The Pennsylvania State University

The Graduate School

College of Agricultural Sciences

ECOLOGICAL WEED MANAGEMENT: THE ROLE OF GROUND BEETLES

IN WEED SEED PREDATION

A Thesis in

Agronomy

by

Meredith J. Ward

© 2008 Meredith J. Ward

Submitted in Partial Fulfillment of the Requirements for the Degree of

Master of Science

December 2008

The thesis of Meredith J. Ward was reviewed and approved* by the following:

William S. Curran Professor of Weed Science Thesis Adviser

David M. Mortensen Professor of Weed Ecology

Mary E. Barbercheck Professor of Entomology

Tracy S. Hoover Professor and Head of the Department of Agricultural and Extension Education

David M. Sylvia Professor and Head of the Department of Crop and Soil Science

*Signatures are on file in the Graduate School

iii ABSTRACT

Weed management in cropping systems is a constant challenge for farmers. Crop- weed competition causes an overall crop yield loss of 12% annually, costing farmers around $15 billion dollars. Conservation of weed seed predators may help reduce weed infestations. This practice consists of enhancing populations of natural biocontrol agents that helps limit weed abundance. Ground beetles are one such biocontrol agent. Harpalus pensylvanicus and Amara aenea are omnivorous ground beetles that occur naturally in

Pennsylvania and in the Northeast. These beetles are known to feed on weed seeds of species frequently found in Pennsylvania farm fields. Successful integration of seed predation as a weed management practice requires the incorporation of “suitable” cropping systems at specific times throughout the growing season. By evaluating seed preference, beetle activity density and their relationship to crop phenology and timing of weed seed rain, the importance of weed seeds as a food source to H. pensylvanicus’ survival and its potential impact on weed seed abundance and weed management can be assessed. This study investigated the 1.) feeding preference of H. pensylvanicus when presented with three common summer annual weeds and its preference for fresh and aged giant foxtail (Setaria faberii Herrm.) seed, 2.) H. pensylvanicus and A. aenea activity density and seed predation in six cropping systems, and 3.) measured giant foxtail seed rain and evaluated the coincidence of giant foxtail seed rain with H. pensylvanicus activity density.

Weed species seed preference experiments were conducted using giant foxtail, common lambsquarters (Chenopodium album L.), and velvetleaf (Abutilon theophrasti

iv Medikus). The seed age preference experiment focused on fresh and field aged giant foxtail seed. H. pensylvanicus consumed giant foxtail and common lambsquarters seed, but not velvetleaf seed. When given a choice amongst the three weed species H. pensylvanicus preferred seeds of common lambsquarters and giant foxtail equally over velvetleaf seed. H. pensylvanicus consumed both newly dispersed and field aged seeds but when given the choice, preferred newly dispersed seed. This result indicates H. pensylvanicus would likely target newly dispersed weed seed rather than older seed in the seed banks.

The activity density of Amara aenea (DeGeer) and Harpalus pensylvanicus

(DeGeer) was monitored during the summers of 2004 and 2005 in five cropping systems and sweet corn (Zea mays L.) having different crop species and levels or timing of soil disturbance. Seed predation was assessed from June to September in three of the five cropping systems in 2005. In 2004, A. aenea had peak activity density in the beginning of July in a brassica/buckwheat/brassica cover crop rotation. A. aenea was not detected in

2005. H. pensylvanicus activity density peaked in early August both in 2004 and 2005. H. pensylvanicus activity density was unaffected by cropping system early in the summer; however, soil disturbance in the five cropping systems may have influenced beetle activity density in the fall. Cropping systems with little to no soil disturbances had equal or greater activity density than frequently disturbed treatments. This suggests that H. pensylvanicus may not tolerate one cropping system better than another, but may tolerate specific crop types or practices during specific times of the year. Results from the sweet corn (which followed the cropping systems) suggest that the previous year’s crop may

v not negatively affect activity density in that field the following year. Seed predation rates

in the two cropping systems averaged between 38 to 63% and rates in the sweet corn

averaged between 38 to 61%. Peak seed predation rates in the two cropping systems

occurred in early spring and in August while peak seed predation in the sweet corn

peaked towards the end of July.

Giant foxtail seed dispersal was evaluated by seeding on four dates approximately

10 days apart from mid May to mid June and measuring dispersed seed from August

through October. H. pensylvanicus activity density was monitored using pitfall traps in

foxtail plots over a 72h period. Sampling occurred from June to October. Dispersed seed

was quantified every 10 to 14 days and beetle activity density was monitored on 72 hour

intervals. In 2005, giant foxtail seed rain began in mid August and ended in late October.

The total amount of seed collected over this period averaged between 12,000 and 18,000

seeds/trap. Peak seed rain occurred in early October at the two locations. In 2006, seed

rain began in late August and did not peak until the Oct. 19 collection, then quickly

declined. H. pensylvanicus was most abundant in August. In the foxtail plots in 2006,

beetle activity density was more constant with a spike in late August with fewer beetles

captured in October and none in November.

H. pensylvanicus activity density does not appear to have coincided with the production of giant foxtail seed in Pennsylvania. Other research has suggested large crabgrass (Digitaria sanguinalis) and fall panicum (Panicum dichotomiflorum) may be better candidates (Brust, 1994) as well as yellow foxtail (S. glauca) (Curran, unpublished).These findings suggest that giant foxtail seed may not be a key food source

vi for the survival of H. pensylvanicus in the Northeast, especially in tilled cropping systems. However, H. pensylvanicus may prefer and may have coincided with other annual weed lifecycles. Additional research should focus on the phenology of weeds and the timing of weed seed rain in relationship to this and other potential weed seed predators.

vii TABLE OF CONTENTS

LIST OF FIGURES ...... ix

LIST OF TABLES...... xii

ACKNOWLEDGEMENTS...... xiii

Chapter 1 Effect of cover cropping systems and weed species on invertebrate weed seed predation...... 1

Introduction...... 1

Chapter 2 Feeding preference of H. pensylvanicus for the seeds of three summer annual weeds...... 25

ABSTRACT ...... 25 INTRODUCTION ...... 26 METHODS...... 29 RESULTS AND DISCUSSION...... 31 LITERATURE CITED...... 34

Chapter 3 Activity Density and Weed Seed Predation by Potential Weed Seed Predators in Five Cropping Systems...... 40

ABSTRACT ...... 40 INTRODUCTION ...... 41 MATERIALS AND METHODS ...... 45 RESULTS AND DISCUSSION...... 49 LITERATURE CITED...... 56

Chapter 4 Occurrence and abundance of Harpalus pensylvanicus relative to giant foxtail seed rain...... 71

