Yousef et al (1995): Following an initial six-week observation period, male NZW rabbits (4/group, 8 months old, mean initial bodyweight 2863 g) were given oral doses of glyphosate (from Monsanto, USA) in gelatin capsules for six weeks at 1% or 10% of the LD50. The study authors did not explicitly identify the dosing interval or specify the doses in terms of mg/kg bw. The rabbits were then held without treatment for a further six weeks to study reversibility of effects. The animals were weighed weekly in the morning before access to feed and water. Semen was collected weekly throughout the study using a teaser doe and artificial vagina, with ejaculate volume being recorded after removal of the gel mass. Semen osmolality, fructose concentration and methylene blue reduction time was measured together with sperm concentration and assessment of live, dead and abnormal spermatozoa.
No information was provided on survival of the test and control animals, but for reasons unknown, one rabbit was removed from the control, low and high dose groups during the recovery period. Other than stating that most treated animals showed indications of reduced libido (especially at the high dose), the study authors did not comment on clinical signs. Control mean bodyweight increased by ca 2 and 3% respectively during the treatment and recovery periods. By contrast, the low and high dose groups lost weight during treatment, with weight loss being greatest at the low dose (see following table). During recovery, there was little bodyweight change at the low dose, whereas the high dose group showed a bodyweight gain of ca 8%.
Table 4.13: Bodyweight (g) of rabbits over the experimental period
GLY 1/100th LD50
GLY 1/10th LD50
Bw change over treatment^
Bw change over recovery^
^Calculated by evaluator *p<0.05 vs control
Treated rabbits displayed a partially reversible, non-dose related ca 25% reduction in semen volume during the treatment period, accompanied by a reversible 3-fold increase in the percentage of dead sperm and partially reversible, dose-related depression in initial semen fructose concentration and prolongation in methylene blue reduction time (see next table). According to the study authors, fructose formation by the accessory glands is dependent on testosterone production by the testis; hence, decreased semen fructose suggested a corresponding decline in testosterone secretion. Yousef et al considered that prolonged MBRT could reflect deficits in nutrition status, viability, activity and oxygen consumption by sperm from treated rabbits.
Interpretation of sperm concentration data is confounded by a progressive doubling in the control group between the pre-treatment and recovery periods. By contrast, sperm concentration in the high dose group remained constant during treatment, but declined by ca 8% at the low dose. In both treated groups, sperm concentration then rose by ca 1.8-fold during recovery. The percentage of abnormal sperm became significantly (p<0.05) elevated in the treated groups during the dosing and recovery periods, but again, interpretation is confounded by a two-fold increase in abnormal sperm occurring in controls (mainly) during recovery. The most common types of abnormalities were claimed to be coiled or double tail and tapering or small head. Semen osmolality in treated rabbits changed little during the dosing period, but became statistically significantly lower than in controls. This was caused by increased osmolality in the control group, rather than any effect of treatment.
Table 4.14: Semen characteristics of rabbits. Values are overall means over 6 weeks before, during and after treatment. [n = 4 before and during treatment and n = 3 during recovery]
GLY 1/100th LD50
GLY 1/10th LD50
Semen volume (mL)
Sperm conc. (X 106/mL)
Abnormal sperm (%)
Dead sperm (%)
Methylene Blue Reduction Time (min)
Initial fructose conc. (mg/100 mL)
Semen osmolality (units unstated)
P = Pre-treatment T = Treatment period R = Recovery *p<0.05 vs control
The study has significant shortcomings in its design and reporting of the experimental methods and results. The dosing interval and administered doses of glyphosate are unknown, and the authors did not explain how the reference LD50 value was derived. Although glyphosate treatment does appear to have caused decreases in ejaculate volume, sperm viability and sperm activity (the latter possibly resulting from depression in semen fructose concentration), the causal mechanism is unidentifiable. It is uncertain whether the results were obtained in the presence of systemic toxicity, as bodyweight loss during treatment was three-fold more severe at the low dose than the high dose. Any effects on semen osmolality, sperm concentration and sperm morphology are uninterpretable due to major, unexplained variation over time within the control group. The small size of the experimental and control groups may have contributed to the experimental outcome.
Furthermore, Williams et al (2000) have observed that:
The rabbits used in this study were small for their age, bringing into question their health status and reproductive maturity;
The proper method of semen collection was not used. Multiple ejaculates were not pooled to decrease the inter- and intra animal variability in sperm number and concentration;