Danilewicz et al.: Reproductive biology of male Pontopona blainvillei from Rio Grande do Sul, southern Brazi 



583 



(a commercial mixture of acids) and sectioned on a freez- 

 ing microtome. The 15-20 fim sections were stained with 

 Mayer's hematoxylin and mounted on microscope slides 

 with Canadian balsam or in glycerin. Poor and off-center 

 sections were discarded in favor of new preparations. 

 Three readers counted independently the number of 

 growth layer groups in both the dentine and cementum. 

 When reader estimates differed, the sections were reex- 

 amined together and a best estimate was agreed upon. 

 In this study, we considered one GLG to represent one 

 year of age, which is the accepted model for the francis- 

 cana (Kasuya and Brownell, 1979; Pinedo, 1991; Pinedo 

 and Hohn; 2000). 



Reproduction 



In the laboratory, the testes were separated from the 

 epididymis, weighed to the nearest 0.01 g (n = 107), and 

 measured in three dimensions (length and two diam- 

 eters perpendicular to each other in the middle of the 

 testis) to the nearest 0.1 mm (;? = 104). The mean of 

 these two diameters was called mean testis diameter. 

 The weight of one of the gonads could not be recorded on 

 some occasions (« = 8) and we assumed that both testes 

 had the same weight. Then, relative testis weight was 

 determined as the ratio of the combined testis weight to 

 the animal weight. 



A 1-cm 3 subsample of each testis from the central 

 portion of the organ was removed and examined by 

 using standard histological preparations. The tissue 

 was embedded in paraffin, sectioned in 4-10 fim thick 

 slides through a manual microtome, and stained with 

 hematoxylin and eosin (H&E). Male sexual maturity 

 status was determined by examining the testicular 

 sections at a magnification of lOOx. In this study, we 

 followed the classification criteria suggested by Hohn 

 et al. (1985): 



1 Immature — seminiferous tubules containing main- 

 ly spermatogonias. Abundant interstitial tissue 

 present between the seminiferous tubules and lu- 

 men totally closed. 



2 Pubertal — seminiferous tubules containing sper- 

 matogonias and spermatocytes. Less interstitial 

 tissue present between the seminiferous tubules 

 than in immature animals. The lumen is partially 

 opened. 



3 Mature — seminiferous tubules containing sper- 

 matogonias, spermatocytes, spermatids and, in 

 many cases, spermatozoa. Interstitial tissue al- 

 most nonexistent between the seminiferous tu- 

 bules. The lumen is totally opened. 



The diameters of ten random circular seminiferous 

 tubules were measured for each specimen (« = 93) with 

 a scale present in the lens of the microscope in order 

 to calculate the seminiferous tubule mean diameter. A 

 maturity index (MI) was calculated as the ratio of the 

 combined testes weight by the combined testes length 



aw/iL). 



An analysis of the variation along the year of the val- 

 ues of relative and combined testes weight, and seminif- 

 erous tubule mean diameter, was employed to assess re- 

 productive seasonality. Values of these parameters were 

 compared between months when mating and conception 

 occur ("reproductive months": November-March) and 

 months when they not occur ( "nonreproductive months": 

 April-October). In order to increase the sample size of 

 mature animals collected in reproductive months, data 

 on testes weight from mature male franciscanas from 

 Uruguay were included in the analysis (data supplied 

 by Kasuya 1 ). 



The mean age at attainment of sexual maturity (ASM) 

 was estimated through the DeMaster (1978) method and 

 the logistic regression. 



The DeMaster (1978) equation computes the mean 

 as 



ASM = £a(/ Q -f^), 



where f a = the fraction of sexually mature animals in 

 the sample with age a; 

 j = the age of the youngest sexually mature 



animal in the sample; and 

 k = the age of the oldest sexually immature 

 animal in the sample. 



The variance of the DeMaster method estimate is cal- 

 culated as 



k 

 variASM)=^[if a a-f a )/N a -D], 

 «=j 



where N = the total number of animals aged a. 



The logistic regression approach fits a sigmoid curve 

 representing the probability that a franciscana of age 

 a is sexually mature to the distribution of sexually 

 mature and immature animals by age as 



Y = l/(l+e° +/,v ) or In (1/1-1) = a + bx, 



where x = the age of the dolphin; 



b = the slope of the regression; and 

 a = the intercept. 



To obtain the age when 50% of the animals are sexu- 

 ally mature (Y=0.5), the last equation is simplified as 

 ASM = -alb. 



Mean length and weight at sexual maturity was also 

 estimated by the DeMaster (1978) method, by substitut- 

 ing age for length and weight, respectively. The meth- 

 od was slightly modified, as suggested by Ferrero and 

 Walker (1993), and was calculated as 



1 Kasuya, T. 1970-73. Unpubl. data. Teikyo University of 

 Science and Technology. Uenohara, Yamanashi Prefecture, 

 409-0193, Japan. 