ABSTRACT ...... 71 INTRODUCTION ...... 72 METHODS...... 75 RESULTS AND DISCUSSION...... 77 LITERATURE CITED...... 81

Appendix A Fact Sheets...... 90

Appendix B Lesson Plans ...... 99

Weed Identification ...... 100

viii Mechanical Weed Control: White Thread Stage ...... 113 Weed Seed Predation...... 121

ix LIST OF FIGURES

Figure 1-1: The carabid beetle, Harpalus pensylvanicus DeGeer...... 24

Figure 1-2: The carabid beetle, Amara aenea...... 24

Figure 2-1: Mean percentage of seeds consumed by Harpalus pensylvanicus over a five day exposure period. Choice experiment provided 3 seeds of each species at the same time, while no choice experiment included 9 seeds of an individual species. Columns within an experiment with the same letter are not significantly different using Tukey’s multiple comparison test at a significance level of 0.05...... 37

Figure 2-2: Percentage of newly dispersed and aged giant foxtail seeds consumed by Harpalus pensylvanicus over a five day exposure period. Choice experiment provided 4 fresh and 4 aged seeds at one time, while the no choice experiment included 8 seeds of one or the other. Columns within an experiment with the same letter are not significantly different using Tukey’s multiple comparison test at a significance level of 0.05...... 38

Figure 2-3: Percent of Harpalus pensylvanicus consumption during the duration of the ‘choice’ seed age preference trial. Choice experiment provided 4 fresh and 4 aged seeds at one time. Day and seed species was significant. Columns within an experiment with the same letter are not significantly different using Tukey’s multiple comparison test at a significance level of 0.05...... 39

Figure 3-1: Pitfall construction ...... 62

Figure 3-2: Closed seed predator trap construction...... 62

Figure 3-3: Open seed predator trap construction...... 63

Figure 3-4: Seed card construction for seed predator traps...... 63

Figure 3-5: Average Amara aenea activity density over time in five cropping systems in 2004. Sampling occurred in 72 h periods and started on the sampling dates. Activity density within a sampling date with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05)...... 64

Figure 3-6: Average Amara aenea activity density (number of beetles/pitfall) in five cropping systems in 2004. Activity density is the number of beetles captured per pitfall over a 72h period. Beetle activity density values within a cropping system with the same letter are not significantly different according

x to the Tukey test for mean separation (P < 0.05). (F, fallow; S, soybean; OP/RHV, oat-pea/ rye-hairy vetch; B/BW/B, brassica-buckwheat-brassica; O/RC, oat/red clover)...... 65

Figure 3-7: Average Harpalus pensylvanicus activity density in five cropping systems over time in 2004 and 2005. Activity density is the number of beetles captured per pitfall over a 72h period. The five cropping systems observed were fallow, soybean, oat-pa/rye-hairy vetch, brassica-buckwheat- brassica, and oat/red clover. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05)...... 66

Figure 3-8: Average Harpalus pensylvanicus activity density in five cropping systems over time in 2004 and 2005. Activity density is the number of beetles captured per pitfall over a 72h period. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation. Bars represent standard error of the estimate. (F, fallow; S, soybean; OP/RHV, oat- pea/rye-hairy vetch; B/BW/B, brassica-buckwheat-brassica; O/RC, oat/red clover)...... 67

Figure 3-9: Average Harpalus pensylvanicus activity density over time in sweet corn in 2005. Sampling occurred in 72 h periods and started on the sampling dates. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation...... 68

Figure 3-10: Average percentage of weed seed consumed in two cropping systems (fallow and oat-pea/ rye-hairy vetch) and in sweet corn in 2005. The dates indicated are the start dates for the 14 day sampling period. Percentage weed seed consumed within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation...... 69

Figure 3-11: Effect of tillage type on giant foxtail seed predation in fallow and oat/pea- rye/hairy vetch rotated into sweet corn in 2005. Averaged percent weed seed loss was observed over a 14 day sampling period. Percentage weed seed consumed tillage-type with the same letter are not significantly different according to the Tukey test for mean separation...... 70

Figure 4-1: Average giant foxtail (Setaria faberi Herrm.) seed rain over time in 2005. Two sites were sampled including a naturally established giant foxtail plot and an amended giant foxtail location. Average seed rain within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05)...... 84

xi Figure 4-2: Average giant foxtail (Setaria faberi Herrm.) seed rain over time in the amended plot in 2006. Averaged seed rain within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05)...... 85

Figure 4-3: Figure 4-3a Average influence of disturbance date and sampling date on naturally established giant foxtail seed rain in 2005. The naturally established location had a severe giant foxtail seed bank that had established in a corn-soybean rotation prior to this study. The disturbance date indicates the date at which the giant foxtail plots were tilled to kill emerged plants and stimulate the foxtail germination. Figure 4-3b Average influence of disturbance date and sampling date on amended giant foxtail seed rain in 2005. The amended location consisted of amending the seed bank by over- seeding the target weed species...... 86

Figure 4-4: Average Harpalus pensylvanicus Herrm. activity density in naturally established giant foxtail in 2005. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05)...... 87

Figure 4-5: Average Harpalus pensylvanicus Herrm. activity density in giant foxtail in 2006. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.0659)...... 88

Figure 4-6: Hypothesized Harpalus pensylvanicus activity density synchronization with giant foxtail (Setaria faberii Herrm.) seed rain in central Pennsylvania. Although there is some overlap, these two organisms are not well synchronized...... 89

xii LIST OF TABLES

Table 3-1: Time line of management events that occurred in the five cropping systems during 2004 and 2005...... 59

Table 3-2: Monthly precipitation at Russell E. Larson Research farm during 2004 and 2005 growing seasons...... 60

Table 3-3: Monthly average temperatures at Russell E. Larson Research farm during 2004 and 2005 growing seasons...... 61

xiii ACKNOWLEDGEMENTS

I want to express my gratitude to my advisor Dr. William Curran and to Drs.

Dave Mortensen, Mary Barbercheck, Tracy Hoover and the University of Maine collaborators for their guidance, assistance and perseverance throughout my thesis.

I would also like to thank all of my colleagues that have contributed to my education and research throughout this experience. I particularly would like to thank Tim

Leslie the “Carabid God”, Randa Jabbour “my little ray of sunshine”, Gretchen

Rebarchak my second mom and life preserver and Amy Lassen my closest friend, office mate, stress reliever and confidante. I do not think I could have made it through and completed my thesis without you. I miss you all.

To my cats, Velvet and Meeka, you are my comic and stress relief. All the time I was writing my thesis you two were curled up beside me and ready for playtime when I needed a break.

Finally, I want to thank my husband, Anthony, for all his help and support throughout this very long experience. You were always patient, understanding and there when I needed you throughout all of this. You and soon-to-be baby Ward are the reasons why I finished my thesis and never gave up when it would have been so easy to. Thank you for loving me through the good times, the bad and writing this thesis.

Chapter 1

Effect of cover cropping systems and weed species on invertebrate weed seed predation

Introduction

Weeds are defined as any plant growing where it is not wanted (Norris et al.,

2003). There are approximately 8,000 species of plants that behave as weeds and of those

200 to 250 are major problems in cropping systems worldwide (Holm, 1978). Most agricultural weeds have been found to possess one or more of the following characteristics: abundant seed production, rapid population establishment, seed dormancy, long-term survival of buried seeds, adaptation for spread, presence of vegetative reproductive structures, and the ability to occupy sites disturbed by humans

(Holm, 1978).

