QUESTIONS AND ANSWERS

Disentangling ecological complexity: food web research on a tropical floodplain river.
Kirk Winemiller, Texas A&M

Q: How do flood pulses affect the distribution of species (perhaps through changes in structural complexity of habitat)?

A: We believe flood pulses in tropical lowland rivers have profound effects on not only fishes, but all aspects of the ecosystem. Fish density fluctuates as the extent of aquatic habitat changes during the annual hydrological cycle. During the peak of the flood pulse, when the aquatic realm extends over many hundreds of square kilometers in the savanna surrounding the Cinaruco River, the fishes are spread out at very low per-unit-area densities. It is during this time that most fishes spawn, and the larvae and young probably suffer relatively low predation mortality because encounter rates with predators are very low. Also, as your question implies, there is a lot of structural complexity in the flooded riparian forest and savanna. This probably benefits small fishes that can escape predators more effectively in these complex habitats. But as floodwater subside, fish populations (larger due to recruitment of young of the year) become increasingly dense within shrinking habitats. We have evidence that predator stomachs are fuller during this time of year. As the water recedes well inside the confines of the river and lagoon banks, fishes and aquatic macroinvertebrates may attain densities that lead to competition for access to limited habitats of high structural complexity. This is what results from our field experiment with artificial habitat patches suggested. 

Q: You support a top-down control of the food web, with Cichla pretty much on top and lower species like "bocachico" not as important, but what about the lower, primary producers role on the food web supporting a bottom-up controlled food web? Isn't "bocachico" an arbitrary "middle" species?

A: The bocachico is indeed a "middle species" in the food web, but we selected it for study because of some special characteristics. It is very abundant. There is evidence from prior research by Flecker's group at Cornell that a related prochilodontid has strong effects on organic sediments in Andean piedmont rivers.  Also, the bocachico is a prominently migratory fish, so we anticipated that its effects on the benthic ecology of the Cinaruco River would be seasonally variable. This is what our experiments are showing. We do not discount bottom-up dynamics in this system, and in fact, I argue that the bottom-up aspect dominates the benthic subsystem during the flood season.

Q: Structural complexity increased species composition over a time of 26 days, but what about between the wet and dry seasons?

A: Again, this is a very good question, and one that would be interesting to test experimentally. It would be more difficult to conduct the artificial habitat experiments during the flood season, but this could be done. I would predict that we would see more randomness in assemblage composition on artificial patches, and also that colonization rates would be vastly lower relative to rates observed during the dry season when fish densities are high.

Q: The effect of fishing large piscivores on the food web in the Cinaruco is spatially variable overall, then, is the effect of fishing significant?

A: I suppose the answer depends on what you mean by significant. Certainly for the habitats (lagoons) that are fished heavily by the netters, the cascading impacts are significant. The question I suppose is how many habitats in this system are being impacted in this way, and can the remainder of the landscape compensate for losses and changes in those habitats. Nobody knows the answer to that question, and I suppose it would require modeling, spatially explicit, to make such an estimate. It is my contention that large floodplain rivers in the tropics have a great deal of natural resiliency to fishing losses. A large aspect of this resiliency is the annual flood that makes fishing unprofitable (due to low catch rates in the expanded aquatic habitat) during the period when most fishes spawn.  Also, there is a great deal of dispersal during the flood period, so that areas that are depleted are restocked each year based on fish production derived from less-impacted areas.