BARLOW: REPRODUCTIVE RATES OF SPOTTED DOLPHINS 



Variability in percent mature with sampling con- 

 ditions may result from several interacting factors. 

 Preliminary data have indicated that spotted 

 dolphins in the ETP may segregate on the basis of 

 reproductive maturity (A. A. Hohn fn. 4 and M. D. 

 Scott fn. 5). Schools that consist principally of im- 

 mature dolphins may have a smaller characteristic 

 school size, may be less likely to have large numbers 

 of tuna associated with them, and may be more 

 vulnerable to high kills-per-set due to the inex- 

 perience of younger dolphins. Also, the consistent 

 underrepresentation of immature age classes in the 

 spotted dolphin age distribution (Barlow and Hohn 

 1984) indicates that a very significant bias may oc- 

 cur in the sampling of immature animals. These are 

 largely speculations, and until a well-supported ex- 

 planation for sampling variability is presented and 

 until some method of removing this bias is found, 

 percent mature should not be used as an index of 

 changes in spotted dolphin reproduction. 



Gross Annual Reproductive Rate 



Changes in GARR have been used as a measure 

 of changes in the net rate of growth for a popula- 

 tion (Smith 1983). This approach has been faulted 

 on the basis that it does not consider age structure 

 effects (Polacheck 1982), and more critically on the 

 basis that such an approach is theoretically unsound 

 (Goodman^). These criticisms do not, however, 

 detract from the usefulness of GARR as an index 

 of gross per capita reproduction for a popula- 

 tion. 



If GARR were robust to sampling conditions, it 

 could be one of the most useful indices of popula- 

 tion reproduction. One advantage is that a GARR 

 index considers percent pregnant and percent 

 mature simultaneously, and hence compensatory 

 changes in these two do not affect the index. Sim- 

 ply stated, GARR is calculated as (the fraction of 

 females in a population) x (the fraction of females 

 that are mature) x (the fraction of mature females 

 that are pregnant)/(gestation time). Unfortunately, 

 percent mature is a major component in these 

 calculations, and this parameter has been found to 

 be dependent on sampling conditions. Until sampling 

 problems associated with estimating percent mature 

 are resolved, GARR is not likely to be a useful in- 

 dex of change in reproductive rates. 



Between-Population Comparisons 



The northern and southern stocks of spotted 

 dolphins have been subjected to very different levels 

 of fishing-related mortality. Smith* has estimated the 

 northern stock to be at 38-55% of its 1959 level and 

 the southern stock to be at 93-98% of its historical 

 level. Density dependent increases in reproductive 

 rates might be predicted for the northern stock 

 relative to the southern stock. 



The percentage of mature females that were preg- 

 nant differs significantly between the northern and 

 southern stocks. Surprisingly, however, the southern 

 stock was found to have the higher percent pregnant 

 (36% vs. 33%). Another exploited population of spot- 

 ted dolphins in the western Pacific was found to have 

 an annual pregnancy rate of 0.254 (Kasuya 1976), 

 which (with a gestation time of 11.2 mo) would give 

 an average percent pregnant of about 24%. Con- 

 siderable variability in percent pregnant can thus 

 exist between spotted dolphin populations, none of 

 which is obviously related to density compensatory 

 effects. Sampling of the southern population has, 

 however, been sporadic, and if annual variability in 

 pregnancy rates is greater for that stock, a few years' 

 data may not be sufficient to accurately estimate a 

 long-term mean. Nonetheless, the tendency for a 

 more exploited stock to have lower reproductive rates 

 is worrisome, and future life history comparisons 

 between the northern and southern stocks would 

 probably be useful. 



Evidence for density compensatory changes in 

 pregnancy rates were also lacking when two spin- 

 ner dolphin, 5. longirostris, populations were com- 

 pared (Perrin and Henderson 1984). They found 

 similarly that the more heavily exploited stock 

 (eastern spinners) had a lower percent pregnant than 

 the less heavily exploited stock (whitebelly spinners). 

 The opposite would be predicted based purely on 

 density compensatory effects. 



The overall percentage of pregnant females that 

 are lactating is significantly higher for the north- 

 ern spotted dolphins than for the southern stock. 

 The biological significance of this result is ques- 

 tionable given the year-to-year variability in this 

 parameter. Between-population comparisons of this 

 percentage are not likely to be meaningful until the 

 cause of this large annual variability is identified. 



As was noted above, the percentage of females that 



'Goodman, D. 1984. Uses of the gross annual reproduction rate 

 calculation in the dolphin assessment. Admin. Rep. LJ-84-22C, 17 

 p.; available from Southwest Fisheries Center La JoUa Laboratory, 

 National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, 

 CA 92038. 



*Smith, T. D. (editor). 1979. Report of the status of porpoise 

 stocks workshop (August 27-31, 1979, La Jolla, California). Ad- 

 min. Rep. LJ-79-41, 120 p.; available from Southwest Fisheries 

 Center La Jolla Laboratory, National Marine Fisheries Service, 

 NOAA, P.O. Box 271, La Jolla, CA 92038. 



667 