Many weeds are capable of out competing crop plants due to their rapid rates of germination and establishment. If a field is not properly managed, weeds can thrive and some species can produce viable seeds within weeks. Most cultivated crops such as corn, soybean and small grains produce several hundred seeds per plant, whereas many common summer annual weeds like giant foxtail (Setaria faberi Herrm) can produce up to 10,000 seeds, common lambsquarters (Chenopodium album L.) 72,000 seeds, and velvetleaf (Abutilon theophrasti Medicus), about 17,000 seeds per plant (Lingenfelter and

2 Hartwig, 2002). As a result, weed management in cropping systems is a constant

challenge faced by farmers. Competition between weeds and crops for limited resources

causes an overall yield loss of 12% annually worldwide, costing producers over $15

billion (Pimentel et al., 1999). To control weeds in most major crops, many farmers have

relied on synthetic herbicides. Since 1966, herbicide use in the United States has

increased four-fold. In 1997, farmers worldwide spent approximately $17 billion on

herbicides to control weeds (Liebman et al., 2001). Due to the negative impacts of

pesticides on human and environmental health, scientists, farmers and others in the

community are continually seeking alternative methods of weed control.

Conservation biological control may reduce reliance on herbicides. Weed seed predation by ground beetles is an example of conservation biocontrol. Previous research has demonstrated that some ground beetles can be responsible for up to 90% of weed seed predation in agroecosystems (Honek et al., 2003). Ground beetles in the family

Carabidae can consume up to 11 seeds daily and seed removal can be as high as 1200 to

1400 seeds m-2day-1 (Honek et al., 2003).

Harpalus pensylvanicus (DeGeer) and Amara aenea (DeGeer) are omnivorous

ground beetles that inhabit Pennsylvania and other parts of the Northeast. These carabid

beetles are believed to feed on a number of plant seeds and insects frequently found in

Pennsylvania farm fields. Minor adjustments in farming practices could increase the

abundance of ground beetles and associated ecosystem services. For example, increasing plant residues on the soil surface and decreasing tillage can favor certain seed predators.

In addition, delaying post-harvest tillage in the fall and adopting no-till systems leaves

3 weed seeds on the soil surface longer, giving carabids and other seed predators more time to consume seeds prior to incorporation into the seed bank (Liebman et al., 2001).

Effect of herbicides on Farming and the Environment

In the U.S., conventional farming (i.e. not organic) relies heavily on herbicides to control weeds. Nearly 98% of the U.S. corn and soybean crops are treated with herbicides

(NASS, 2005). In 2005, atrazine was the most widely applied corn herbicide with 66% of the crop being treated. Followed by, glyphosate at 31% of acres treated, a 19% increase from 2003. This is followed closely by s-metolachor and acetocholor at 23% of acres treated. Herbicides were also used on 98% of the U.S. soybean crop in 2005 (NASS,

2005). Glyphosate use dominated in soybean with 88% of the crop being treated with a total of 29 million kg (NASS, 2005).

Multiple factors promote the use of herbicides as a primary tool for weed management. First, herbicides reduce labor requirements for weed management

(Gunsolus and Buhler, 1999). Second, government policies have helped foster input intensive agricultural practices, providing technical advice, loans and subsidies to purchase herbicides (Pretty, 1995). Third, herbicides are effective and provide ease and low risk of crop yield loss, and fourth, basic manufacturers’ use aggressive advertising and networks of sales people to market herbicides. In addition, publicly funded research often demonstrates their utility (Benbrook, 1996). Finally, the numbers of small and medium sized farms in the US are decreasing with larger farms taking their place. These farms are more standardized in their production techniques and more dependent on

4 capital-intensive technologies (Kirschenmann, 1991). Herbicides reduce the management complexity and the need for locally adapted methods of management.

Since the 1980s, increased attention has been paid to the contamination of surface and drinking water by herbicide use (Liebman et al., 2001; United States Geological

Survey, 1999). Approximately 57% of U.S. agricultural streams contain at least one or two herbicide active ingredients that surpass at least one aquatic-life level (Gilliom and

Hamilton, 2006). Some of the herbicides most frequently found in water include acetochlor, alachlor, cyanazine and metolachor (Liebman et al., 2001; United States

Environmental Protection Agency, 1999). The concentrations of these herbicides in surface water can be directly correlated with the intensity of herbicide use (Barbash et al.,

2001; Hickman, 2004). Under normal conditions, 1 to 3% of applied herbicides are lost through surface runoff; however, during heavy rainfall as much as 5 to 10% of applied herbicides can leave the field with runoff (Crain, 2002). These streams often flow into reservoirs, lakes and rivers which provide much of the U.S. drinking water.

Many of the herbicides that are frequently detected in surface water have also been classified as probable, likely or possible carcinogens by the US EPA (1999). Several herbicides identified in water supplies have been implicated as possible disrupters of human immune, endocrine, and reproductive systems (Liebman et al., 2001; United

States Environmental Protection Agency, 1999). Much remains to be determined about the acute and chronic effects of herbicides on human health. However, public health reports and epidemiological studies indicate that certain herbicides can cause poisoning, cancer and other human health disorders (Liebman et al., 2001). Farmers, farm families

5 and agricultural workers may be subject to greater health risks as a result of being

exposed to higher concentrations over a prolonged period of time (Stone et al., 1988).

In addition to water quality and nontarget health concerns from herbicides,

herbicide resistant weeds are also a concern facing many farmers. Resistant weeds

evolve over time as the result of repeated exposure to an herbicide leading to the

predominance of genotypes that can survive and thrive when treated with herbicide rates

that would normally be fatal (Liebman et al., 2001). Before 1980, herbicide resistance

was mainly limited to the triazine herbicides and was observed in only a few weed

species (Warwick, 1991). Over the years, herbicide resistant weeds have gone from just a

few to 318 resistant biotypes in 184 species resistant to 40 herbicide active ingredients

(Weed Scientists for Weed Scientists, 2008).

Our aim is to research methods that reduce reliance on herbicides as the sole weed

control method. Conservation biological control is one weed management tactic that

could lessen our dependence on chemical weed control.

Conservation Biological Control of Weeds

Biological control of weeds is the use of suppressive agent populations to decrease weed infestations (Liebman et al., 2001). Numerous herbivores known to thrive in agroecosystems are biological control agents. Biocontrol agents represent a diversity of taxa and include: mammals, insects, nematodes, bacteria, and fungi. These organisms reduce weeds in agroecosystems by targeting multiple structures and stages throughout a plant’s lifecycle. Every plant organ and stage of the lifecycle is potentially open to attack by these organisms (Liebman et al., 2001).

6 Biocontrol agents decrease weed infestations in a variety of ways. Populations of

soil-borne fungi decrease the germinability of buried seeds. Insects defoliate weeds,

reduce plant biomass production and consume weed seeds. Nematodes attack root

systems of weeds weakening or killing the plant (Norris et al., 2003). Vertebrate and

invertebrate seed feeders can greatly enhance fungal attack by piercing the weed seed

coat. Up to 98% of velvetleaf seeds were infected by Fusarium spp. when gaps were created in the weed's seed coat by the scentless plant bug, Niesthrea louisianica Sailer

(Kremer and Spencer, 1989).

