In July I was lucky enough to attend a lecture in San Francisco by Uwe Romer a
German scientist who specializes in dwarf apistos. His experimentations pretty
well proved that temperature plays an incredible part in sex determinations in
dwarf apistos (and some west african dwarfs). I came away from that lecture
remembering 3 numbers 23, 26 29. 23 C (~74F) produced mainly females, 26 C
(~79F) equal mixtures, 29 C (~84F) mainly males. The cichlid folks had for
years assumed that the most important factor was pH. He also showed that this
temperature was important from time of birth up to some 33 days. So during the
first 33 days proper temperature was critical if you wanted to obtain a equal #
of males and females. After that timepoint it didn't seem to matter. A
question was asked whether this also occurred for other African cichlids.
Although not as carefully proven as his apisto data he stated that it didn't
seem to be a factor for many of the rift lake cichlids.
Recently Wayne Leibel has rephrased Uwe Romer's studies in the latest TFH (Jan
1997) pg 124-127) so you will be able to get an idea for the conditions used in
his experiments. He had used so many spawnings and really seemed to try and
control all of the factors, I was very convinced, by his data, that it was
temperature and not pH that the cichlid world had previously thought. The
original paper is (U. Romer & W. Beisenherz 1996 Journal of Fish Biology 48:714-
725). Unfortunately for me not many fish specialists at the local universities
so most of these journals are lacking. Anyone that would like to send me a
How does that relate to rainbowfish? Could they follow the pattern of apistos
in producing males/female ratio or is it unimportant as possibly suggested with
the observations with the rift lake cichlids? How come Roy couldn't seem to
buy a male M. pygmae to save his soul even though he produced hundreds of
females? I had almost the opposite effect and produced some 3 males for every
one female. At the time we felt that maybe it had more to do with the pH (or
maybe hardness). My pH tends to fluctuate much more because of low amount of
buffering capacity (KH). The water he was using at the time tended to be much
much harder with a lot higher pH. Could it have actually been a temperature
factor instead? Mind you rainbows don't necessarily have to fall into a 23,
26, 29 temperature gradient (if it holds true for them).
The instruments most of us use to measure temperature are pretty inaccurate. I
took a bunch of my <cheap> tank thermometers to work and calibrated them with
an accurate lab thermometer. Some of them were off by almost 5 degrees F.
This could be enough of a difference that could cause some skewing of the sex
ratios (if it holds for rainbows). It could have been possible that we had
enough of a difference between our tank raising temperatures that we produced
very different sex ratios. I am still leaning toward the pH (or hardness)
differences though because of observations I've heard from others in the US.
Softer more acid water tends to produce more males, harder water tends to
produce more females. So far most haven't produced as an extreme though as Roy
and I had observed with the M. pygmae. Still this would be a great experiment
for someone to try out. I would also suggest that they try several very
different species too and see if it holds the same. One apisto in Romer's
study tended to flip-flop the sex ratios with the temperature changes.
Chairman, Rainbowfish Study Group