This research was partially funded by the Michigan Agricultural Experiment Station, the Horticulture Research Institute (HRI) and the Michigan fruit industry, and with the cooperation of the Michigan Department of Natural Resources

Introduction

Animals avoid several plant species even though they are succulent and nutritious. This may be due to smell, taste, thorns and spines, or general palatability. Daffodils and iris which have become naturalized are two examples of plants that are not eaten. In a population of daffodils, tulips, crocus and iris, deer and rabbits will selectively eat the tulips and crocus.

The best answer to the problem, but not one for the near future, is to transfer the genetic information responsible for this repellent activity into fruit trees. The alternative is to find chemicals or extracts that provide practical repellent activity. This project, initiated in 1996, has focused on extracting the repellent factors from plants and identifying them. This has been accompanied by efforts to develop an effective and economical formulation with purified extracts. Over 320 feeding studies totaling more than 2,100 field plots have been conducted.

The problem and some answers

Deer and rabbits cause severe damage to thousands of acres of fruit trees annually. This may be due to urban development which has reduced their natural habitat, as well as an increase in the populations of these animals.

To date, there are few effective controls or preventive measures, and none that are both inexpensive and practical. Commercial formulations of repellents are available which contain putrescent whole egg solids, capsaicin from peppers, garlic, denatonium benzoate (provides a bitter flavor), oil of mustard, and urine from coyotes. These are formulated either alone or in a mixture with various additives and surfactants.

Extracting repellents

The ideal animal repellent from plants would be applied as ground-up plants formulated as a dust or wettable powder. Water extracts also would be convenient. It is necessary, however, to concentrate the materials to achieve sufficient repellent activity in a smaller volume. Using organic solvents for extracting the active compounds from plant materials permits more convenient separation, concentration and analysis of active compounds. In addition, solvent extracts lend themselves to formulation into suspensions or emulsions for spraying.

Identification of active compounds

Crude extracts derived by sequential organic solvent extraction of bulbs and leaves of several plant species were fractionated by chromatographic procedures. Active fractions were determined through feeding studies. These fractions were purified further to yield four compounds with repellent activity. Structure elucidation was made by mass spectroscopy, nuclear magnetic resonance and other chemical and spectral methods. Three of these compounds were tentatively identified as alkaloids and a triglyceride with mono unsaturated fatty acid moieties. The fourth compound was identified as a sterol. In a test using corn seed treated with these compounds, squirrels consumed 100% of the control and 30, 0 and 15% of the alkaloid, triglyceride and sterol, respectively. In another test with these compounds, deer and raccoons consumed 60, 30, 20 and 13%, respectively. All of the treatments were significantly more effective than the controls. Other active compounds in daffodils and iris have been isolated and are being characterized and identified.

Procedures for testing repellent activity

Research with deer was conducted at the Michigan Department of Natural Resources Rose Lake Wildlife Center near East Lansing. Boughs of white cedar 5-8 inches long and 3-5 inches wide which were readily available and desired by deer were treated with the formulations. The boughs were tied to the fence inside an enclosure with 2 to 7 deer. Each test contained up to ten treatments placed randomly along the fence and repeated 3 times. The percent of each cedar bough consumed was estimated and recorded. The following table shows the results of three tests conducted in a pen with 4 adult deer and 3 fawns. Normal feed was withheld during all or part of the 1 - 3 day duration of the tests.

Percent of Cedar Boughs Consumed
Treatment	Rate	Test 1	Test 2	Test 3
	(mg/bough)	6/16/98	8/20/98	9/8/98
Control	Untreated	100	100	100
Putrescent egg solids	A	53	32*	0**
Denatonium benzoate	A	97	100	100
Capsaicin	A	95	100	98
Animal urine	A	---	67	67
Daffodil 138	10-15	33*	0**	0**
Daffodil 138	30-45	0**	---	0**

A - These commercial formulations were applied as recommended by the manufacturer.

Tests with rabbits were conducted using pots containing pansy plants. The plants were hardened off in growth chambers and treated with formulations containing the active fractions. They were then moved outside to areas known to have a high population of rabbits. These tests were evaluated by rating the plants based on feeding activity. Even during Michigan winters the plants remain desirable to rabbits over several weeks.

At another test site in northern Lower Michigan, a feeder which automatically dispenses corn was used to attract deer as well as wild turkeys, raccoons and squirrels. The repellent treatments were applied to corn seeds. Seven to 10 treatments of 10 seeds per plastic tray were arranged randomly among the scattered feed and replicated 3 times. The feeder was turned off during each test to prevent contamination of the treatment trays. The percentage of corn consumed from the trays was recorded.

A repellent for the future

We have extracted active repellents from 5 different plant species and identified 4 different chemicals responsible for repellent activity. Although our formulations seem to be better than the commercial products tested, we have yet to put together a formulation including surfactants and other enhancers that would be economical, stable in the environment and most importantly would remain active for long periods. As optimists, our ultimate goal is to develop a spray that will exude the "essence" of daffodils to protect fruit crops.