There are three approaches to using biological control to manage weed populations. Classical biological control is the introduction of host-specific, exotic natural enemies, adapted to a specific introduced weed (Stacey, 2003). This tactic focuses on a single weed species and involves the release of single to multiple species of control agents. Classical biological control can be successful because it exposes a exotic pest to its native natural enemies. This method of control works slowly, but once established it can be permanent. The majority of classical control agents are insects. Insects and other organisms that are chosen for classical biological control can easily locate the host, have high reproductive rates relative to prey, are specific to the target pest and can tolerate the climatic conditions of their host’s location (Stacey, 2003).

Inundative biological control is the mass rearing of a control agent that is released at very high numbers, inundating the site of the pest. This method of weed management relies on human intervention to increase and disperse host specific control agents onto target weed species (Stacey, 2003). Inundative control is one of the few biological weed management practices that can have immediate results. These biocontrol agents severely

7 damage and kill nearly all susceptible weeds in a matter of days or weeks. This tactic can

suppress weeds early in the growing season avoiding crop yield loss due to competition.

The biocontrol agent and targeted weeds are usually native to one another and persistence

of the agent is short-lived in the environment. More native rather than exotic biocontrol

species are used because it is easier to collect these organisms, there are less regulatory

requirements and native species have pre-adaptation to the local climate. One such

biocontrol agent is Niesthrea louisianica Sailer (Rhopalidae), an insect native to the US

(Arizona to Florida north to New York and West to Iowa). N. louisianica is commonly used for biological control of velvetleaf, a major exotic weed of corn, soybean, cotton, and sorghum. Weed infestations are reduced when the larvae and adults feed on seeds of this malvaceous plant (Spencer, 2006).

A third type of biological control is conservation biological control. Conservation biological control involves managing a site in such a way that favors naturally occurring weed suppressive organisms (Norris et al., 2003). Marino et al. (1997) assessed that a majority of seed loss could be explained by the activity of invertebrates. Therefore, a larger number of suppressive organisms may result in a larger number of plant species and lifecycle stages that can be targeted to reduce the weed population. By creating a habitat that is attractive to suppressive agents at a time when inhabiting weed populations are vulnerable, farmers can increase the effectiveness of biological control.

Biological weed management is can be reasonably permanent; achieving populations of agents that reach suppressive levels can take several field seasons. Results are not immediate and are not always apparent. Diversity and stability of biological control species within agricultural systems depends on the diversity of plant species

8 within the field, the surrounding plant communities, the type of crop grown, and how all

these factors interact (Stacey, 2003).

In some circumstances, the biocontrol agent, like the host, can also become a pest.

The agent’s host preference may not be limited to the targeted weed species. They may

influence populations of other native species that are beneficial. In addition, the

movement of some biological control agents cannot be restricted. A control agent that

may reduce weed infestations on a farm may increase in number and relocate to an area

where it can become a pest (Louda et al., 1990).

Sole reliance on of biological control may not be effective enough to limit crop yield loss. Release of natural enemies that attack weed pests may sometimes provide little reduction in crop damage depending on the life stage of the weed that is attacked by the agent. Combining biocontrol of weeds with cultural, mechanical or chemical controls can be more effective in reducing weed numbers than using a single tactic.

Weed Seed Predation

If a plant population is to thrive, individual plants must produce viable seed, disperse, establish and survive to maturity (Louda et al., 1990). Summer annual weeds such as giant foxtail, common lambsquarters and velvetleaf germinate in the spring, mature, produce seed and die in one growing season (Lingenfelter and Hartwig, 2002).

From maturity forward, weed seeds are vulnerable to the weather, desiccation and predation.

At least 95% of all plant mortality occurs during the seed lifestage (DeSousa et al., 2002). Weed seed predation is the removal of potentially viable seeds from the weed

9 seedbank (Tooley and Brust, 2002). Numerous organisms have been found to be weed

seed predators. Some of the most common are rodents, moles, ants and carabid beetles.

As an example, larvae of Harpalus rufipes, a carabid beetle, is known to dig vertical

burrows and feed exclusively on seeds lining the walls of its burrow (Zhang et al., 1997).

Seed predation has been studied in numerous cropping systems. Cromar et al.

(1999) found that invertebrates were responsible for 80 to 90% of seed predation in corn and wheat; while only 10 to 22% of seed predation was attributed to vertebrates. In particular, ground beetles were responsible for greater than 50% of measured predation in no-till and chisel-disk soybean. Fall seed predation was more than two times greater in no-till than in conventionally tilled soybean, supporting the hypothesis that tillage reduces seed predation (Brust and House, 1988).

Plant populations are reduced by seed predation by various means such as reducing sites that are suitable for germination and establishment, by overlapping peak seed predator activity density with mass seed production and by high seed predator abundance in high weed infestation areas (Zhang et al., 1997). Louda et al. (1990) found that insect weed seed predation caused a six-fold reduction in seedling establishment around adult plants and that early feeding by insects reduced the release of viable seeds three-fold. The breeding period of Harpalus species, particularly H. pensylvanicus is thought to coincide with seed rain (early autumn) of grass seeds (Holland, 2002).

A number of seed factors affect seed predation including seed abundance, seed morphology and the nutritional value, seed coat thickness, and proportion of endosperm

to seed coat (Tooley and Brust, 2002). Seeds are nutritionally rich organs. As a result,

some organisms require a mixed diet of seeds for optimal reproduction and growth. H.

10 rufipes had a 50% shorter development time when fed a mixed seed diet compared to a mixed insect diet. Further, the larval growth rate of H. rufipes was more rapid when fed common lambsquarters, groundsel and various grass seeds than on an insect diet (Zhang et al., 1997). The importance of seeds may also be directly linked to beetle fecundity

(Tooley and Brust, 2002). Female H. rufipes laid more eggs when fed a mixed seed diet than when fed a purely insect diet (Jorgensen and Toft, 1997).

Seeds can be taken by predators before or after they disperse. Predispersal seed predation occurs while seeds are still attached to the parent plant (Nurse et al., 2002), thus limiting dispersal of viable seed to the ground. De Sousa et al. (2002) found that average redroot pigweed (Amaranthus retroflexus L.) predation in corn fields ranged from 4 to

11%. More importantly, the proportion of damaged seeds per attacked infloresence was as high as 42 to 93%.

Predispersal predators are organisms that are specialized to one plant species, genus or family (Nurse et al., 2002). Predispersal weed seed predators include a range of diverse taxa including beetles, flies, ants, insect and larva. Predispersal insect predation may be responsible for up to 80% of seed mortality in many grassland and forest habitats

(Zhang et al., 1997). Ants and insect larvae were also found to be responsible for more than 70% of predispersal seed predation of understory rainforest species (Nurse et al,

2002).

