Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Individual study: Effect of seed addition and native ungulates on emergence of prairie forbs and grasses, Neal Smith National Wildlife Refuge, Iowa, USA

Published source details

Martin L.M. & Wilsey B.J. (2006) Assessing grassland restoration success: relative roles of seed additions and native ungulate activities. Journal of Applied Ecology, 43, 1098-1109


Areas restored as grassland often lack rare herbs and grasses that occur in native grasslands. Reasons for this may include lack of a seed source and/or microsite limitation. Native ungulates can create microsites for seedling establishment in prairie restorations by grazing dominant species and exposing patches of bare ground. This study tested for seed and microsite limitation at Neal Smith National Wildlife Refuge, the largest tallgrass prairie restoration in the USA, by adding seeds of rare native prairie plants inside and outside ungulate exclosures.

Study site: Beginning in 1992, around 1,200 ha of the 2,104 ha refuge has been seeded with tallgrass prairie species. Bison Bison bison and elk Cervus elaphus were introduced into a 303 ha enclosure (in 1996 and 1998, respectively), with around 35 bison and 15 elk present during the duration of this study. Prior land use was arable with scattered pastures. Weather in 2003 was much warmer and drier than 2004.

Planting and management: There were 20 different plantings in the enclosure (approx. 14 ha each) using seed mixes collected from local prairie remnants. Eight plantings (seeded between 1994 and 1996) were randomly selected for sampling. Management after planting included yearly spring burning during the early years followed by 2-year burn rotations, a common practice at the commencement of such restorations. Mowing was undertaken to control weeds and invasive species as required. Plots were not burned or mowed in 2003 or 2004, the years of sampling.

Experimental design: A randomized block split-plot design with grazing or exclosures (to prevent grazing) was applied to main plots and seed addition treatments (see below) in subplots. In each of the eight plantings (blocks), two 6 × 8 m grazed plots were established 5 m from, and on both sides of, a 6 × 8 m permanent exclosure in June 2003. Two grazed plots were sampled per planting because of heterogeneity with grazing.

Main plot sampling design: Biomass and above-ground net primary productivity (ANPP) were estimated in each plot in June and August 2003 and in March, June and August 2004. Live and dead biomass was divided, live material was sorted by species, and dried and weighed. Plants were designated as native or exotic. ANPP and grazing intensity were estimated for grazed plots (n = 16) during: June–August 2003, March–June and June–August 2004. One 3 × 4-m temporary exclosure was established at each site in March 2004 and was moved in June 2004 to measure consumption and grazing intensity. Above-ground biomass from the centre of each temporary exclosure was collected within quadrats. Biomass from the permanent exclosure was used to estimate consumption during June–August 2003 (i.e. the first growing season).

To determine if grazing created microsites favourable for seedling emergence, soil moisture and light intensity at the soil surface were measured at random points, monthly from July to September 2003 and May to October 2004.

Subplot seed additions: Two different seed additions of rare prairie forbs and grasess were made to separate, randomly located 1 m² subplots within each main plot, also with a control subplot (no seed addition). The first seed addition treatment consisted of adding seeds of 10 species in June 2003, and the second adding seeds of 25 species in April 2004. Seeds were added at a rate of 19,700 seeds m². Seed numbers were based on that of a typical seed rain rate found in a tallgrass prairie. Seeds were hand-scattered in each subplot and vegetation and litter shaken to help seeds reach the soil.

Species added in the first experiment were: Bouteloua curtipendula, Sporoblus asper, Solidago speciosa, Pycnanthemum virginianum, Dalea purpurea, Chamaecrista fasciculata, Amorpha canescens, Lespedeza capitata, Monarda fistulosa and Eryngium yuccifolium. Species in the second experiment included these 10, plus: Potentilla arguta, Silphium laciniatum, Echinacea pallida, Ratibida pinnata, Artemesia ludoviciana, Liatris pycnostachya, Verbena stricta, Helianthus rigidus, Gentiana andrewsii, Tradescantia bracteata, Viola pedatifida, Anemone cylindrica, Phlox pilosa, Schizachyrium scoparium and Solidago rigida.

Beginning the month after sowing, monthly seedling counts were made during the growing season. Forb seedlings were identified to species and counted in a randomly placed 20 × 50 cm quadrat within each subplot, to estimate emergence. Grass seedlings were only counted if the species had been added.

Grazing effects: Above-ground NPP/m² was 1.2, 1.1 and 8.0 times more in grazed compared with exclosed plots depending on time period (differences only significant for June-August 2004). Light availability at the soil surface was 1.7 times greater in grazed plots, and this was fairly consistent across time. Standing dead biomass and litter was marginally lower in grazed plots.

Native species seedling emergence: Adding seeds increased native seedling numbers by 2.5 times in 2003 and 2 times in 2004 in the first experiment. In the second, seedling enhancement was 3.8 in May, 5.2 in June, 5.5 in July, 6.6 in August and 17.9 in October.

Seedling emergence did not differ between grazed and exclosed plots in the first experiment but was on average 1.4 times more in grazed plots in the second. Grazing alone, without seed additions, did not increase seedling numbers. No significant relationship was apparent with grazing intensity.

Biomass was negatively related to light and soil water availability, and water was positively related to seedling enhancement more regularly and strongly than light. Grazing intensity was positively related to NPP in the first (significant for 2004 only) and second experiments, but NPP never significantly explained seedling enhancement beyond effects of light and water.

Exotic species seedling emergence:Weed seedling emergence were not clearly affected by treatments. Exotics such as dandelion Taraxacum officinale and wild carrot Daucus carota were among the most abundant species in both experiments, and hairy aster Aster pilosus and common ragweed Ambrosia artemisiifolia were also abundant in the first. The average number of exotic seedlings did not significantly differ between grazed and exclosed plots in either experiment (average number of exotics 0.1/m², n = 48; first experiment 2003: grazed 1.51, exclosed, 1.72; 2004: grazed 2.03, exclosed 1.91; second experiment, grazed 2.19, exclosed 1.92). The number of exotics was not related to grazing intensity for either experiment.

Seedling species diversity: Seed addition increased seedling diversity and richness. Average diversity in the first experiment was 1.2 times as great with seed additions in 2003 and 2004, but this difference was only marginally significant (average 1/D 0.1/m², n = 48; 2003, 2.76, control 2.24; 2004, 2.10, control 1.82). In the second experiment diversity was 1.9 times as great in the seed addition subplots than controls (average 1/D 0.1/m², n = 48; addition 3.37, control 1.82). Average richness in the first experiment was 1.3 and 1.2 times as great in seed addition than controls in 2003 and 2004, respectively, and was 2.3 times as great in addition subplots in the second experiment (average S 0.1/m², n = 48; first experiment, 2003, addition 3.81, control 2.88; 2004, addition 2.96, control 2.53; second experiment, addition 5.82, control 2.52).

Seedling species diversity did not differ between grazed and exclosed plots for either experiment. Diversity enhancement declined with grazing intensity in the first experiment but effects were unclear in the second.

Conclusions: These results suggest that this tallgrass prairie restoration is  primarily seed limited, and under certain conditions seedling emergence was enhanced by ungulate grazing when seeds were added.


Note: The compilation and addition of this summary was funded by the Journal of Applied Ecology (BES). If using or referring to this study, please read and quote the original paper, this can be viewed at: