More Pitfalls with sperm viability staining and a viability-based stress test to characterize sperm quality

Abstract

Sperm viability (SV), the proportion of live sperm in a sample, is a widely applied measure of sperm quality but few studies test its robustness. At least three reasons make SV problematic as a surrogate for sperm quality. First, reviewing the ecological literature revealed that previously identified methodological pitfalls have not been overcome, including low cross-study standardization of protocols, inadequate statistical treatment, and unaccounted for within-sample heterogeneity. Second, SV is affected by biological variation such as between species, reproductive organs, or sperm age cohorts. Third, the proportion of live sperm extracted from males appears more related to male than to sperm quality in the sense of the future performance of sperm. We propose an alternative method to assess sperm quality by characterizing the temporal decrease of SV in a stressor medium and illustrate in two species, the common bedbug (Cimex lectularius) and the fruit fly (Drosophila melanogaster) how some common methodological pitfalls may be circumvented. Our data empirically support the well-known but little-considered facts that (i) non-blind measurements may alter SV and (ii) that SV frequently have non-significant repeatability within one sample. (iii) Cross-sectional sampling of ejaculates showed that this heterogeneity even masked a biological pattern-the sperm stratification within males. We show (iv) that this shortcoming can be overcome by following the temporal decline of SV of a sperm subsample in a stress test. Finally, (v) comparing the staining pattern of sperm between Cimex and Drosophila, we found that in the latter, the visibility of sperm is substantially delayed (30 min) when sperm density is high. We show that this delay in stained sperm visibility was, however, not biased toward dead or live sperm. To measure sperm quality, we advocate analyzing the temporal decline in SV in a stressor medium over current protocols that use SV per se and blinding samples for SV measurements. As cell viability is widely used in biological and medical laboratory studies, our protocol may be useful to characterize cell quality beyond ecology and evolution.