Postdispersal seed predation occurs after seeds leave the plant. Seed predation can occur in the soil or on the surface (Zhang et al., 1997). Postdispersal predators are usually more mobile and their diet less specialized (Nurse et al., 2002). Cromar et al. (1999) found that postdispersal seed predators removed up to 11% of available seeds per day. In

11 no-till soybean, postdispersal seed predators can remove greater than 68% of seeds (Brust

and House, 1988). Vertebrates have been found to be responsible for 13 to 15% of

velvetleaf predation in continuous corn (Cardina et al., 1996) and 30 to 88% in

organically managed cereals (Gallandt et al., 2005).

Previous research showed that mice and ground beetles can reduce seed numbers

by 21 to 65% (Culver and Beattie, 1978). Carabid beetles, ants and rodents are the most

promising postdispersal seed predators. Predation behaviors differ greatly between each

organism. Rodents and carabids find seeds under the soil surface by olfaction and are

more efficient than ants at harvesting seeds (White and Landis, 2004; Zhang et al., 1997).

A second difference is the timing of foraging. Carabids and rodents are nocturnal while ants are mostly diurnal (Larochelle and Lariviere, 2003; Zhang et al., 1997). The third major way that rodents, carabids and ants differ is that rodents consume larger seeds first then move to smaller sized seeds while ants and carabid beetles tend to prefer small seeded weed species (Brust and House, 1988; White and Landis, 2004; Zhang et al.,

1997).

Farm routines can be altered to increase populations of weed suppressive organisms. Integrating a legume cover crop after small grains in a farming rotation, may enhance predation by providing herbivores protection from predators. Creating refuge stripes of perennial grasses at 200 m intervals around the border of crop fields can create an ideal overwintering site for ground beetles, fungi and nematodes (Liebman et al.,

2001). Increasing plant residue and decreasing tillage favors persistence of invertebrate populations. Delaying post-harvest tillage or using no-till systems leaves seeds exposed on the soil surface resulting in higher losses (Liebman et al., 2001).

12 Cover Cropping Systems and Weed Seed Predation

Annual cropping systems can be harsh environments for weed seed predators because of the high frequency and intensity of physical disturbances and xenobiotic stressors. Intensive management of agricultural landscapes such as frequent cultivation and pesticide application can negatively affect abundance, diversity and efficiency of ground beetles and other weed seed predators (Carmona and Landis, 1999). Intensive tillage methods can immediately affect emigration due to habitat disturbance and disruption of life-history processes while indirectly causing habitat deterioration and decreasing prey populations (Thorbek and Bilde, 2004). To minimize the use of herbicides and tillage to control weeds, greater attention has focused on the use of cover crops. Cover crops help improve soil physical characteristics, reduce nutrient leaching and erosion, decrease runoff, increase infiltration of rain water, help with the retention of soil moisture, add nitrogen (in the case of legume species), and can suppress weed establishment and growth (Liebman and Davis, 2000; Teasdale, 1996). Cover crops may not provide significant additional income; however, in the long run they will improve soil quality and could become an important component of a weed management plan.

Cover cropping may also affect important seed predators. For example, fall cover crops increased the activity density of carabids by 73% (Gallandt et al, 2005). Liebman and Davis (2000) also reported a 200% increase in the daily rate of seed predation in a wheat (Triticum aestivum L.) -red clover (Trifolium pratense L.) intercrop compared to a monoculture of wheat. This may be the result of the eightfold increase in cricket (Gryllus pensylvanicus Burmeister) activity density. Cover crops also produce abundant plant residue which, when left on the soil surface can form a dense mat. This mat reduces

13 losses of seed predators to higher order predators. Cromar et al. (1999) found that residue type influenced predation rates with 31% of seed lost to predation in corn, 24% in soybean, and 21% in wheat.

Carabid Beetles

Ground beetles in the family Carabidae are one of the most studied insect families. They are also one of the most diverse families: the latest worldwide count revealed 32,561 species in 1,859 genera (Holland, 2002). They were first studied because of their aesthetic appeal and ease of collection. More recently, they are being studied because of their role as bioindicators of pesticide and as biocontrol agents (Holland,

2002).

Since the 1980’s, there has been increasing interest in the toxic effects of pesticides and pesticide contamination. As a result of carabid sensitivity to chemical pollutants (i.e. heavy metal and pesticides) their distribution and abundance in agricultural settings are being used to evaluate crop management regimes. Most of the initial methodologies for using carabids as indicators of pesticide risk were developed by the International Organization for Biological and Integrated Control of Noxious Animals and Plants (IOBC). A combination of laboratory, semi-field and field tests are now used to assess the impact of new pesticides on carabid beetles and are a requirement for pesticide registration in Europe (Holland, 2002).

Ground beetles can contribute to weed seed reduction and have been identified as one of the most abundant and promising weed biocontrol agents (Carmona and Landis,

1999). The key to using carabids as biocontrol agents is understanding their preference

14 for certain weed seeds, factors that influence their feeding preferences and factors that

influence their abundance in agroecosystems.

Cardina et al. (1996) found that carabid beetles are responsible for approximately

50% of all weed seed predation that occurs in an agroecosystems. Other studies have concluded that carabid beetles may be responsible for consuming up to 90% (Cromar et al., 1999) of weed seed in long-term tillage and crop rotation studies that included corn,

soybean and wheat. Greenhouse tests show that H. pensylvanicus selectively feeds on dicot weed seeds (Holland, 2002) and their grass seed predation can reduce foxtail seedling emergence by 67% (Cromar et al., 1999). These results suggest that carabids and

possibly other weed seed predators can selectively influence the community structure of

weedy populations (Zhang et al., 1997).

Weed seeds are a major component of some ground beetles’ diet. Multiple

sources indicates (Brust, 1994; Holland, 2004; Larochelle and Lariviere, 2003) H.

pensylvanicus and other beetles may have synchronized their lifecycles to coincide with

summer annual weed seed production. If temporal coincidence of the beetle with seed

maturation and dispersal would result in the most effective biocontrol tactic producers

may be able to take advantage of this “free” weed control method by increasing beetle

habitat at a time when weed seeds disperse.

Life History of Harpalus pensylvanicus

H. pensylvanicus is a common omnivorous carabid beetle found throughout North

America and is common to Pennsylvania (Figure 1.1). This beetle is found inhabiting

lowlands and mountains; mainly, forests, grasslands, pastures, cultivated fields and field

edges. H. pensylvanicus prefers open ground, drier sites in sand or sandy loam soils,

15 covered with moderately dense but frequently tall vegetation. They are mostly nocturnal,

remaining sheltered during the day mostly in cracks in the soil or holes dug by adults and

infrequently are active during daylight hours (Larochelle and Lariviere, 2003).

H. pensylvanicus is a larger beetle 1.3 to 1.6 cm in length that is black or brown in

color with all appendages being light brown in color (White and Landis, 2004). Though it

can fly, it prefers walking and is a frequent climber and strong burrower (Holland, 2002).

They are gregarious beetles; tending to move with others of the same species or in groups

(Larochelle and Lariviere, 2003). When alarmed, H. pensylvanicus can emit a brachinus-

like smoke from their pygidigal gland.

