Agriculture affects amphibians (part 2):
Pesticides, fungi, algae, higher plants, fauna, management recommendations

by Robert L. Bugg, SAREP, and Peter C. Trenham, postdoctoral research fellow, Section of Evolution and Ecology, UC Davis

California red-legged frog (threatened). (photo by Matthew Fujita, UC Davis/UC Berkeley)

Part 1 of this article (viewable here) addressed the consequences of climatic change, landscape-scale dynamics, hydrology, and mineral enrichment of water on amphibians. In Part 2, we discuss pesticides, associated organisms (e.g. fungi, algae, higher plants, parasites, predators, and competitors), and on-farm modifications that may favor native amphibians. The list of references is also provided.

Pesticides
At sufficient doses many pesticides negatively affect amphibians. The issue is complex. Davidson et al. (2001, 2002) found that several California frogs have declined disproportionately from sites that are downwind from areas with agricultural activity. This suggests that windborne agricultural pesticides might be contributing to declines; however, it is unclear which pesticide(s). These chemicals can travel surprisingly long distances. Even frogs collected from high in the Sierra Nevada contain detectable concentrations of organochlorine pesticides that appear to be compromising their immune systems (Sparling et al. 2001).

An additional concern is that standard assays often underestimate the toxicity of pesticides. These tests typically expose tadpoles over short intervals under unnatural conditions. Problems with short-term studies, focusing only on tadpole mortality, were emphasized by Hayes et al. (2002), who demonstrated that exposure to extremely low concentrations of atrazine was not lethal, but resulted in the dramatic feminization of male frogs. Further, Relyea and Mills (2001) found that the insecticide carbaryl was much more toxic to tadpoles when the latter were stressed by the presence of a predator. This calls into question the wisdom of unnatural test conditions. Clearly, where herbicides and insecticides reduce food sources these chemicals can have indirect effects on amphibians. Also, exposure to certain chemicals can make larval amphibians more vulnerable to parasites and predators (see “Native Fauna” below).

Associated Organisms
Fungi. A chytrid fungus, now recognized as Batrachochytrium dendrobatidi, was implicated by Berger et al. (1998) as the causal agent for die-offs of amphibians in Australia and Central America. This fungus attacks keratinized tissues, including the mouths of tadpoles and various structures of the adults like the pelvic patch, which is the site of cutaneous respiration and osmoregulation for many frogs and toads. Fungal infections of the mouth that do not appear to reduce larval survival often spread to other organs and kill the metamorphosed young. Rollins-Smith et al. (2002a, 2002b) discovered that several peptides found in the skin of amphibians, including ranid frogs, inhibit infection and growth by the fungus and other pathogens. Rollins-Smith et al. (2002b) speculated that if environmental factors inhibit formation and exudation of the peptides, this resistance mechanism could be compromised. The fungus has been isolated from frogs in the Sierra Nevada of California, although the role it may have played in declines here remains unclear (Fellers et al. 2001).

Algae. Diatoms (Chrysophyta) are important foods for larvae of several frog species, including Pacific tree frog (Hyla regilla) (Kupferberg et al. 1994) and foothill yellow-legged frog (Rana boylii) (Kupferberg 1997). In streams, diatoms may occur as films on rocks (a type of periphyton) or as ephiphytes attached to some species of filamentous green algae (e.g., Cladophora spp., Chlorophyta). In either case, diatoms are a fat-rich food that enables more rapid growth and metamorphosis than occurs on diets of filamentous green algae alone. This suggests that algacidal chemicals may have deleterious side effects on frog reproduction.

Higher Plants. Aquatic, emergent, and terrestrial higher plants influence shading, food availability to larvae and adult amphibians, run-off, siltation, insulation from thermal extremes, and refuge from predators. Different amphibians have different optima and tolerances when it comes to vegetation types and degree of cover. Management that influences vegetation in and around breeding habitats clearly may affect amphibians. In Madagascar, Vallen (2002) found that, comparing the amphibian fauna of intact rainforest with 1) secondary forest, 2) eucalyptus plantation, and 3) rice fields, that these habitats had richnesses reduced by 46, 54, and 88%, respectively. As noted by Mitchell and Power (2003), invasive introduced plants, through release from fungal and viral pathogens, may be at a competitive advantage with respect to native plants. This might be expected to have indirect consequences for associated native fauna, such as amphibians, but no formal correlative or causal linkage has been made.

