A field study was conducted in alfalfa fields in the Arlington Agricultural Research Station, Dane County, Wisconsin, USA.

Three fields were studied. Each field was divided into three sections and in each section a one of three treatments was implemented:

1.       Control – harvesting normally, once a month (medium disturbance)

2.       Insecticide – harvesting as in the control + implementing insecticide spray (permethrin insecticide Pounce 32 EC (8 oz/ac)) immediately after harvest (high disturbance)

3.       Strips – harvesting in strips (low disturbance)

In each field and each section insects were collected during the summer of 2001 (exact dates are in the files). Insects were collected with sweep nets, the number of sweeps changed in according to the number of aphids in the field, and is given in the files for each sample. Mummies of Apidius ervi were collected by visual searches of three minutes.

The study was conducted in the Arlington Agricultural Research Station, Dane County, Wisconsin, USA.

The disturbance experiment was conducted during the summer of 2001. The experiment was run in three fields of 3.2, 3.5, and 7.5 ha that were located between 0.6 and 3.7 km apart. Each field was divided into three equal sections to which one of the following treatments was applied: control, insecticide, or strip. In the control, alfalfa was harvested normally. In the insecticide treatment, the plot was harvested normally and then the permethrin insecticide Pounce 32 EC (8 oz/ac) was applied. In the strip treatment, strips of alfalfa were left uncut during the normal harvesting event. These uncut strips were harvested two weeks later, when the cut parts of the plot had already regrown. This represents a lower degree of disturbance to the arthropods, because they can survive in the uncut strips and immigrate into the cut strips after they have started to regrow. In the first round of harvesting the uncut strips were meant to be five meters wide. This was mechanically hard to execute which resulted in uneven strips, some thinner than the desired five meters. In the second and third harvesting events, the strips were made to be the width of the harvester (7.3 meters), so the cut and uncut strips were the same width.

Insects were sampled with sweep nets from May 17 (day zero) to September 22, 2001 (day 128). Samples were taken once or twice a week depending on weather conditions (rain sometimes prevented sampling) and alfalfa height. Immediately after harvest, the alfalfa was too short for sampling, except for the uncut parts of the strip treatment. Ten samples were taken in each treatment/field on every sampling date. All samples were taken at least 5m away from the field edge. The number of sweeps per sample changed according to aphid abundance: at the extreme low aphid abundance, 40 sweeps were taken (which resulted in 400 sweeps per treatment-field), while at the extreme high abundance (more than 50 aphids per sweep) only one sweep was taken per sample (10 sweeps per treatment-field). In order to measure parasitism directly, aphids from the sweep samples were returned to the lab. Fifty adults (and fourth instars if 50 adults were not obtained) were dissected from every treatment-field on every sampling event to count second and third-instar parasitoid larvae. Finally, three 3-min sets of visual searches were conducted to assess A. ervi mummy abundances; searches were conducted by walking slowly through the field and recording visible mummies. All mummies were collected and reared in the lab in petri dishes to estimate hyperparasitism rate.