H. pensylvanicus is abundant from January to December (Larochelle and

Lariviere, 2003). First to third instar larvae overwinter in burrows that are approximately

30 cm deep (White and Landis, 2004). A majority of larvae emerge from pupae in July-

August. The beetles copulate during June, August and September and females are gravid from July to October. Eggs are laid singly beneath the soil surface, approximately 5 to 15 cm deep. The first instar larvae hatch about three weeks after oviposition and by

October, up to third instar larvae are present in underground burrows (White and Landis,

2004). Adults overwinter in cultivated fields, vacant lots and fence rows in underground

burrows that range in depth from 8 to 20 cm (Larochelle and Lariviere, 2003). One

generation is produced per year; however, many adults can survive through the winter

and emerge in early June. The overwintered females begin ovipositing in early August

(White and Landis, 2004). New adult generations begin to emerge after mid-July and

have peak activity density in early September (Larochelle and Lariviere, 2003).

16 Beetle larvae consume seeds, insect larvae and insect adults. Larvae find weed

seeds in the soil and store them by pressing them into the walls of their burrows. Adults

consume seeds, plant tissues, pollen, fungi, other insect eggs, larva and adults. Adults

feed on seeds; particularly seeds of foxtail species, common ragweed (Ambrosia

artemiisifolia L.), pigweed species, (Larochelle and Lariviere, 2003) and common

lambsquarters (Cromar et al., 1996). Laboratory experiments showed that H.

pensylvanicus preferred foxtail seeds over velvetleaf and pigweed seed and reduced green

foxtail (Setaria viridis L.) and yellow foxtail (Setaria glauca Beauv.) seedling emergence

by 67% (White and Landis, 2004). Laboratory studies also indicated that H.

pensylvanicus can locate seeds buried up to 1 cm deep in the soil as easily as it can locate

seeds on the surface (White and Landis, 2004).

Life History of Amara aenea

Amara aenea inhabits lowlands to mountains; dwelling in lawns, gardens,

orchards, cultivated fields and field edges (Figure 1.2). This beetle tends to prefer open,

dry sandy soils with some vegetative cover. They are diurnal, being extremely active on

the ground and on plants during daylight hours. On cool or cloudy days they have been

found to seek shelter under dead leaves, weeds, stones and other debris (Larochelle and

Lariviere, 2003).

A. aenea is a small beetle approximately 0.8 to 1.0 cm in length with narrow parallel sides. The body is relatively flat with very fine lines on the elytra and is generally copperish and shiny metallic in color (Landis and White, 2004). This beetle can fly and is

17 macropterous; however, it has a preference for walking (Holland, 2002). They are also moderate runners and frequent climbers and tend to congregate in well lit areas and move towards artificial lights. A. aenea is also considered an effective colonist due to the beetle’s tendency to thrive in disturbed areas (Larochelle and Lariviere, 2003).

The lifecycle of A. aenea is not completely known. A. aenea is abundant from

January to December. Adults mate from May to June and emerge from pupae between

July and September. A. aenea overwinter as adults and emerge in spring, where they are active during April and May (Landis and White, 2004). Adults have been found to overwinter on high, dry ground, at the edges of fields, roadsides and woods. They overwinter under dead leaves, stones and in soil up to 7.5 cm deep (Larochelle and

Lariviere, 2003). Because A. aenea are abundant in crop fields in spring, they may be more likely to impact winter annual weed seeds and certain insect pests that are abundant in spring (Landis and White, 2004).

The larval diet of A. aenea is unknown; adult A. aenea beetles feed on various insect larvae, pupae and adults along with the seeds of several plant species. A. aenea feeds on seeds less than 3 mm in length (Holland, 2002) including common chickweed

(Stellaria media), large crabgrass (Digitaria sanguinalis ) and dandelion (Taraxacum officinale) (Landis and White, 2004). Carabid beetles feed by holding seeds between their tarsal spurs; as a result of this behavior, there is likely a maximum seed size upon which each beetle species is capable of feeding. They tend to avoid seeds protected by dense seed coats and glumes that make the seeds hard to manipulate (Larochelle and Lariviere,

2003).

18 This thesis research investigated the weed seed predation potential for two carabid

beetles in Pennsylvania. Several experiments were conducted from 2004 to 2006

examining the activity density of A. aenea and H. pensylvanicus and assessing giant foxtail seed rain. This research was conducted to determine the feeding preference of H. pensylvanicus for three different summer annual weed species; to determine feeding preference of H. pensylvanicus for fresh vs. field aged giant foxtail; to monitor A. aenea and H. pensylvanicus activity density in six different cropping systems managed with and without tillage, and to determine if giant foxtail seed rain is coincides with H. pensylvanicus beetle activity density. Finally, the appendices of this thesis includes

lesson plans and fact sheets used to educate young adults and future farmers about weed

seed predation and its potential in reducing the weed seed bank.

Literature Cited

Barbash, J.E., G.P. Thelin, D.W. Kolpin, and R.J. Gilliom. 2001. Major herbicides in

ground water: results from the national water-quality assessment. Journal of

Environmental Quality. 30:831-845.

Benbrook, C. 1996. Pest management at the crossroads. Mount Vernon, NY: Consumers

Union.

Brust, G. E. 1994. Carabids affect the ability of broadleaf weeds to compete.

Agriculture, Ecosystems, and Environment. 48:27-34.

Brust, G. E. and G. J. House. 1988. Weed seed destruction by arthropods and rodents in

low input soybean agroecosystems. American Journal of Alternative Agriculture.

3:19-25.

19 Cardina, J., H. M. Norquay, B. R. Stinner, and D. A. McCartney. 1996. Post-dispersal

predation of velvetleaf (Abutilon threophrasti) seeds. Weed Science. 44:534-539.

Carmona, D. M and D.A. Landis. 1999. Influence of refuge habitats and cover crops on

seasonal activity-density of ground beetle (Coleoptera: Carabidae) in field crops.

Environ. Entomology. 28:1145-1153.

Crain, A.S. 2002.The loads of selected herbicides in the Ohio River Basin, 1997-2000.

Fact Sheet-U.S. Geological Survey, Report: FS 0089-02. 4pp.

Cromar, H.E., S.D. Murphy, and C. J. Swanton. 1999. Influence of tillage and crop

residue on postdispersal predation of weed seeds. Weed Science. 47:184-194.

Culver, D. C. and A. J. Beattie. 1978. Myrmecochory in Viola. Dynamics of seed-ant

interactions in some West Virginia species. Journal of Ecology. 66: 63-72.

DeSousa, N., J. T. Griffiths, and C. Swanton. 2002. Predispersal seed predation of

redroot pigweed (Amaranthus retroflexus). Weed Science. 51: 60-68.

Gallant, E. R, T. Molloy, and R. P. Lynch. 2005. Effect of cover cropping systems on

invertebrate weed seed predation. Weed Science. 53:69-76.

Gilliom, R.J. and P.A. Hamilton. 2006. Pesticides in the Nation’s streams and ground

water, 1992-2001-A summary. Fact Sheet- U.S. Geological Survey, Report: FS

2006-3028. 6p.

