Feeding was the most sensitive endpoint examined, the 10-day NOEC (feeding) being 1.7 ug/g algae. With aldrin, the three endpoints produced common NOECs of 7.5 ug/g algae. Effects were not seen at lower concentrations as all larvae remained mobile, with no inhibition of feeding observed below the NOEC. Permethrin was the least toxic of the three pesticides examined. No significant mortality or immobilization occurred at any of the treatments, but feeding was adversely affected at 24 ug/g algae, the NOEC (feeding) being 8.1 ug/g algae.
To understand these toxic effects further the body burdens of these substances in affected larvae were examined. The body burden data (Table 2) indicated that exposure of the larvae to lindane at 102 ug/g in their diet resulted in accumulation of 2.7 ug/g in the larvae, 20% mortality, and complete immobilization after a 10-day exposure period. Exposure of the larvae to aldrin at 95 ug/g in the diet resulted in their accumulation of diedrin, presumably by an epoxylation reaction similar to that observed with plants and soil , although in this study dieldrin was not detected in the spiked algae. At this treatment a total body burden of 5 ug/g aldrin and 22 ug/g dieldrin killed 60% of the larvae and immobilised a further 20% after a 10-day period. Permethrin was not detected in the larvae exposed to the highest concentration (204 ug/g algae), which caused only reduced feeding and 20% immobility when compared with the controls. This indicates either that larvae failed to absorb permethrin from the algae or that larvae metabolised permethrin rapidly to substances of low toxicity that were not detected using the analytical techniques described above or otherwise eliminated it.
Exhaust Emission Simulation
The results for exposure of bagmoth larvae to algae amended with vehicle exhaust gases showed that the lead content of the algae increased with increasing time of exposure to exhaust gas/air mixture. Larvae exposed to algae containing lead at 200 ug/g and above did not feed during the 96-h exposure period, by which time they had become immobilized.
Lead concentrations in the algae were highest at the downwind site closest to the M40 (site 4), and declined gradually with increasing distance downwind from the motorway. Chromium concentrations were elevated at both sites closest to the motorway (sites 3 and 4), but were higher downwind than upwind. Algae from the control site contained less lead and chromium than at any of the transect sites. Feeding was reduced at site 4, the closest downwind site (10 m) to the M40, where the concentration of metals was highest. Feeding increased at sites 5 and 6, further downwind of the M40 (300 and 700 m, respectively). Feeding was also depressed at site 1, an area surrounded by arable farmland located furthest upwind of the M40 (700 m), but increased upwind of the M40 as the sites approached the motorway and distance increased from the farmland (sites 2 and 3, located 300 and 10 m from the M40, respectively). It is significant that larvae supplied with algae from the "clean" control site, an isolated rural woodland far removed from arable farming and the influence of road traff|c, fed much more than those supplied with algae from the transect sites. This is evidence of the impacted nature of the environment along the transect. However, traces of lindane (0.2 ug/g algae) were found in samples of control algae analysed prior to spiking, indicating that the control site was not pristine.
The most important feature of this work is that epiphytic algae have a high affinity for atmospheric pollutants and that bagmoth larvae feed on this material. This makes them front-line or sentinel indicator organisms as far as the toxicological effects of airborne pollutants are concerned. The epiphytic algae Diplococcus sp., on which Luffia ferchaultella feeds, can be attached to filters and used to absorb pollutants from an air stream. This material can then be used to establish air / algae equilibrium kinetics and for laboratory feeding trials. Laboratory experimentation has demonstrated that bagmoth larvae are adversely affected by common atmospheric pollutants absorbed by this material. Feeding was found to be the more sensitive of three toxicological endpoints investigated. The metabolism of aldrin to dieldrin, detected by analysis of larval body burdens, is significant when one compares the toxicity of aldrin with that of lindane. As the average third instar larval wet body weight was 1.5 mg, we calculate that the lethal quantity of aldrin ingested over a 10-day period was 35 ng/larva. The corresponding value for lindane was 1ng. The lethal dose for permethrin could not be calculated because of its low toxicity to these organisms. A subsequent field investigation using a transect across an impacted environment identified that reduced feeding by bagmoth larvae was related to proximity to sources of atmospheric pollution. The presence of large quantities of heavy metals in samples collected near a motorway confirms the ability of epiphytic algae to accumulate heavy metals from the air. The high concentrations of chromium and lead proximally downwind of the M40 indicate the strong effect of prevailing wind direction in dispersal of pollution from motorway traffic at this locality. Lead is a commonly used additive in petrol, while chromium is an additive used in the cement that this section of the M40 was surfaced with at that time. The detection of measurable amounts of lindane in algal samples collected from a "clean" site demonstrates the ability of these unicellular plants to accumulate this and probably other organic substances from the atmosphere.
There are many features of the biology of Luffia ferchaultella that make it a useful organism in ecotoxicological studies. Of most importance for assessing the toxic effects of atmospheric pollutants is that the diet of its larvae is capable of sequestering such pollutants, resulting in measurable detrimental effects on their feeding. Results of laboratory experiments designed to simulate pollution from agricultural sprays and vehicle exhaust emissions identified feeding as the most sensitive endpoint of those examined. A field transect demonstrated that larvae supplied with algae collected from an environment impacted by atmospheric pollution exhibited reduced feeding compared with those fed algae from a "clean" control site.