9695 - Laserfiche WebLink

Home Browse Search

n populations in Texas in Siles o 123 123 125 125 124 124 ]21 4 125 52 64 1]6 87 34 121 72 51 123 33 75 ]6 124 42 65 ]8 ]25 105 18 123 101 10 III o 124 124 4 121 125 size per spawn assuming :ings occur and that aU fspring; (B) the observed all families have an equal erved number of pairwise :h family. Scenario B C 2.00 2.00 2.67 2.62 2.67 2.62 4.80 2.8] 2.67 2.64 2.67 2.66 2.67 2.51 3.00 4.18 based on genotype data, however, revealed that only a subset of the breeders present in a spawning tank contributed to the offspring produced. Considering only the number of dams and sires that actually spawned, and assuming that family size per spawning pair was distributed binomially across all spawning pairs, the observed Ne for each of the 13 spawns ranged from 2.0 to 4.8 and averaged 3.02 (Table 2, column B), approximately 34% less than the predicted maximum expected average Ne of 4.55. Parental assignments also revealed considerable variation in family size among spawning pairs (Table 1). As an example, while all six dam X sire combinations in tank BB-7 contributed to the spawn, 98% of the sampled progeny came from two of the three dams, while 97% came from one of the two sires. Similarly, in tank VB4-I, 84% of the sampled progeny came from one of two dams (the third dam did not contribute to the sampled progeny), while 97% of the sampled progeny came from one of two sires. In both of these examples, the actual effective size per spawn was over 40% less than that expected had family size per mating combination been equivalent. Estimates of Ne derived by considering the variation in family size per mating pair per spawn (equation 2) ranged from 2.00 to 4.18 per spawn and averaged 2.59; Table 2, column C). The average estimate of Ne (2.59) based on actual family size per spawning pair is approximately 43% less than the predicted maximum expected average Ne of 4.55. The above results lead to four not-unexpected generalizations regarding the effective size of an offspring population generated in a single spawning event. First, the expected maximum N is not strictly a function of the number of fish in ae spawning tank (given a space limitation in this case of five broodfish per tank) but rather of the number of dam X sire combinations (irrespective of sex) that contribute to a spawn. As examples, compare tank BB-2 (which had four females and one male and a maximum N of 3.2) with most other spawning tanks (three females eand two males; maximum Ne, 4.8) and tank BB-II (two females and three males), for which the expected maximum Ne was the same (4.8) as for most other spawning tanks (three females and two males). Second, assuming that the expected family sizes per mating pair follow a binomial distribution, Ne is a function of the number of dam X sire combinations (families) in a spawning event. Examples here include spawn I in tank BB- 7 NOTE 1331 in size among different families generated within a spawning event. Examples include spawn I in tank BB-2 (two females X one male = two families in the proportions 61.6% and 38.4%; Ne = 2.62) versus spawn 1 in tank BB-12 (one female X two males = two families in the proportions 83.7% and 16.3%; Ne = 2.31) or spawn 1 in tank VB3-1 (one female X two males = two families in the proportions of 91 % and 9%; Ne = 2.00). Finally, the number of actual dam X sire combinations had a proportionally greater effect on reducing Ne than did the variation in family size per spawning pair. The reduction in Ne stemming from the actual number of mating combinations was approxi- mately 34%, while the reduction in Ne further generated by variation in family size was approximate- ly 9%. The bootstrap resampling simulations to assess the effect on NeR (effective size of a released population) when progeny from different spawning tanks (spawns) were mixed to generate a release population demon- strated, as expected, that NeR was reduced when the number of progeny from different spawns in the mixture varied. Briefly, over the spawning year 2003, the mean :!: SD number of spawns contributing to each released population was 2.84 :!: 1.65. The estimate of NeR for the simulated released populations when different spawns were mixed in a release population and when the number of progeny per spawn varied (according to TPWD records) was 5.38 :!: 2.32, whereas the estimate of NeR for the simulated released populations when the number of progeny per (differ- ent) spawn was equalized was 7.17 :!: 4.20. The difference between these two simulation-based esti- mates indicates that equalizing the number of progeny from different spawns in a released population would generate, on average, a 33% increase in NeR' We also generated NeR estimates when spawns contributing less than 10% of the progeny in a mixture were included as is and the contribution of the remaining spawns in the mixture were equalized. The estimate of NeR for these simulated "pseudoequalized" released populations was 6.53 :!: 3.49. The average difference between this estimate of NeR and that when the number of progeny per spawn was equalized (NeR = 7.17 :!: 4.20) was approximately 9%. The purpose of this study was to assess the potential for a Ryman-Laikre effect (a reduction in the effective size of a wild population stemming from the small effective size of II hlltr.h",rv-TP]Pll.PI! nnnll]"tinn I in thp