plants per bed). Plant spacing determined bed area, which varied by treatment (Table 1, Figure 1). Treatments were re-randomized before planting in 2002.
Table 1. Treatments, bed areas, plant spacing, and plant densities used in this study (after Jeavons 2006).
| Treatment | Code | Bed area (m2) | Interplant spacing (cm) | Density (plants m-2) |
|---|
| Primary crop | Secondary crop | Overall |
| 1. Basil – pure stand | BAS | 2.5 | 30* | 13.8 | - | 13.8 |
| 2. Brussels sprout – pure stand | SPT | 6.0 | 46 | 5.9 | - | 5.9 |
| 3. Tomato – pure stand | TOM | 7.9 | 53 | 4.4 | - | 4.4 |
| 4. Tomato and basil | T&B | 4.8 | 41.5 | 2.2 | 4.5 | 7.2 |
| 5. Tomato and Brussels sprout | T&S | 6.9 | 49.5 | 1.6 | 3.1 | 5.1 |
| 6. Brussels sprout and basil | S&B | 4.1 | 38 | 2.7 | 5.3 | 8.6 |
Figure 1. Mixtures consisted of a primary and secondary crop (P and S, respectively); the primary crop was the one requiring the greater plant spacing in monoculture (c.f. Table 1). Each plot had 35 plants. Mixed plots had 12 plants from the primary crop and 23 from the secondary crop. Plant spacing in mixtures (s) was the mean of the spacing used for each crop in monoculture. Bed dimensions were a function of plant spacing.
Resource availability and use. Plots were rain-fed. Precipitation and minimum and maximum temperatures were
recorded daily at the Morgantown municipal airport, approximately 1 km from the study site. Degree days were calculated from 1 June using the single sine method and a 10 circC minimum cutoff (UC IPM online 2007). Tomato vines were suckered to leave a single dominant stem, which was staked and tied. Plots were kept weed-free by regular hand weeding. Canopy light penetration was measured by Sunfleck Ceptometer (Decagon, Pullman WA) between 1000 and 1500 h on three cloudless days in each year (Figure 2). Eight readings were taken at soil level in each bed. The upright meter was rotated 45circ with each reading, to eliminate effects due to row orientation. Each reading recorded the proportion of a 1 m probe exposed to direct sunlight (Sunfleck proportion). The mean of all eight readings was used as the sample reading for a plot.