Even the evidence for humans is a bit patchy - a third of studies come to the opposite conclusion, and plenty of those that agree have massive methodological flaws. So I doubt we can say that with any real confidence for other mammals.
Just want to highlight that you put a blog named "Astral Codex Ten" claiming patchiness and methodological flaws because "data are very noisy" against the peer-reviewed journals like Oxford Scientific which for example states:
"This comprehensive meta-regression analysis reports a significant decline in sperm counts (as measured by SC and TSC) between 1973 and 2011, driven by a 50–60% decline among men unselected by fertility from North America, Europe, Australia and New Zealand. "
If the "data are very noisy" argument were a sound argument it would have dismissed most of the science that we have today.
I'm not putting anything against anything. That paper [0] completely agrees with the blog - they threw out all but 38 of the 2936 studies they considered, for crying out loud - and neither makes any call on sperm count in animals other than humans.
"This comprehensive meta-regression analysis reports a significant decline in sperm counts (as measured by SC and TSC) between 1973 and 2011, driven by a 50–60% decline among men unselected by fertility from North America, Europe, Australia and New Zealand. "
If the "data are very noisy" argument were a sound argument it would have dismissed most of the science that we have today.