Kiesecker, JM 2002. Synergism between trematode infection and pesticide exposure: A link to amphibian limb deformities in nature? Proceedings of the National Academy of Sciences 99: 9900–9904.

Keisecker presents data from a series of field and laboratory experiments indicating that frog deformities caused by parasites are more common in habitats contaminated by pesticides because contaminant-exposed frogs are less able to resist parasite infection. This study may neatly knit together what had been viewed as two mutually-exclusive and competing explanations for why frog deformities occur and why their incidence has increased so dramatically, contamination vs. parasites.

What did he do? In his field experiments , Keisecker put wood frog (Rana sylvatica) tadpoles in ponds that had populations of snails infected with the trematode parasites known to cause limb deformities in frogs. These snails are a host for the parasites during a different stage in their life-cycle.

Keisecker organized the tadpoles into two separate treatments in these ponds. One set were placed in containers with sides made of a mesh large enough to allow the larval stage of the parasite to pass into the container and infect the tadpoles, while the other set were in containers whose mesh was so small it kept the larval stages out.

The ponds themselves differed with respect to pesticide levels. Some were in close proximity to agricutural fields whereas others were distant. Keisecker engaged a chemical laboratory to measure the levels of a range of organochlorine and organophosphate pesticides. Three ponds were clear of measureable pesticides whereas the other three had a series of compounds.

This combination of treatments—high or low pesticides, accessible or inaccessible to parasites—allowed Keisecker to examine separately the effects of pesticides and parasites, and look at their interaction.

In his laboratory experiments, Keisecker tested the impact of three pesticides on parasite infection rates, while also measuring the levels of a type of white blood cell (called eosinophils) thought to be important for resisting parasite infections. The pesticides tested were atrazine, malathion and a synthetic pyrethroid, esfenvalerate. Keisecker compared the effect of the pesticides at three different levels: control (no pesticide), ; low (the EPA maximum permissible standard for drinking water), and high (ten times the lower level). The lower level was chosen to ensure that the levels used were biologically plausible, i.e., that frogs in the real world would encounter them.

What did he find? The pond experiments revealed that the parasites were necessary for limb deformities. None of the tadpoles developed deformities in the cages made with mesh too small for their passage. Deformities occurred, however, in the cages with larger mesh sizes, in all ponds.

Deformities were strikingly less common, however, in the ponds that were uncontaminated by pesticides.

The results of Keisecker's pond experiments: The 0's indicate that cages without parasite access never had deformities. Cages with access had proportions of limb-deformed frogs >0. Unpolluted ponds with parasites had fewer deformities than polluted ponds with parasites. adapted from Keisecker 2002

Keisecker also noted size differences among treatment groups in the ponds. Frogs in cages exposed to parasites were smaller than those unexposed. This size difference was more pronounced in cages in contaminated ponds than in clean ponds.

The lab experiments revealed that even very low levels all three pesticides (atrazine, malathion and esfenvalerate) altered the frog's immune systems and increased the rate of parasite infections. This is reflected in the figure below, summarizing the results for one of the three pesticides, atrazine. [note that Keisecker used two different types of trematode in his experiments, Telorchis and Ribeiroia; the results were similar for both.

The graph on the left shows that both levels of atrazine used in the experiment (3 µg/liter and 30 µg/liter) suppressed the number of eosinophils compared to the control (dechlorinated tap water) and a "solvent control" (needed to test for the effect of the solvent used to dissolve the pesticide into solution).

The graph on the right shows that parasite infection rates (the proportion of trematode larvae that encysted successfully in the exposed frogs) were greater for both levels of atrazine exposure.

adapted from Keisecker 2002

It should also be noted that "high atrazine" was high only relative to the low level used. Frogs near agricultural fields will regularly encounter 30 µg/liter atrazine.

What does it mean? First, this paper adds to the growing weight of evidence that pesticides, particularly atrazine, affect frog development at levels far beneath those previously thought to be relevant.

Second, Keisecker's work provides important new insights into why frog limb deformities have increased so dramatically. As he notes in his paper, frog limb deformities have been observed in nature for several hundred years. Only in the last decade, however, have they been detected at plague-like proportions. This paper provides strong evidence that pesticide impacts on frog immune systems makes them more vulnerable to pathogens they normally would have been able to resist.

Many questions remain, nonetheless. A number of authors have noted frog deformities in ponds where they have not detected parasites (see Bill Souder's book, "A Plague of Frogs," for an excellent, broad review). Are some contaminants capable of inducing deformities without parasites present? Some patterns of deformities do not appear associated with parasites at all. What is causing them? Some evidence suggests that changes in water quality because of agriculture made habitat conditions more favorable for the snails, and that this may be contributing to the increase in deformities. Some evidence indicates that snails can be exquisitely sensitive to endocrine-disrupting contaminants. Are there circumstances where an excess of contaminant would actually suppress snail populations, deprive the parasites of hosts for another part of their life-cycle, and actually lead to fewer deformities?

More generally, this paper highlights two important issues. More attention needs to focus on the interactions of developmental contaminants with other factors affecting health, survival and function. In this system, adverse effects of natural pathogens are exacerbated via erosion of immune system function. An analogous example, also with frogs, was reported by Relyea and Mills in 2001; they discovered that predator-induced stress increased the toxicity of a contaminant, carbaryl, to tadpoles.

Animals—and people—live in the real world, with multiple factors impinging upon health, survival, reproduction, etc., all at once. Laboratory experiments that work with only one factor at a time can provide fabulous, reductionist insight into the actions of single mechanisms by themselves, but totally miss what may be happening in real life. Our regulatory standards are based on these reductionist approaches and hence are likely to be far too lax.

The second general point is the impact of contaminants on immune system development and function. As noted elsewhere, the way we compile statistics about the causes of death and disease attributes cause to the disease agent... a virus, bacteria, or parasite...with no recognition that death or disease might not have resulted from infection, were the victim's immune system not impaired. A highly relevant example was published in 1997 on the interactions of PCBs and Epstein Barr virus in increasing risk for non-Hodgkin's lymphoma. If this mechanism of contaminant impact is as widespread as early assessments suggest, we are grossly underestimating the role of contamination in undermining human health.