There is much interest in restoration ecology, yet there are few data documenting the conditions that prevailed prior to European colonization of California. Thus, valid models or targets for restoration are controversial. A case in point occurs with the bunchgrass dominance paradigm. For many years, the prevalent belief was that purple needlegrass (Nassella pulchra) and other native bunchgrasses dominated much of the floor of California’s Great Central Valley prior to a great drought in 1860. However, Holstein (2001) asserted that stands of rhizomatous graminoid plants (e.g., grasses, rushes, and sedges) and annual forbs were—and are—more prevalent, citing several examples where observations occurred prior to 1860 and minimal disturbance has since occurred. This and related issues may have consequences for survival not only of flora, but also of associated fauna, such as amphibians. Different California native plant complexes have yet to be explored as determinants of amphibian assemblages.

Britson and Kissell (1996) reported that when tadpoles of upland chorus frog (Pseudacris triseriata feriarum) were fed oak or pine pollens exclusively during any phase of development, size and incidence of metamorphosis were reduced. Thus, although tadpoles may feed on suspended pollens, these may not by themselves sustain metamorphosis.

Hazell et al. (2001) studied frog diversity in farm ponds of New South Wales, southeastern Australia. They found the highest diversity in ponds with abundant emergent vegetation, higher percentages of ground cover in the riparian zone, and higher percentages of cover by native vegetation within 1 km. Emergent and riparian vegetation are thought to be important as shelter for adults and recent metamorphs, reducing predation and desiccation.

Driscoll and Roberts (1997) in Western Australia found that controlled burning of native vegetation, to reduce fuel load and the likelihood of catastrophic wildfire, led to a short-term 29% decline in calling males of the frog Geocrina lutea.

Waldick et al. (1999) in eastern Canada found that black spruce (Picea mariana) plantations harbored much lower densities of redback salamander (Plethodon cinereus) than did natural forests that included deciduous angiosperms.

In Maine, DeMaynadier and Hunter (1999) found higher densities of adult and newly metamorphosed wood frog (Rana sylvatica) and spotted salamander (Ambystoma maculatum) in closed canopy forest rather than in a cleared power-line right-of-way. Further, they showed that newly metamorphosed wood frogs have strong preference for closed canopy habitat.

In agricultural landscapes of southern Quebec, Canada, Maissonneuve and Rioux (2001) found that amphibian and reptile (herpetofauna) species richness and diversity were greatest in shrubby, intermediate in woody, and lowest in herbaceous riparian strips. By contrast, for small mammal species richness and diversity, highest values were in this order: woody > herbaceous > shrubby. For herpetofauna, abundance increased with increasing complexity of vegetational structure, i.e., vertical stratification of vegetation.

Kruess and Tscharntke (2002), near Hamburg, Germany, found that increased grazing intensity of pastures led to no changes in plant diversity, but insect diversity declined. This pattern might have consequences for amphibians, such as northern leopard frog (Rana pipiens), which forages in meadows.

Woodford and Meyer (2003) found that lakefront housing and the associated removal of emergent and riparian vegetation was associated with greatly reduced densities of male green frog (Rana clamitans melanota) in Wisconsin glacial lakes. The authors pointed out that lakeside developmental and vegetation management standards could be altered and enforced to better accommodate the green frog.

Native Fauna. A parasitic fluke Ribeiroia ondatrae (Trematoda) causes limb deformation in several native California amphibians; these deformations are thought to have survival value to the flukes because they increase amphibian susceptibility to predation by birds and mammals, which serve as final hosts to the parasites (Johnson et al. 2002). In Penn-sylvania, Kiesecker (2002) found that hatchling tadpoles of wood frog exposed to the herbicide atrazine and the insecticides malathion and esfenvalerate, all at legally permissible levels for drinking water, led to compromised immune systems and increased infection and limb deformation by the flukes Ribeiroia sp. and Telorchis sp. In cage studies carried out in ponds, Kiesecker (2002) also found that limb deformations caused by flukes did not occur in ponds that lacked agricultural run-off containing these pesticides, although exposure to flukes did lead to reduced body mass of newly metamorphosed frogs.

Permanent water bodies enable overwintering by larvae of various dragonflies, including darners in the genera Aeshna and Anax (Odonata: Aeschnidae), which in the succeeding spring are important predators of native amphibian larvae (see Petranka and Hayes 1998). Perhaps seasonal draining of such ponds would lessen these problems, as has been shown in other parts of the United States (Adams 2000).

Because amphibians have permeable skin, they are very vulnerable to desiccation. To avoid desiccation and predators, amphibians seek shelter. For aestivation, or summer dormancy, some amphibians find sufficient refuge under surface debris or in dense vegetation, whereas other species move underground. Although some amphibians construct their own burrows, many depend on tunnels created by other organisms.