Gunsolus, J. L. and D. D. Bulter. 1999. A risk management perspective on integrated

weed management. Journal of Crop Protection. 2:167-187.

Hickman, E.R. 2004. Scientific Investigations Report: SIR 2004-5105, 36pp.

Holland, J. M. 2002. The agroecology of carabid beetles. Intercept Limited. Andover,

Hampshire, UK. 1-30.

20 Holm, L. 1978. Some characteristics of weed problems in two worlds. Proceedings of the

Western Society of Weed Science 31: 3-12.

Honek A., Z. Martinkova, and V. Jarosik. 2003. Ground beetles (Carabidae) as seed

predators. European Journal Entomology. 4:531-544.

Jorgensen, B. T. and S. Toft. 1997. Role of granivory and insectivory in the life cycle of

the carabid beetle Amara similata . Ecological Entomology. 22:1-7.

Kirschenmann, F. 1991. Fundamental fallacies of building agricultural sustainability.

Journal of Soil and Water Conservation. 46: 165-168.

Kremer, R. J. and N. R. Spencer. 1989. Impact of a seed-feeding insect and

microorganisms on velvetleaf (Abutilon theophrasti) seed viability. Weed

Science. 37: 211-216.

Larochelle A. and M. C. Lariviere, 2003: A natural history of the ground-beetles

(Coleoptera: Carabidae) of America north of Mexico. 583 pp.

Liebman M. E. and A. S. Davis. 2000. Integration of soil, crop, and weed management

in low-external-input farming systems. Weed Research. 40:27-47.

Liebman, M. and E.R. Gallandt. 1997. Many little hammers: ecological management of

crop-weed interactions. Pages 291–343 in L.E. Jackson, ed. Ecology in

Agriculture. San Diego, CA: Academic.

Liebman, M., C. L. Mohler, and C. P. Staver. 2001. Ecological management of

agricultural weeds. Cambridge University Press. Cambridge. UK.

Lingenfelter, D. D. and N. L. Hartwig. 2002. Introduction to weeds and herbicides. The

Pennsylvania State University. University Park, PA.

21 Louda, S. M., M. A. Potvin and S. K. Collinge. 1990. Predispersal seed predation,

postdispersal seed predation and competition in the recruitment of seedlings of a

native thistle in sandhills prairie. American Midland Naturalist, 124:105-113.

Marino, P. C., K. L. Gross and D. A. Landis. 1997. Weed seed loss due to predation in

Michigan maize fields. Agr. Ecosyst. Env. 66:189-196.

NASS. 2007. National Agriculture Statistics Service.

http://www.nass.usda.gov/QuickStats/Create_County_Indv.jsp (accessed March

19, 2005).

Norris, R. F., E. P. Caswell-Chen and M. Kogan. 2003. Concepts in integrated pest

management. Prentice Hall: Upper Saddle River, NJ.

Nurse, R. E., B. D. Booth, and C. J. Swanton. 2002. Predispersal seed predation of

Amaranthus retoflexus and Chenopodium album growing in soyabean fields.

Weed Research. 43:260-268.

Pimental M., and D. Pimental. 1999. Population growth, environmental resources, and

the global availability of food. Social Resources. 66:417- 428.

Pretty, J. 1995. Regenerating agriculture: policies and practice for sustainability and self-

reliance. Washington, DC: Joseph Henry Press.

Spencer, N. R. 2006. Inundative biological control of velvetleaf Abutilon theophrasti

(Malvaceae) with Niesthrea louisianica (Hem.: Rhopalidae). Biocontrol. 33:421-

429.

Stacey, D.A. 2003. Climate and biological control in organic crops. International Journal

Pest Management. 49:205-214.

22 Stone, J. F., M. L. Eichner, C. Kim and K. Koehler. 1988. Relationship between clothing

and pesticide poisoning: symptoms among Iowa farmers. Journal of

Environmental Health. 50:210-215.

Teasdale, J. R. 1996. Contribution of cover crops to weed management in sustainable

agricultural systems. Journal of Production Agriculture. 9:475-432.

Thorbek, P. and T. Bilde. 2004. Reduce number of generalist arthropod predators after

crop management. Journal of Applied Ecology. 41:526-538.

Tooley, J and G. E. Brust. 2002. The agroecology of carabid beetles. Intercept Limited.

Andover, Hampshire, UK. 215-229.

United States Environmental Protection Agency. 1999. Pesticidal chemicals classified as

known, probable and possible human carcinogens. Washington, D.C.: U. S.

Environmental Protection Agency, Office of Pesticide Programs.

United States Geological Survey. 1999. The quality of our nation’s waters: nutrients and

pesticides, Circular 1225. Washington, DC: US Department of Interior and

US Geological Society.

Warwick, S. I. 1991. Herbicide resistance in weedy plants: physiology and population

biology. Annual Review of Ecology and Systematics. 22:95-114.

Weed Scientists for Weed Scientists. 2008. International survey of herbicide resistant

weeds. http://www.weedscience.org/in.asp (accessed April 27, 2008).

White, S. and D. Landis. 2004. Biocontrol agent profile Harpalus pensylvanicus.

Technical Bulletin. Michigan Sate University.

http://www.cips.msu.edu/biocontrol/fact_sheets/H.%20pen.htm

23 Zhang, J., F. Drummond, M. Liebman, and A. Hartke. 1997. Insect predation of seeds

and plant population dynamics. Technical Bulletin 163. University of Maine.

http://www.umaine.edu/mafes/elec_pubs/techbulletins/tb163.pdf

24

Figure 1-1

Scale

Figure 1-1: The carabid beetle, Harpalus pensylvanicus DeGeer.

Figure 1-2

Scale Figure 1-2: The carabid beetle, Amara aenea

Chapter 2

Feeding preference of H. pensylvanicus for the seeds of three summer annual weeds

ABSTRACT

Harpalus pensylvanicus (DeGeer) is a known weed seed predator found in the

northeastern United States. H. pensylvanicus is believed to feed on a number of different

weed seeds, although few experiments have investigated species preference. This study

investigated the feeding preference of this carabid beetle for three common summer

annual weeds. Secondly, the preference for fresh and aged giant foxtail (Setaria faberii

Herrm.) seed was also investigated.

Over a two-year period, carabid beetles were captured in the field and laboratory

feeding trials were conducted. Weed species seed preference experiments that included

choice and no choice treatments were conducted using giant foxtail, common

lambsquarters (Chenopodium album L.), and velvetleaf (Abutilon theophrasti Medikus).

H. pensylvanicus consumed giant foxtail and common lambsquarters seed, and did not consume velvetleaf seed. When given a choice amongst the three weed species, H.

pensylvanicus preferred seeds of common lambsquarters and giant foxtail equally over

velvetleaf seed. This suggests that H. pensylvanicus does show preference among seed

species and does not equally prefer seed species. In the seed age preference experiment,

H. pensylvanicus consumed both newly dispersed and field aged giant foxtail seeds.

However, when given the choice, H. pensylvanicus preferred newly dispersed giant

26 foxtail seeds over field aged seeds. These results showed that H. pensylvanicus will consume two smaller seeded annual weeds that possess brittle seed coats over a larger seeds with a thicker seed coat even though all three seeds are common to many northeastern agroecosystems. Secondly, the age preference experiment suggests that H. pensylvanicus may better impact newly dispersed weed seeds rather than aged seed or pre-established seed banks.

