Drill seed rather than seeding by hand
Overall effectiveness category Evidence not assessed
Number of studies: 5
Background information and definitions
Seed-drilling is the sowing of seeds in uniform rows to a standard soil depth (Bufton 1978). This is usually done with the aid of specialized machinery called a seed drill. Sowing at depth rather than sowing seeds by hand on the soil’s surface may reduce seed removal by seed predators, such as birds. Rolling after sowing with seed drills may allow light to reach seeds to allow germination.
Bufton, L. (1978) The influence of seed-drill design on the spatial arrangement of seedlings and on seedling emergence. Symposium on the Timing of Field vegetable Production, 72.
Supporting evidence from individual studies
A replicated, randomized, paired, controlled study in 1998–1999 in a former arable field in California, USA (Montalvo et al. 2002) found that drill seeding did not increase the abundance of four of six plant species compared to applying a slurry of mulch and seeds (‘hydroseeding’). The abundance of three of six plant species was lower in areas where drill seeding was used (0.2–10.0 plants/m2) than in areas where seeds were applied in a slurry (0.3–17.0 plants/m2). However, the abundance of two species was higher where drill seeding was used (drill: 2–18 plants/m2; seeds in slurry: 1–12 plants/m2), and in one case there was no significant difference (drill: 313–370 plants/m2; seeds in slurry: 228–368 plants/m2). In February 1998, eight blocks each with three 27 × 4.5 m plots were established. In each block, one plot was seeded with a seed drill to a depth of 6–12 mm, and one plot had a slurry of seed, water, and wood fibre applied at a rate of 560 kg/ha. A straw mulch was applied to all plots at a rate of 1,680 kg/ha and a hydromulch slurry of water, wood fibre, and soil stabilizer was sprayed over the straw. In July 1998 and January and May 1999, plant abundance in each plot was estimated using six 4 × 4 m quadrats.Study and other actions tested
A replicated, paired, controlled study in 2005–2007 in three former arable fields in Iowa, USA (Yurkonis et al. 2010) found that sowing grassland seed by drill seeding resulted in similar plant species richness to sowing by hand, but higher abundance of warm-season grasses. Two years after seeding, there was no significant difference in species richness between drill-seeded plots (6.3–7.2) and plots seeded by hand (5.1–6.3). However, native warm-season grasses were more abundant in drilled plots (relative abundance: 0.23–0.34) than plots seeded by hand (relative abundance: 0.11–0.28). In spring 2005, seventy-two 12 x 12 m plots across three sites were seeded with a commercial mix of 13 forb and seven grass species at a rate of 430 seeds/m2. Half of the paired plots were drill seeded, and half were seeded by hand. Abundance of plant species was recorded in July 2007 in a randomly placed 1-m2 quadrat in each plot.Study and other actions tested
A replicated, paired, controlled study in 1999–2007 in two former arable fields in Iowa, USA (Yurkonis et al. 2010) found that drill seeding resulted in similar plant species richness to broadcast seeding at both sites, but drill seeded areas had fewer native warm-season grass and more non-native species at one of two sites. At both sites, species richness was similar in drilled (10–15 species) and broadcast-seeded areas (10–18 species). However, at one site, drill-seeded areas had a lower relative cover of native warm-season grasses (16%) and a higher relative cover of non-native species (72%) than broadcast-seeded areas (native warm-season grasses: 54%, non-native species: 35%), while at the other site there was no significant difference (native warm-season grasses: 87% vs 89%; non-native species: 9% vs 7%). At two sites, one area was drill-seeded and another was broadcast-seeded, using the same seed mix in both areas. At Site 1, one 1.9-ha area was drilled, and one 3.5-ha area was broadcast, with a local seed mix of 20 native species (sown at 16–17 kg/ha) in autumn 1999. The site was burned in spring 2004–2006. At Site 2, one 1-ha area was drilled, and one 1-ha area was broadcast, with a seed mix containing 37 forbs and nine grasses (sown at 12 kg/ha) in spring 2003. The site was mowed twice yearly. Plant species and cover were recorded in 10 random 1-m2 quadrats in each area in June 2007.Study and other actions tested
A replicated, controlled study in 2005–2010 in nine ex-arable fields in Minnesota and Iowa, USA (Larson et al. 2011) found that using drill seeding had mixed effects on the cover of sown and non-native plants and plant species richness compared with sowing by hand. In four of 10 comparisons, cover of sown plant species was higher where a seed drill was used (35–71%) than where seeds were sown by hand (33–50%). In six of 10 comparisons, cover was lower or not significantly different where a seed drill was used (drill: 0–36%; hand: 0–52%). In three of 10 comparisons, the cover of non-native plant species was lower in areas where a seed drill was used (11–34%) than where seeds were sown by hand (25–41%), but in seven of ten comparisons it was not significantly different (drill: 11–34%, hand:11–45%). In two of four comparisons, species richness was lower where a seed drill was used (10 species) than where seeds were sown by hand (12–14 species), and in two comparisons, there was no significant difference (both 10 species). In each of nine fields, thirty-six 6 x 2 m plots were established. Seeds were sown in 12 plots using a seed drill, while in 24 plots seeds were sown by hand. Seed mixes contained 10–36 species representing a mixture of grasses, legumes and non-legume forbs. In mid-June to August 2005–2007 and 2010, vegetation cover in each plot was estimated using a 4 x 0.25 m quadrat, while species richness was estimated using 6 x 2 m plots.Study and other actions tested
A replicated, randomized, controlled study in 2006–2011 in arid rangelands in Arizona, USA (Bernstein et al. 2014) found that drill seeding increased the density of the sown grasses Indian ricegrass Achnatherum hymenoides and needle-and-thread grass Hesperostipa comata compared to broadcast seeding by hand. After five years, the average density of Indian ricegrass and needle-and-thread grass was higher in drill-seeded plots (0.09–0.11 plants/m2) than in plots where broadcast seeding was done by hand (0.01–0.02 plants/m2). In November 2006, twenty 3 x 3 m plots were sown with native C3 grass seeds. Drill seeding was simulated by using a hoe to create furrows (40 cm apart, 0.6–10.1 cm deep) in half of the plots, and broadcast seeding was done by hand in the other half. Counts of grass species were made in all plots in May 2007, 2010 and 2011.Study and other actions tested