Associations with Plants

Food For Freedom

Ideas for Surviving Food Shortages

Get Instant Access

Associations between Lepidoptera and their larval host plants are remarkably conserved. Related butterfly species are almost always found on related host plants, suggesting that colonization of new hosts is an evolutionary hurdle (Ehrlich and Raven 1965). The number of plant species that a butterfly can consume is thought to be limited by the physiological machinery needed to detoxify plant chemicals (Berenbaum 1995; see Bernays and Graham 1988). Specialized associations with host species create an opportunity for climate change to impact larval feeding (Dennis and Shreeve 1991). Weather may change the butterfly-host plant interaction by altering the temporal overlap between larval development and plant availability (Harrington et al. 1999). To illustrate this phenomenon, an extended example is drawn from my own research on E. e. bayensis.

E. e. bayensis has a sensitive temporal interaction with its host plants that is controlled by weather and climate. This butterfly is univoltine (one generation per year) and inhabits patches of annual grassland on serpentine soils in the region of the San Francisco Bay.






Figure 2.1. Life cycle diagram of Euphydryas editha bayensis and one of its two larval host plants, Plantago erecta. Dotted lines refer to the among-year variance in the timing of each organism. The location and length of the bars is determined by climatic conditions in a single year. In years of drought or deluge, larval development is delayed and can push beyond the date of senescence for P erecta. A second host plant, Castilleja spp., matures up to 2 weeks later than P. erecta, and can serve as an important resource to extend the period of larval development. (Drawing by Lawrence Lavendel.)

Figure 2.1. Life cycle diagram of Euphydryas editha bayensis and one of its two larval host plants, Plantago erecta. Dotted lines refer to the among-year variance in the timing of each organism. The location and length of the bars is determined by climatic conditions in a single year. In years of drought or deluge, larval development is delayed and can push beyond the date of senescence for P erecta. A second host plant, Castilleja spp., matures up to 2 weeks later than P. erecta, and can serve as an important resource to extend the period of larval development. (Drawing by Lawrence Lavendel.)

Obligatory hibernation, or diapause, of larvae is required during the grassland's dry season (Fig. 2.1).

An important driver of total population size in E. e. bayensis is annual variation in larval mortality. Larvae experience food shortages and die in large numbers when host plants senesce before larvae are sufficiently developed to diapause (Ehrlich 1965, Singer 1972, Ehrlich et al. 1980, Dobkin et al. 1987, Cushman et al. 1994). Food shortages occur when plant growth and death are accelerated by high temperature or drought or when larval hatching and growth are delayed by cloudy days or low temperatures. In years with moderate temperatures and precipitation, in contrast, host plants can sustain sufficient numbers of larvae to diapause, and populations remain stable or expand.

The oligophalous nature of E. e. bayensis and the properties of the butterfly's host species complicate the larval race to diapause in extreme years. Pre-diapause larvae feed on two types of host plants: Plantago erecta and species in the genus Castilleja (Singer 1971, Hickman 1993), and the timing of senescence in these two hosts is distinct. In particular, Castilleja develops and dies approximately 1 to 2 weeks later than P. erecta, and larvae that forage in Castilleja patches have been noted to have a higher probability of survival than larvae outside these patches in extreme years (Singer 1971, 1972).

To investigate the relationship between larvae and their hosts, larvae were given the option to feed on both host plants over the length of larval development in field trials, and their diet choice was tracked through time (Hellmann 2000). In separate enclosures, larvae hatched either from eggs on P. erecta or from eggs on Castilleja. Approximately half of the larvae from P. erecta eggs moved to Castilleja shortly after hatching. Similarly, approximately half of all larvae that hatched on Castilleja moved to P. erecta early in the study period. However, the fraction of larvae feeding on P. erecta quickly decreased as time passed while the corresponding fraction of larvae foraging on Castilleja increased. As P. erecta quality declined, proportional use of Castilleja grew. This suggests that Castilleja sustains feeding longer than P. erecta, a quality that enables it to partially buffer larvae against climatic extremes. If these extremes occur with greater frequency under climate change, Castilleja could play an increasingly important role in larval diet.

Further evidence of the importance of Castilleja under extreme conditions can be seen in experiments of plant and larval performance under elevated temperature (Hellmann 2000). Persistently elevated temperature accelerates both the senescence of hosts and the growth of larvae. However, larval survivorship under these conditions is greater when Castilleja is present than when it is absent. If regional climate brings temperatures that are consistently higher than prechange conditions, this result suggests that the availability of a long-lasting host (i.e., Castilleja) is key to maintaining butterfly populations (Hellmann 2000).

Exactly how much Castilleja can help sustain checkerspot populations under change is limited by the plant's total abundance. Unfortunately, Castilleja is patchily distributed and highly variable among years. At low abundance (with small patches and large distances between patches), the fraction of larvae that are able to feed on this species is relatively small. P. erecta, in contrast, is widely dis persed throughout the grassland, and its abundance fluctuates little annually.

Further, it is not yet known how climate change may influence the distribution and abundance of Castilleja, but several scenarios are possible (Hellmann 2000). If Castilleja abundance is high under a future climate, the likelihood of butterfly population persistence under warming also may be high because many larvae could feed on this key resource. Conversely, if the changing climate significantly decreases Castilleja abundance, few larvae will be able to exploit its buffering potential, and butterfly populations could decline. To account for this latter possibility, management activities that promote Castilleja would be helpful in preserving Bay checkerspot populations.

