Choosing Species to Enhance Native Plant Abundance Following Biological Control of Leafy Spurge (Euphorbia esula)

In many cases biological control agents have been successful at reducing the abundance of invasive, exotic weeds (McFayden 1998, Moran et al. 2005). For example, leafy spurge (Euphorbia esula) is an herbaceous perennial native to Eurasia which has invaded over 1 million ha in North America forming dense colonies in grasslands, woodlands, and riparian areas, and reducing native species diversity (Lajeunesse et al. 1999). The most effective biological control agents to control leafy spurge are flea beetles in the genus Aphthona (Anderson et al. 2003). The small flea beetle larvae move through the soil, aggregating to feed on the roots of leafy spurge for several months, overwinter in the soil, and resume feeding the following spring (Jonsen et al. 2001). However, a reduction in target weed abundance does not always mean that native plants increase concomitantly (Butler and Wacker 2010, Setter and Lym 2013).

It is becoming increasingly apparent that seeding natives may be required in addition to biocontrol to restore native habitats (Bush et al. 2007, Reid et al. 2009). Furthermore, seeding native plants in association with biological control may help reduce the abundance of the invasive species through competition (Cutting and Hough-Goldstein 2013). We introduced flea beetles (Apthona nigriscutis) for the control of leafy spurge at three sites in north-central Montana, USA in 1994 (Lesica and Hanna 2004). Although the biomass of leafy spurge was reduced 60–69% by 2008, there was only modest recovery of forb diversity (Lesica and Hanna 2009). Our current study was designed to determine which species would be best to enhance native forb species diversity and to test the hypothesis that the slow speed of return of native forbs following biocontrol was due to low propagule pressure (Simberloff 2009). Here we report on changes in planted and resident forb abundance at these sites four years after seeding.

We conducted our study at Pine Butte Swamp Preserve on the west edge of the Great Plains in north-central Montana at 1350 m elevation (Lesica and Hanna 2004, 2009). The six circular 530-m² macroplots, two at each of three sites (Cook North, Cook South, Island), were originally part of a controlled study in which flea beetles (Apthona nigriscutis, A. lacertosa) were released into treatment plots in 1994 (Lesica and Hanna 2004); however non-release control macroplots had become colonized by flea beetles by 2000, and there has been little apparent difference in spurge density between treatment and control macroplots at the same sites for the past decade. Climate of the study area is semiarid and continental. The closest climate recording station with complete data for the study period was at Great Falls, 100 km east of the study area. Annual precipitation (September through August) was well above the 30-year average in 2011 and 2014 and somewhat below average in 2012 and 2013 at Great Falls, the closest climate recording station with complete data for the study period. Plant communities were dominated by native perennial grasses and leafy spurge in 1994 prior to biocontrol release (Lesica and Hanna 2004). At the start of this current study, western wheatgrass (Pascopyrum smithii), junegrass (Koeleria macrantha) and Kentucky bluegrass (Poa pratensis) were common graminoids at all sites. Fringed sagewort (Artemisia frigida), spotted gayfeather (Liatris punctata), nodding onion (Allium cernuum), blue flax (Linum lewisii), smooth horsetail (Equisetum laevigatum), bastard toadflax (Comandra umbellata), and tufted fleabane (Erigeron caespitosus) were common forbs. Vascular plant nomenclature follows Lesica (2012).

We laid out three permanent 13-m transects (25-m at Cook North) in each macroplot for a total of six transects/site. We randomly assigned transects in each macroplot to one of three treatments: (1) harrowed/then seeded/harrowed again (hereafter referred to as harrowed; (2) harrowed/then seeded/then rolled (hereafter referred to as rolled); and (3) control—no seed, no soil preparation. Seeds of each species were broadcast separately by hand evenly across each treatment transect. Harrowing presumably

Table 1.

Seeding density and viability as determined by standard tetrazolium tests for eight...