Plant carbohydrate content limits performance and lipid accumulation of an outbreaking herbivore

Abstract

Locusts are major intermittent threats to food security and the ecological factors determining where and when these occur remain poorly understood. For many herbivores, obtaining adequate protein from plants is a key challenge. We tested how the dietary protein : non-structural carbohydrate ratio (p : c) affects the developmental and physiological performance of 4th-5th instar nymphs of the South American locust, Schistocerca cancellata, which has recently resurged in Argentina, Bolivia and Paraguay. Field marching locusts preferred to feed on high carbohydrate foods. Field-collected juveniles transferred to the laboratory selected artificial diets or local plants with low p : c. On single artificial diets, survival rate increased as foods became more carbohydrate-biased. On single local plants, growth only occurred on the plant with the lowest p : c. Most local plants had p : c ratios substantially higher than optimal, demonstrating that field marching locusts must search for adequate carbohydrate or their survival and growth will be carbohydrate-limited. Total body lipids increased as dietary p : c decreased on both artificial and plant diets, and the low lipid contents of field-collected nymphs suggest that obtaining adequate carbohydrate may pose a strong limitation on migration for S. cancellata. Anthropogenic influences such as conversions of forests to pastures, may increase carbohydrate availability and promote outbreaks and migration of some locusts.

Keywords: herbivores; locust outbreaks; macronutrients; migration; nutritional ecology; range expansion.

Figure 1.

(a) Marching Schistocerca cancellata preferred to feed on diets high in carbohydrate. The x-axis indicates the per cent protein to non-structural carbohydrate ratio (p : c) of the dry artificial diets. The y-axis indicates the number of locusts observed feeding in the bowls over 30–60 min during three different trials. In all cases, χ² analysis indicated that locusts fed non-randomly. (b) Protein and carbohydrate contents of the most abundant plants where marching locusts were found, compared to the preferred intake target (from artificial diet intake target experiment) of S. cancellata (0.5 p : c, black dashed line) and a p : c of 1 (grey dashed line). Open circles represent local plants that were fed to locusts in plant choice and performance assays.

Figure 2.

Performance of S. cancellata nymphs consuming single artificial diets or a single local plant. (a) Survival declined as dietary p : c increased on the artificial diets (binary logistic regression model: ${\chi }_2^2=7.033,$ p = 0.03; p : c ratio slope = 0.316, p = 0.007). Sex did not affect survival (sex slope = −0.108, p = 0.816). (b) Macronutrient ratio in artificial diets but not sex significantly affected body masses (two-way ANCOVA: diet effect: F_9,117 = 6.323, p < 0.001; sex effect: F_1,117 = 1.511, p = 0.221; diet × sex effect: F_4,117 = 1.346, p = 0.257; initial wet mass used as a covariate). (c) Body masses differed significantly between different plant fed locusts (ANOVA, F_6,62 = 12.357, p < 0.001). (d) Mass gain was only positive on Paspalum sp. (ANOVA, F_6,62 = 33.0, p < 0.001). In this and subsequent figures, letters indicate significant difference (Bonferroni post hoc tests, p < 0.05), and the number inside the black bars indicate the number of individuals in each treatment group.

Figure 3.

Schistocerca cancellata nymphs preferred local plants that were high in carbohydrate. During the buffet experiment, S. cancellata consumed more of plants with higher carbohydrate content (a), or plants with a higher carbon to nitrogen ratio (b). The small letters near each data point represent a plant which was used in experiments: a, Paspalum sp.; b, Prosopis sp.; c, Digitaria sp.; d, Celtis sp.; e, Celastraceae; f, Mikania sp.; and g, Grabowslia sp.

Figure 4.

Effects of diet on lipid concentration (mg of lipid weight per mg dry mass) and estimated flight capability for field-collected S. cancellata. (a) Locusts with a choice of diets (intake target experiments) or feeding on a single artificial diet with a p : c of 1 : 2 or lower, had higher lipid content than locusts confined to feeding on diets with 1 : 1 or higher p : c (ANCOVA using dry mass as a covariate, F_5,153 = 36.886, p < 0.001,). (b) Lipid content of locusts reared on local plants increased with relative C : N content of the plants.