq 4)°fi4/N fl Where P probability t = the standardized unit of Student's distribution i the species in the ith cluss j = non-productivity correction term in the jth class N = total number in sample £ = frequency in the sub-class --- Page 4 --- The change will be in the computation of the variable t in the equation. Originally t was: t = Sample K - Regional K / Sample Standard Error. The regional means being highly variable were smoothed by a moving average which allowed them no dispersion values. Since K was derived from the function w= kK°(L)” where W is weight, L is the length, and the exponent n held constant at 3, the computer can easily handie the linear form log W = log K + n (log L). The value of t will now be computed by comparing regression values of log weight on log length. The significance of the mean weight of the sample compared with the regional mean (both adjusted to the regression values at the pooled mean length) will be tested by t = Sample Adj. W ~ Regional Adj. W / Pooled St. Error. The use of the length-weight regression will eliminate the use of length intervals. 1 have also eliminated the stages of sexual development as non- productivity correction terms and substituted in their place monthly correction terms for males of each species and females of each species. I feel that this will minimize most of the weight variability associated with factors other than water productivity. Results: To illustrate the use of the index, a sample problem is shown. Table 1 shows part of a gill net collection taken at Delta Lake in November 1966. Those fishes which were not sexed are not included and the freshwater drum are not included due to insufficient regional length-weight values for November. Since regression analysis requires at least three individuals in each sample, those species and/or sexes which did not have three members were not used in the calculation of the index. The Delta Lake collection was tested against the regional values and the results are shown in Table 2. Only one of the three game fish species appears to be thriving: the white crappie. This in itself is significant, since several years ago crappie disappeared from this lake after providing a number of years of good fishing. A massive stocking program restored their numbers, and this high KI-value proves its effectiveness. I am not surprised that white bass are doing poorly, however, since this lake is not one in which they normally live. Both male and female blue catfish show unusually low values of KI. The population seems large and is reproducing, and the only thing to which I can attribute this low condition is parasites. Data from the collection sheets list a heavy infestation of internal parasites. --- Page 5 --- Table 1 Gill Net Collection from Delta Lake | St. L. Weight St. L. Weigh Species Sex (mm. ) (gm. ) | Species Sex (mm. ) = | \ | Longnose gar F 720 1318 White bass M 169 | F 855 3005 F 176 | M 624 1233 F 172 F 639 1049 M 197 | M 518 638 F 159 | M 514 510 M 170 | M 692 1588 M 166 | Blue catfish F 522 1828 | White crappie F 192 209 M 392 907. | F 229 356 | M 274 273} F 179 172 F 203 114 | M 183 184 | M 201 108 M 172 156 | F 197 97 | M 173 155 | F 199 115 | F 177 174 | F 257 236 | M 174 157 | M 230 160 | M 188 175 F 229 155 | M 173 161 i M 199 109 F 179 184 M 225 149 F 191 208 M 196 95 F 174 165 F 221 146 M 167 144 M 162 130 White bass M 256 510 M 171 158 | F 209 251 | M 163 131 | M 207 248 M 124 55 | F 203 231 | !Gizzard shad M 237 253 | F 199 166 M 192 126 F 173 105 F 163 B40 F 141 47} i M 140 50 |j i F 133 4300 M 138 49 | F 133 420 i --- Page 6 --- Table 2 Computation of KI for Delta Lake Collection tf Noe | t- Value ___ Probability | | Longnose gar males | 4 0.37 64.4% females 3 -0.58 28.1 Gizzard shad males DL 30.5 225 98.8 2 | | - “1.55 | 6.0 | | = females lon a t ho oO i | Blue catfish | males females White bass 39.4 { | | | i { { i males 6 females 30.2 i i 100.0 White crappie males | { | | | females 100.0 KI Ul 4(64.4) + 3(28.1) + 4(30.5) + °*° + 7(100.0) / 4434+44°°°+7 3,329.50 / 60 55 She monte hei oR tS ht Aer teen --- Page 7 --- “- , r Over-all the K Factor Index for the lake is 55.5%. Lf it were equal with the regional values, it would be 50.0%. Much of the KI is contributed by the crappie, and without them the total is only 36.4%. Both rough fish species tested appear to be in above average condition, if the sexes are combined. None of the values of this test should be con- sidered as firmly established due to the small size of the sample, but their pooling most likely will be very indicative of the total KI. What remains to be done now computation-wise is to determine the theoretical distribution of KI so that confidence limits can be found. If that can be done, lakes can also be compared with one another as with regional means. oy ip OPP Da OY x L Prepared by John C. Barron Approved by “ A) Api 7 OD a Project Leader , aad alain 0 7 9 eeeaeemaiaaeiall Project Leader Coordinator Date ___Jume 1}, 1905 a gin M,C. Dietz Inland Supervisor