Introduction
The large asymmetry between plant and consumer nitrogen (N) contents (1) has led to the N-limitation hypothesis, which stipulates that in most terrestrial systems low plant N content generally impairs herbivore performance (1–3). Indeed, many studies have shown that herbivore growth, survival, and fitness respond positively to increase N or protein availability (4–11). Nevertheless a handful of studies have challenged the universality of the N-limitation hypothesis either by showing that herbivores can cope with N-deficiency, or that increasing N inputs actually decreases herbivore performance (12–16). One proposed mechanism to explain these exceptions is an imbalance of N (or protein) relative to carbohydrates (15). However, limited studies have tested this carbohydrate limitation hypothesis in situ in field settings. We propose that species with high carbohydrate demands (e.g. self-propelled migratory species) are most likely to be negatively impacted by high plant N and we tested the carbohydrate limitation hypothesis in the field using a migratory locust in a West African agroecosystem.
The importance of macronutrient balance on foraging behavior and performance has been unequivocally demonstrated in laboratory settings with artificial diets and a theoretical approach termed the Geometric Framework (17–20). For herbivores, protein and carbohydrate have been shown to be key nutrients, as illustrated by the fact that animals have evolved the capacity to regulate their intake to a specific ratio (21–23). Despite these advances, the link between macronutrient ratios selected by herbivores in the laboratory and plant nutritional content in the field has seldom been explored. Plant C:N ratios are commonly reported in the literature (1, 24–28) but not plant protein:carbohydrate ratios (for exceptions see (29, 30)). This is problematic because while N can be used as a proxy for plant protein content, the majority of plant carbon is structural (i.e. cellulose or lignin) and undigestible for most herbivores (31, 32). Furthermore, despite the demonstrated importance of both nutrients on foraging and performance, the effect of carbohydrate limitation on herbivore reproduction has yet to be tested in a field setting.
To test the carbohydrate-limitation hypothesis, we used the grass-feeder locust Oedaleus senegalensis in an annual cropping system (millet, Pennisetum glaucum ) and manipulated millet protein:carbohydrate ratios using N fertilizer. We used a grass-feeder species because grasses typically rely on physical defenses like silica (33, 34) and not chemical defenses; so our treatment would affect nutrients and not secondary compounds. Furthermore, all locust species undergo long, and energetically costly, migrations fueled by lipid reserves (Hunter et al., 1981; Weis-Fogh, 1952) that come from carbohydrate consumption (Simpson et al., 2002). O. senegalensisis no exception and can fly up to 350 km in 24h (38–41).
To our knowledge this is the first study to measure the effect of plant macronutrient content on insect survival and reproduction in the field and a key step to linking the Geometric Framework methodology with plant nutritional landscapes in the field.