In much of California, amphibians must survive a long, hot summer, with almost no rainfall for six months; therefore, aestivation, and underground refuges to accommodate it, are critical. Western spadefoot (Spea hammondii) digs its own burrows in loose sandy soils or gravel beds, where it is inactive until re-emergence. California tiger salamander (Ambystoma californica) relies on other animals to create its burrows. In two studies, this salamander was found exclusively in the underground burrows of mammals like California ground squirrel (Spermophilus beecheyi) and Botta’s pocket gopher (Thomomys bottae) (Loredo et al. 1996, Trenham 2001). Other California species such as western toad (Bufo boreas) and Pacific tree frog have also been observed in mammal burrows. We believe that aestivation sites provided by fossorial (digging) mammals may be especially important on farmland, where logs and boulders are typically lacking.

Introduced Fauna. As indicated by Torchin et al. (2003), the success of introduced versus native animals may derive in part from their lower incidence of parasitism. The implied competitive advantage may enable invasion by an exotic species, some of which may compete with, or be natural enemies of, sensitive native species.

In California, bullfrog (Rana catesbeiana), which is native to the eastern United States, is the most common introduced amphibian. Replacement of extensive seasonal wetlands by permanent ponds water can enable survival of invasive exotic species such as bullfrog. Bullfrog interfere with red-legged frog (Rana aurora draytonii), foothill yellow-legged frog, and Pacific tree frog through larval competition for food and through predation by adult bullfrog, (Kupferberg 1997, Lawler et al. 1999, Chivers et al. 2001).

In Queensland, Australia, Crossland (2000) reported that eggs and hatchling tadpoles of the introduced toad Bufo marinus are toxic to the predatory larvae of the native frog Limnodynastes ornatus. In artificial ponds, this led to increased survival by the native frog Litoria rubella, which in the absence of B. marinus is preyed upon by Lim. ornatus. This work illustrates community re-structuring through direct and indirect effects, in that the presence of the introduced toad suppresses a native predator, thereby releasing another native species from predation.

Introduced sportfish and mosquito fish (Gambusia affinis) also interfere with native fish and amphibians (Goodsell and Kats 1999, Lawler et al. 1999), and are present in many permanent water bodies in the Central Valley, such as agricultural ponds. These fishes demonstrate a strong negative correlation with native amphibians, and have been proposed as a major threat to several species (Fisher and Shaffer, 1996). In the Sierra Nevada of California there is extensive evidence that the decline of mountain yellow-legged frog (Rana muscosa) is at least partly due to predation by introduced trout (Knapp and Matthews, 2000). In Australia, introduced trout species also prey on the tadpoles of native frogs; these tadpoles are unpalatable to native fish, but acceptable to the introduced ones (Gillespie 2001).

Management recommendations
For farmers and other land managers committed to enhancing native amphibians, we recommend that they:

1) Create new ponds and avoid filling low areas that flood during winter rains.
These are breeding habitats for many of our native amphibians. Wetland loss due to filling and draining is one of the main threats to amphibians worldwide. However, farmers and ranchers regularly create ponds that often provide productive amphibian breeding habitat.

2) Manage aquatic habitats to simulate natural Californian conditions, avoiding the substitution of permanent ponds or wetlands for ephemeral ones. As noted earlier, seasonal drainage of ponds disrupts the establishment of detrimental populations of fish, bullfrogs, and predatory insects. Seasonal wetlands with long hydroperiods may sustain the highest diversity of native amphibians, but even short-lived puddles and ditches can support tree frogs and toads to metamorphosis.

3) Avoid introducing non-native fish, including mosquito fish, to seasonal or permanent bodies of water. This may run counter to mosquito and othervector-control priorities, and if so requires consideration of alternative control tactics.

4) Retain non-cultivated, preferably native, vegetation near ponds, streams, and wetlands. Submerged, emergent, and terrestrial vegetation are important as shelter for larval, adult, and newly metamorphosed amphibians. Vegetation provides essential cover from predators and moist shelter sites.

5) Minimize the introduction of agrichemicals, including pesticides and fertilizers, into aquatic and terrestrial habitats. These chemicals can have important direct and indirect deleterious effects on native amphibians.

6) Retain potential shelter for amphibian aestivation. Uncultivated woody or grassland patches, rodent burrows, woody debris, and rock piles provide important refugia. Pocket gopher and ground squirrels should be tolerated, where possible; this is not always the case.

Landscape-scale management issues include proximity of multiple breeding sites to one another and width of uncultivated zones that link breeding sites to forested areas or to other habitats required in the amphibian life cycles. Narrow hedgerows are probably insufficient to provide linkages (see Joly et al. 2001, Le Coeur et al. 2002).

Farm ponds and ditches are still poorly understood ecological resources in the United States; research cited here highlights the possibility of enhancing these resources (also see Pokorny and Hauser 2002, Maezono and Miyashita 2003).

Additional studies on the above and related themes will enable us to augment and refine recommendations and thereby enhance conservation of native amphibians that are influenced by agricultural practices.