V I I . DESERT FISHES
459
small fish are present to October, but are in greatly reduced numbers after
July. Some adults, however, were ripe in other periods (including winter),
and presumably could have reproduced. James indicated that mortalities
and reduced condition factors resulting from diminution of food supplies
in winter, and possible prédation by planarian flatworms, among other
things, collectively resulted in a cycle that resembled the reproductive patterns that might occur under direct control of seasonal change in daylength. The fish live in Devil's Hole, a shaft with water > 1 5 meters below
the ground surface. Sunlight contacts the water directly in summer, but
not in autumn, winter, or spring, so that algal production, and secondary
production including that of fish, are intimately linked to incoming energy.
Light effects therefore may be only indirectly responsible for the spawning
cycle in this fish, operating principally through nutritional or other biotic
pathways.
Cyprinodon n. nevadensis in Saratoga Spring also is strongly cyclical.
Ovarian sizes increase in January-February to a peak in June or July, then
slowly decline to a low in November. Variation from fish to fish was high,
however, and some females appeared capable of reproducing through midsummer and into autumn at least, but little recruitment occurred after
June. The same pattern was unexpectedly evident in Saratoga Marsh, a
fluctuating habitat connected directly to the spring.
Cessation of recruitment similar to that observed in C. n. nevadensis
has been repeatedly duplicated in small experimental ponds with a number
of pupfish species from both stenothermal and eurythermal habitats. We
are tempted to attribute this to food limitations and/or population density in smaller habitats, and this certainly appears the case in Saratoga
Spring (although perhaps not so in the marsh). Hubbs and Strawn (1957)
similarly negated the importance of day length in breeding of greenthroat
darters (Etheostoma lepidum) from central Texas, stressing that nutritional condition of the fish and suitable temperatures were far more
important.
Termination of reproductive success in midsummer does not appear to
result from high water temperatures, since pupfish continue to exhibit reproductive behavior and to spawn apparently at temperatures approaching
their upper lethal limits. However, a myriad of other factors may come
into play. The barraging effect of inter- and intraspecific encounters involving territorial males may suppress successful spawning, and population
sizes most often are largest at about the time reproductive success declines
in pupfish populations. Cannibalism, either of young or ova, may be a
factor in some species, as may be the simple presence of a large population
which creates a stress syndrome. Additional research on these problems
is obviously called for. It is our impression that density-dependent fac-
459
small fish are present to October, but are in greatly reduced numbers after
July. Some adults, however, were ripe in other periods (including winter),
and presumably could have reproduced. James indicated that mortalities
and reduced condition factors resulting from diminution of food supplies
in winter, and possible prédation by planarian flatworms, among other
things, collectively resulted in a cycle that resembled the reproductive patterns that might occur under direct control of seasonal change in daylength. The fish live in Devil's Hole, a shaft with water > 1 5 meters below
the ground surface. Sunlight contacts the water directly in summer, but
not in autumn, winter, or spring, so that algal production, and secondary
production including that of fish, are intimately linked to incoming energy.
Light effects therefore may be only indirectly responsible for the spawning
cycle in this fish, operating principally through nutritional or other biotic
pathways.
Cyprinodon n. nevadensis in Saratoga Spring also is strongly cyclical.
Ovarian sizes increase in January-February to a peak in June or July, then
slowly decline to a low in November. Variation from fish to fish was high,
however, and some females appeared capable of reproducing through midsummer and into autumn at least, but little recruitment occurred after
June. The same pattern was unexpectedly evident in Saratoga Marsh, a
fluctuating habitat connected directly to the spring.
Cessation of recruitment similar to that observed in C. n. nevadensis
has been repeatedly duplicated in small experimental ponds with a number
of pupfish species from both stenothermal and eurythermal habitats. We
are tempted to attribute this to food limitations and/or population density in smaller habitats, and this certainly appears the case in Saratoga
Spring (although perhaps not so in the marsh). Hubbs and Strawn (1957)
similarly negated the importance of day length in breeding of greenthroat
darters (Etheostoma lepidum) from central Texas, stressing that nutritional condition of the fish and suitable temperatures were far more
important.
Termination of reproductive success in midsummer does not appear to
result from high water temperatures, since pupfish continue to exhibit reproductive behavior and to spawn apparently at temperatures approaching
their upper lethal limits. However, a myriad of other factors may come
into play. The barraging effect of inter- and intraspecific encounters involving territorial males may suppress successful spawning, and population
sizes most often are largest at about the time reproductive success declines
in pupfish populations. Cannibalism, either of young or ova, may be a
factor in some species, as may be the simple presence of a large population
which creates a stress syndrome. Additional research on these problems
is obviously called for. It is our impression that density-dependent fac-