What was done
The authors grew five herbaceous plant species in greenhouses receiving atmospheric CO2 concentrations of 350 and 700 ppm.  Following natural senescence, leaf litter was collected and incubated in ambient soils for 12, 30, and 45 days at 18 °C before analyzing its quality characteristics and presenting it to the woodlouse Armadillidium vulgare to study its feeding patterns.

What was learned
In general, the percentage weight loss of leaf litter during incubation (which is a measure of decomposition) was significantly less for CO2-enriched litter produced from Medicago, Trifolium, and Lolium species; while atmospheric CO2 enrichment had no effect on weight loss of litter produced from Tyrimnus and Galactites species.  Litter nitrogen contents were also differentially affected by growth CO2 concentrations.  Following incubation, for example, leaf litter collected from CO2-enriched Lolium and Tyrimnus species had lower nitrogen contents than those observed in litter derived from ambiently-grown control plants.  However, no significant CO2-induced effects on leaf litter nitrogen contents were found in Medicago and Trifolium species; while litter nitrogen contents actually increased in CO2-enriched Galactites species.

With respect to litter consumption, the woodlouse exhibited no significant differences in the consumption of Medicago, Tyrimnus and Lolium species, regardless of its growth CO2 concentration.  In contrast, the woodlouse consumed significantly greater amounts of leaf litter generated form CO2-enriched Galactites and Trifolium species.

What it means
In piecing together the many and varied differential responses discussed above, two broad conclusions can be drawn.  First, leaf litter nitrogen contents are not necessarily reduced due to atmospheric CO2 enrichment.  In the present study, for example, leaf litter nitrogen contents either dropped, rose or remained the same in response to atmospheric CO2 enrichment, dependent upon species.  Second, leaf nitrogen contents are not necessarily related to the feeding patterns of this particular woodlouse.

With respect to current thoughts on compensatory feeding, it is often suggested that herbivores will consume greater amounts of plant material generated in elevated CO2 environments due to the presumed lower nitrogen contents of such foliage.  However, in the present study, the two species that exhibited significantly lower leaf litter nitrogen contents -- Lolium and Tyrimnus -- were not consumed at significantly greater rates than those of their respective controls.  In fact, in one case where significant increased feeding did occur, it was for CO2-enriched leaf litter that actually exhibited increased nitrogen contents with atmospheric CO2 enrichment.

Clearly, it is way too early to draw any broad or generalized conclusions about the effects of atmospheric CO2 enrichment on the feeding properties of herbivores.