Dissertations and Theses @ UNI


Open Access Thesis


Granivores--Food--Iowa--Cedar Falls Region; Granivores--Biological control--Iowa--Cedar Falls Region;


Floral diversity of reconstructed prairies is often low compared to remnant prairies. Studies have demonstrated it is feasible to increase forb diversity in these prairies through overseeding and mowing, but overall rates of seedling establishment range from 0.1% to 1% of sown pure live seed. One explanation may be the ubiquitous populations of granivorous animals eating much of the seed . In this thesis I measured the amount of granivory occurring in a reconstructed prairie. I also tested how season, seed species, seed predator, sacrificial food, and chemical deterrents affect granivory in the reconstructed prairie. I hypothesized that granivores remove significant amounts of seeds broadcast onto an established grassland, and vertebrate and smaller granivores would prefer different species of seeds. A third hypothesis was that granivores would influence seedling establishment and it would be possible to reduce granivory through the addition of a sacrificial food or a chemical feeding deterrent. To answer these questions I glued a known number of seeds to sandpaper cards, apply various treatments to the seeds or their surroundings, and count the remaining seeds over the following weeks. The first experiment to quantify levels of seed predation involved Silphium integrifolium. During the summer of 2006,seed cards were randomly placed in 16, 5 x 5-m plots. At the wholeplot level, the plots were treated with the addition of a sacrificial food (Helianthus annus). At the within-plot level, the seeds were treated with the chemical capsaicin. I assessed the rate of removal of these seeds over an 18 day period. During the fall of 2006, this experiment was repeated with modifications. The sacrificial food and capsaicin treatments were at the whole-plot level and one of three species (Silphium integrifolium, Dodecatheon meadia, and Phlox pilosa) were at the within-plot level.

Seeds of these species were broadcast during the fall of 2006. During the spring of 2007, seedlings were counted the following spring and analyzed to detect if the amount of granivory the previous fall affected seedling establishment.

Small wire mesh exclosures were built to test for the difference in granivory by vertebrate and invertebrate granivores. Seeds of Ratibida pinnata, Sorghastrum nutans, and Dale a purpurea were placed inside on seed cards and their rate of loss was recorded.

Seed losses across trials ranged from 60% to over 98%. Significant factors included the time of year, predators involved, and species of seed. It was possible to reduce granivory in some cases. Capsaicin-treated D. meadia seeds yielded 2.1x as many seedlings as untreated seeds. Addition of sacrificial food also significantly (p=0.0006) reduced the amount of seed loss in Summer 2006 but not Fall 2006. There were significant (p

From these results, I have concluded granivory is an important factor in plant establishment within reconstructed prairies. I also found it is possible to reduce granivory, possibly increasing the success of a seed addition.

Year of Submission


Degree Name

Master of Science


Department of Biology


Tallgrass Prairie Center

First Advisor

Laura L. Jackson, Chair

Date Original


Object Description

1 PDF file (vii, 106 pages)



File Format


Included in

Biology Commons