Promising Results Restoring Grassland Disturbances with Native Hay (Alberta)

In Alberta, much of the once dominant rough fescue grassland has been lost to cultivation, overgrazing, and intensive oil and gas activities. Few attempts to restore rough fescue plant communities have been successful (Elsinger 2009, Desserud et al. 2010). Plains rough fescue (Festuca hallii) is a perennial bunch grass, slow growing and long lived, requiring 2 to 3 years to become established from seed. Rough fescue is an erratic seed-setter, seldom producing seed (Johnston and MacDonald 1967). The objective of our study was to assess the potential of native hay as a seed source for restoring rough fescue grassland.

The benefits of native hay include no cost for seeds, a natural mix of adapted native species, protective mulch for emerging seedlings, no requirement for special seed processing or seeding, and increased ground cover. However, the relative hardness of prairie grasses requires specialized harvesting equipment, and seed viability is unreliable. The highly variable production of seed set and the resulting dominance of species in seed at time of harvest influence seed viability in native hay (Romo and Lawrence 1990).

We found no previous research involving native hay for rough fescue grassland restoration. Experiments using native hay to restore grasslands were successful in Germany (Kiehl et al. 2006), England (Jones et al. 1995, Edwards et al. 2007), and Idaho (Gates 1962). In Idaho, native hay resulted in successful native grass establishment, while fertilizer and seeding with sawdust and conifer mulches had poor results (Gates 1962). In contrast, no native seedlings emerged from native hay application in mixed-grass prairie restoration in Saskatchewan (Wilson et al. 2004).

The study area is located in Alberta, Canada, in uncultivated rangeland in the Central Parkland natural region. Topography is an undulating complex of small depressions and hills. The soils are Dark Brown Chernozems on loam textured glacial till. Rough fescue grassland occurs on uplands and upper slopes.

Table 1.

Mean (± SD) cover (%) for selected plant species on the native hay and seeded pipeline ROWs, showing year of growth, adjacent native grassland control, and initial germination (%) from soil seed bank and native hay. Differing letters indicate significant differences within a ROW at P < 0.05. Shading indicates non-native or weedy species.

We studied 2 natural gas pipeline rights of way (ROWs) between 2006 and 2008. One was located on public land, where regulations require vegetative cover of at least 65% of predisturbance species, with no non-native species (native hay ROW). The other pipeline, 15 km to the southeast, was located on private land, and therefore was not subject to the above regulations (seeded ROW).

In August 2005, an energy company removed topsoil from the native hay ROW (15 × 150 m) before pipeline installation, which they spread back within 1 month after construction, and left the ROW unseeded. They cut hay in adjacent grassland on July 16, 2006, after plains rough fescue peak flowering in central Alberta in June and before midsummer seed shattering. A modified combine, with more durable and sharper than traditional crop blades, was used to cut about 67 m³ of hay in grassland approximately 50 to 200 m from the pipeline and immediately spray it on the ROW to a depth of 2 to 3 cm.

In July 2007, a different energy company removed topsoil from the seeded ROW (3 × 150 m) before pipeline installation and spread it back after construction. In August 2007, they seeded the ROW (approximately 15 kg/ha) with annual rye (Elymus sp.), slender wheatgrass (Elymus trachycaulus), and Rocky Mountain fescue (Festuca saximontana).

We collected monthly precipitation data at a well site (Byemoor) about 30 km east of each pipeline between April 2007 and August 2009, which we averaged with Environment Canada data from weather stations 35 km south (Craigmyle), 25 km northwest (Big Valley), and 30 km west (Trochu) of the pipelines, forming a circle around the pipeline areas.

To evaluate native hay seed content, we randomly collected 10 hay samples and spread each approximately 1 cm thick over 3 cm of potting soil (1:4 vermiculite and peat...