This type of temporal interaction between E. e. bayensis larvae and their host plants is not unique, and many host-restricted species may be prone to climate impacts by changes in plant dynamics. For example, large larval mortality has also been observed when hosts and larvae fall out of phase in Pieris virginiensis. Cappuccino and Kareiva (1985) investigated interannual differences in the availability of hosts for Pieris virginiensis and found that in some years, host plants senesced before larvae had completed foraging. During years with early plant death, many larvae were left stranded with no alternative hosts. Sensitivity to weather-driven phenology also extends to butterfly species whose larvae feed on woody species because larval foraging is closely linked to leaf flush, leaf age, and the timing of leaf chemical defense (Feeny 1970, Coley 1980, Holliday 1977, Watt and McFarlane 1991, Hunter and Elkington 2000). Whether temporal shifts in woody species will lead to increases or decreases in lepidopteran numbers is dependent on the direction of the shift in tree phenology versus the timing of larval development (Buse and Good 1996, Visser and Holleman 2000, Buse et al. 1998, Dewar and Watt 1992).

Timing is not the only factor of host accessibility and quality. Changes in soil moisture, soil temperature, cloud cover, precipitation, and air temperature can each affect the nutrient content or palatability of host plants. A change in food quality can be harmful to herbivores if it forces them to increase foraging time, thereby exposing them to predators or increasing the potential for food limitation (Ayres 1993). Extreme weather events also can drive changes in host plant suitability and availability and can even lead to severe plant death and die-back (Inouye 2000). An example of how the impact of weather on plants can translate to high butterfly mortality was observed in Colorado in 1969 (Ehrlich et al. 1972). An unseasonably warm spring was followed by a brief, late, and severe snowstorm that killed or degraded a large majority of lupine inflorescences, the oviposition and larval foraging site of the butterfly Glaucopsyche lygdamus. A local population extinction of this butterfly species was observed at elevations where lupine was severely damaged and the eggs they housed were killed. In a second example, the host plants of a single subpopulation in a source-sink metapopulation of E. editha were killed in a severe summer frost (Thomas et al. 1996). As a result, larvae starved, and the butterfly subpopulation went extinct. A nearby subpopulation also declined from the loss of immigrants.

Increases in the atmospheric concentration of CO2 also can affect the relationship between butterflies and their larval host plants. Under elevated CO2, the nutrient concentrations of plants change, increasing the ratio of carbon to other essential nutrients such as nitrogen and water. The general pattern emerging from studies on the impacts of elevated CO2 on arthropod herbivores is that individuals grow more slowly, eat more plant biomass, and have an extended development time on plants grown under elevated CO2 (Ayres 1993, Watt et al. 1995, Coley 1998). Lengthened development time can result in greater exposure to predators and parasites, and decreased food quality increases the potential for food to be limiting, thereby decreasing total population sizes. Studies of the impacts of elevated CO2 on plant community composition also indicate that plant species respond differently to increased atmospheric carbon, resulting in possible changes in plant dominance (see Baz-zaz 1990, Vitousek 1994). If vegetation composition changes, altering the availability of host resources, butterfly populations would be affected in turn.

In a study with the butterfly Junonia coenia, researchers found decreased larval growth and survivorship on plants grown under elevated CO2 (Fajer et al. 1989). Larval growth was slowed from decreased water and nutrient concentrations in the host plant; the CO2 treatment, however, did not affect the concentration of defensive compounds. Studies of moths also have shown increased feed ing under elevated CO2 (Lincoln et al. 1986, 1993), and one study has seen a change in the digestibility of leaf material due to increased leaf toughness and a change in plants' ability to defend themselves (Dury et al. 1998).

In the E. e. bayensis system, larvae may need to feed longer to meet nutritional demands under elevated CO2, and these foraging needs could delay development to diapause beyond the date of plant senescence. This effect could increase larval mortality at the onset of the grassland dry season and exacerbate phenological shifts driven by climate alone (as already discussed). Elevated CO2 also might alter competitive interactions among host plants on the serpentine, either positively or negatively impacting the availability of larval food resources. Experiments on community composition in the grassland indicate that CO2 can affect the density of native forbs. The primary influence, however, is on late-season annuals; CO2 impacts on early-season forbs (including larval host plants) are restricted to soil types where native forbs co-occur with invasive grasses and are less abundant (Chiariello and Field 1996). The direct impacts of CO2 on plant quality and senescence date of larval plants are not known.

Impacts on plants can indirectly affect adults as well as larvae. Butterflies feed as adults, and most use some form of plant material for essential sugars and nutrients (e.g., nectar, pollen, sap, and rotting fruit; Boggs 1987). Just as weather and CO2 concentration can affect the quality, availability, and timing of larval host plants, they also can change the quality or availability of adult food resources. For those species that allocate a large fraction of adult resources to reproduction and have little larval stores for egg production, the effect of climate change on adult food could substantially impact total reproductive output (Ehrlich and Gilbert 1973, Murphy et al. 1983, Boggs 1986, 1997, Boggs and Ross 1993).

Was this article helpful?

0 0

Post a comment