This study was performed in the Laboratory of Carnivore Reproduction at the School of Veterinary Medicine, State University of Ceará and approved by the Animal Ethics Committee of the institution (protocol No. 12641034–8).
Ten post-pubertal dogs (2 Labrador, 4 Rottweiler, 4 German Shepherd) aged 2 to 4 years and weighing 33 to 42 kg, and 10 pre-pubertal dogs (4 Labrador, 3 Border Collie, 2 Golden Retriever, 1 Flat-Coated Retriever) aged 6 months and weighing 15 to 18 kg were used for the study. The dogs used in this study Belonged to São Lazaro Kennel 4° Cia de Choque/CPCÃES and the Estancia Kirst Kennel. A consent statement was signed by those responsible for the animals used in this research, explaining the purpose of the study and ensuring the welfare of animals.
Veterinary examination and complete blood count at the beginning of the study confirmed that all dogs were clinically normal and healthy. All dogs were fed a maintenance complete dry food with ad libitum water for the duration of the study.
Breeding soundness evaluation
Dogs were selected from a larger group and only those with no previous reproductive disease and normal external genitalia were used ; all pre-pubertal dogs had not been bred and had small external genitalia but otherwise had a normal reproductive tract upon clinical examination.
Three ejaculates were collected from each of the post-pubertal dogs by digital manipulation at 7 day intervals during the study and the second fraction of the ejaculate was subjected to detailed examination. Attempts to collect semen from the pre-pubertal dogs were not successful. The second fraction was assessed immediately and volume was recorded. Subjective microscopic assessment of the percentage total sperm motility was made at ×400 magnification at room temperature. Sperm concentration was measured using a Neubauer chamber after dilution with formal-saline . Membrane integrity was evaluated at ×400 magnification using the hypo-osmotic swelling test  and sperm morphology was evaluated at ×1000 magnification on Rose-Bengal stained slides , in each 200 sperm cells were evaluated.
Three ultrasound examinations were performed on the right and left testis of each dog (after semen collection) with 7 days intervals, using a SonoAce PICO machine (Medison, Korea) with a linear array transducer with 5 to 9 MHz capability. Dogs were positioned in dorsal recumbency, acoustic gel was applied to the skin, and the transducer was positioned initially on the lateral surface of the testis. Longitudinal and transverse B-mode images were made (using the mediastinum as a reference point for measuring the testicular length and width) and testicular volume was calculated using the formula for an elipse; V = length × width × height × 0.5236 . The appearance of the testicular parenchyma was recorded subjectively as heterogeneous or homogenous.
For the measurement of testicular artery flow in three separate regions, colour Doppler ultrasound was used with the transducer initially placed at the neck of the scrotum to identify the tortuous distal (looping) region of the supra-testicular artery (here termed distal supra-testicular artery), immediately cranial to the cranial pole of the testis. The transducer was then moved distally in the front (dorsal) plane to identify the marginal region in longitudinal section (here termed the marginal artery testicular artery), and the relatively straight intra-testicular arteries within the testicular parenchyma (here termed intra-testicular arteries). Within each region the colour gain was adjusted to reduce any excess colour noise and the pulsed Doppler gate was positioned within the lumen of the vessel. Three waves of a cardiac cycle were used to measure mean values for peak systolic velocity (PSV), end diastolic velocity (EDV), and these were used by the machine software to calculate resistance index (RI) and pulsatility index (PI). The sample gate ranged from 2.0 to 3.0 mm across all three regions and for two adult dogs was 5.0 mm for the supra-testicular artery only. The angle of insonation used was 0°. The same operator performed each examination.
The time taken to conduct the clinical, breeding soundness and ultrasound examination was recorded.
Data were tested for normality (Shapiro-Wilk test) and homoscedasticity (Levene test). Data for semen quality, testicular volume and Doppler ultrasound measurements had a normal distribution; thus, analysis of variance of repeated measures was performed for the comparisons between weeks in each group evaluated, and the paired t test was used for the comparisons between the right and left testes. The t test was performed to assess the possible differences between the groups.
There were no significant differences (p > 0.05) between the left or right testes or between the examination time points, and so data were pooled as mean values for each dog. Pooled mean Doppler measurements between the different regions did not show homoscedasticity and so differences were assessed using the Friedman test. A significance level of p < 0.05 was used in all cases, and the results were expressed as the mean ± standard deviation.