Field bands of marching locust juveniles show carbohydrate, not protein, limitation

Abstract

Locusts are grasshoppers that migrate en masse and devastate food security, yet little is known about the nutritional needs of marching bands in nature. While it has been hypothesized that protein limitation promotes locust marching behavior, migration is fueled by dietary carbohydrates. We studied South American Locust (Schistocerca cancellata) bands at eight sites across Argentina, Bolivia, and Paraguay. Bands ate most frequently from dishes containing carbohydrate artificial diets and minimally from balanced, protein, or control (vitamins and salts) dishes-indicating carbohydrate hunger. This hunger for carbohydrates is likely explained by the observation that local vegetation was generally protein-biased relative to locusts' preferred protein to carbohydrate ratio. This study highlights the importance of studying the nutritional ecology of animals in their environment and suggests that carbohydrate limitation may be a common pattern for migrating insect herbivores.

Keywords: Collective movement; Grasshopper; Locust phase change; Migration; Movement ecology; Nutrition; Orthoptera; Swarming.

Fig. 1

Panel A: Feeding behavior by a locust balancing carbohydrate and protein to reach its intake target (IT), RPO and L. Steger locust photo credit. Panel B: Hypothetical nutritional landscapes and predicted nutrient limitation for locusts. Panel C: Predicted compensatory feeding behavior by locusts in the nutrient limiting environments shown in panel B. Panel D: Marching South American Locust (Schistocerca cancellata) nymphs during a field trial in Paraguay, April 2019, ST photo credit. All panels are based on the Geometric Framework for Nutrition (Simpson and Raubenheimer 2012). See text for a detailed description.

Fig. 2

Map of experiment sites.

Fig. 3

Panel A: South American Locust nymphs aggregating to roost in the late afternoon near Cabezas, Bolivia, April 2017; inset of roosting nymphs from Catamarca, Argentina February 2020, RPO and L. Steger photo credit. Panel B-C: Field trial location near Boyuibe (Panel B) and outside Rio Seco (Panel C), Bolivia, April 2017, AJC photo credit. Marching bands were often found in disturbed areas, particularly along roadsides.

Fig. 4

Nutritional status of marching nymphs, as indicated by the proportion of locusts stopping at each diet across 12 trials in Argentina, 17 trials in Bolivia, and 3 trials in Paraguay. Boxplots are medians with interquartile ranges. Letters indicate significant differences from Dunn's multiple-comparison tests, following Kruskal-Wallis tests. Paraguay data were adapted from Talal et al., 2020 and analyzed using multiple chi-squared tests with Bonferroni corrections. Pie charts represent the total number of observations summed across all trials for each country.

Fig. 5

Intake targets for Argentina and Bolivia populations. Points represent means and SEMs of mass specific protein and carbohydrate consumption of males (open symbols) and females (closed symbols), eating diet pair 1 (triangle, 7p:35c and 35p:7c) or diet pair 2 (circle, 7p:35c and 28p:14c), from Argentina (purple) or Bolivia (pink) populations.

Fig. 6

Self-selected protein:carbohydrate intake targets (ITs) of locust populations from Bolivia (pink), Paraguay (blue), and Argentina (purple). Symbols indicate nutrient contents of available host plants at the outbreak sites for graminoids (triangles), forbs (squares), and woody vegetation (circles). The dashed line is the 1p:1c line, for reference. Most available plants fell to the right of the ITs, suggesting that locusts would have a difficult time obtaining sufficient carbohydrate without overeating protein (Fig. 1).

Fig. 7

Plant p/c ratios (A), protein contents (B), and carbohydrate contents (C) collected from field sites in Bolivia (pink), Paraguay (blue), and Argentina (purple). Boxplots are interquartile ranges with the median shown as a bold line and mean as an X. The small dots are the raw data points. In panel A, values above the 1:1 line are protein-biased (more protein than carbohydrate). There were significant main effects of plant type and country on plant p:c ratios and protein. Graminoids had lower p:c and protein values than forbs or woody vegetation, as represented by the bars and lower-case letters in panel A and B. Bolivia plants had lower p:c and protein values than Paraguay and Argentina (Tukey posthoc analysis not shown). For more statistical details, see the section Plant collection and nutrient assays.

Fig. 8

Locusts were more hungry for carbohydrate than protein in all landscapes (all lines fell to the left of the 1:1 line), but the carbohydrate hunger increased in magnitude from Bolivia (A), to Paraguay (B), and to the highest degree in Argentina (C). Bands followed a pattern predicted by nutritional geometry in that the more protein biased the plant leaf nutritional landscape that they were marching through was on average, the greater carbohydrate hunger they expressed (D). For more details see the section Comparing plant nutrients with marching band carbohydrate hunger and the discussion.