Experimental design
Our field site was the village of Nganda, in the West Central Agricultural Region of Senegal where most of millet and groundnut production takes place. The two crops are typically rotated from one year to the next. Pearl millet (Pennisetum glaucum ) is a rain-fed crop with excellent tolerance to drought, sandy soil, low nutrient availability, and high temperatures. On August 2nd 2017 (day 0), we applied two levels of N to a millet field using urea (N:P:K 46:0:0), for a total of 3 treatments: 1) control; 2) medium N (87.5kg N. Ha-1); 3) high N (175kg N.Ha-1). For each treatment we had 4 plots of 400 m2 each.
We collected millet leaves twice during the experiment for nutrient analysis and Specific Leaf Area (SLA), first on August 7th (day 5) for both; and then on August 14th (day 12) for SLA and August 20th (day 18) for nutrient analysis. For both analyses we collected at least 30 leaves per treatment (3-5 g fresh mass) from different plants and from every plot. After collection, we kept the plants in an ice cooler until arrival at the phytosanitary base of Nganda (Direction de la Protection des Végétaux) where we photographed them over a 10 cm x 10 cm grid. We then dried them for 24-36 h in a drying oven (Kowell C1-I) at 60ºC and weighed them. We recorded SLA with ImageJ (National Institute of Health). We ran the chemical assays on the dried plant material in our laboratory at Arizona State University (United States). For this, we ground plant samples for 30 seconds at 200 rpm using a Retsch MM 400 ball mill. We measured plant protein content with a Bradford assay and non-structural carbohydrate content using the phenol-sulfuric acid method, following the protocol of Clissold et al. (15). We recorded stand establishment for all treatments on August 15th (day 13) by counting the number of plants over 10 m transects, with 4 transects per plot (16 transects per treatment).
On each plot we built 3-4 cages (14-16 cages per treatment) using screening mesh (Phifer, Aluminium screen mesh). Each cage was 50 cm wide and 60 cm high, poles were 90 cm rebars, and the metallic mesh was buried underground and held by 50 cm rebars. We closed the cages with staples and binder clips. Each cage contained at least two millet plants. We calculated growth by measuring plant height in the cages or for neighboring plants on August 9th (day 7) and August 18th (day 16). We collected locusts at a nearby location on August 7th (Minna, GPS coordinate 13º 49” 33.7N; 15º25’ 00.7W). The same day, we pooled and weighed 5 females and 3 males adult O. senegalensis for each cage. Every 2-3 days we checked the cages and recorded mortality. On the last day of the experiment, August 21th (day 19), the remaining locusts were removed, and we scraped the first 10 cm of soil with a trowel and manually sieved the soil for eggs. Eggs were kept in alcohol and brought back to the lab to be counted, we then dried and weighed them.
We compared O. senegalensis preference for fertilized vs.unfertilized plants by conducting a choice experiment from August 7th (day 5 after fertilization) to August 10th (day 8 after fertilization). We had two treatments: control vs. medium N, and control vs. high N. For each treatment we tested 24 grasshoppers (12 females and 12 males). Locusts were hosted in a plastic cage with two leaf blades in water tubes and a perch for roosting. Plants were collected in the same fashion as for the nutrient analysis and kept in a water bucket until the beginning of the experiment. At the end of the experiment, plant material was dried, and we calculated dry consumption using a regression equation.