INTRODUCTION

At least 95% of all plant mortality occurs at the seed stage and seed predation is an important form of seed mortality (DeSousa et al., 2002). Weed seed predation is the removal of seeds from the weed seed bank (Tooley and Brust, 2002). Seed predators can affect plant and community dynamics by influencing the density and distribution of seeds in the soil (Cromar et al., 1999). Louda et al. (1990) found that weed seed predation by insects can cause a six-fold reduction in seedling establishment and that feeding by insects reduced the release of viable seeds three-fold.

The main seed predators in agricultural fields have been identified as large and small carabid beetles, mice, ants, insect larvae and crickets. There are two types of weed seed predation; predispersal and postdispersal. Predispersal weed seed predation is predation of weed seeds while they are still attached to the parent plant; postdispersal predation occurs once the seed has left the parent plant (Nurse et al., 2002). Cromar et al.

(1999) reported that postdispersal seed predators can remove up to 11% of available seeds per day and these predators are important in reducing seed supply and seedling

27 emergence in deserts, tropical regions, agricultural fields, grasslands, and forests (Cardina et al., 1996).

Ground beetles in the family Carabidae have been identified as one of the most abundant and promising postdispersal weed seed predators (Carmona and Landis, 1999).

Several studies showed that carabid beetles are responsible for about 50% of all weed seed predation in agroecosystems (Cardina et al., 1996). Additional studies concluded that carabid beetles may remove as much as 80 to 90% of available weed seeds from the weed seed bank (Cromar et al., 1999).

Postdispersal seed predation can also affect community dynamics by modifying the competitive abilities of the component species through selective seed feeding

(Cromar et al., 1999). A seed predator’s preference or ability to consume a particular species could contribute to the decline of dominant species and allow more obscure species to thrive (Zhang et al., 1997). A number of factors affect seed predation including seed abundance, seed morphology, nutritional value, seed coat thickness, proportion of endosperm to seed coat, seed size (Tooley and Brust, 2002) and oil content (Lund and

Turpin, 1997).

Ground beetles feed on a wide variety of weed seeds. A laboratory study established that five species of carabid beetles common to corn fields fed on small grass and broadleaf seeds (Best and Beegle 1977). Barney and Pass (1986) found that H. pensylvanicus, a common ground beetle in the northeastern U.S., fed on small seeds of weedy grasses while inhabiting alfalfa fields. Harrison and Schmoll (2003) found that

62% of a sample population of H. pensylvanicus preferred small seeded species such as smooth pigweed (Amaranthus hybridus L.) and yellow foxtail (Setaria glauca (L.)

28 Beauv.), whereas 17% preferred the larger seeded, giant ragweed (Ambrosia trifida L.),

and 21% had no preference between species. Brust and House (1988) observed that two

species of Harpalus readily fed on the seeds of larger broadleaved weeds such as sicklepod (Senna obtusifolia (L.) Irwin & Barneby), jimsonweed (Datura stramonium L.) and common ragweed (Ambrosia artemisiifolia L.). In another study, four carabid beetle species consumed seed from 12 grass and broadleaved weed species, but did not consume velvetleaf (Lund and Turpin 1977). Cardina et al. (1996) showed that two different carabid beetle species, including H. pensylvanicus, would not consume unimbibed or imbibed velvetleaf seeds.

H. pensylvanicus is a omnivorous carabid beetle widely distributed throughout

North America, preferring open ground and dry soil covered with moderately dense vegetation (Larochelle and Lariviere, 2003). H. pensylvanicus is abundant from January to December in the northeastern U.S. and has peak activity density in early September

(White and Landis, 2004) feeding on pollen, fungi, seeds and other insect eggs, larvae and adults (Larochelle and Lariviere, 2003).

Previous research by Leslie et al. (2007) showed that H. pensylvanicus was frequently identified in farm fields at the Russell E. Larson Research farm in Centre

County. A number of annual grass and broadleaved weed species that infest the research farm fields may serve as a potential food source for this carabid beetle. Therefore, the first objective of this research examined the preference of H. pensylvanicus for the seeds of giant foxtail, common lambsquarters, and velvetleaf, three weeds common to

Pennsylvania field crops. The second objective evaluated the preference of H. pensylvanicus’ for seeds of different ages, either newly dispersed giant foxtail seed or

29 seed that overwintered underground, to help determine if age of seed could affect the rate of seed predation.

METHODS

Experiments were conducted in the laboratory from July through September during 2004 and 2005. H. pensylvanicus adults were collected using pitfall traps at the

Penn State University Russell E. Larson Research and Education Center in Centre

County. Pitfall traps consisted of a plastic container 11 cm in diameter and 14 cm deep.

The top half of a 2 l plastic soda bottle was inverted and placed in the opening of the plastic container to act as a funnel and to direct invertebrate organisms into the collection cup. Six pitfall traps were randomly placed in a field about 1 ha in size that was planted to cover crops. The pitfall traps were emptied after 72 h and live adult beetles were transported to the lab and held individually in aerated 14 by 1.75 cm Petri dishes for the feeding experiments. Each dish contained 1 to 2 grams of field soil and a 1 ml water- filled glass vial with cotton stopper. Food was withheld for 48 h prior to exposure to the weed seeds to decrease the effect of previous feeding experience in the field on the experimental results. After the 48 h pretreatment, a single beetle was placed in a 14 by

1.75 cm Petri dish along with the desired weed seeds and a water source. The inside bottom of each dish was lined with double-sided tape covered lightly with sand to mimic the soil surface.

Weed Species Preference. Giant foxtail, common lambsquarters, and velvetleaf seed were used for the preference experiments. Choice and no choice treatments were

30 conducted with all three weed species. The ‘no choice’ treatments assessed if H.

pensylvanicus would consume the seeds of a particular weed species, while the ‘choice’

treatment determined if H. pensylvanicus’ preferred one species over another. The ‘no

choice’ treatments included nine seeds of an individual weed species per dish, while the

‘choice’ test contained three weed seeds from each of the three weed species (total of

nine seeds per dish). The number of seeds used in the ‘no choice’ tests and the ‘choice’

test were the same to decrease the effect of seed density on consumption. Single beetles

were confined to a dish for 5 days and the number of seeds consumed was examined

daily. Consumption was determined by counting the number of intact seeds that

remained and subtracting that from the original quantity. Treatments were replicated 10

times and the experiment was repeated 5 times over the period of two summers.

Weed Seed Age Preference. Giant foxtail seed were harvested from mature plants

in early September. The “field aged” seed treatment was established by placing

approximately 1000 freshly harvested seed sealed in nylon pouches 7 cm by 7cm (air and

water permeable) and placed 15 cm beneath the soil surface starting in early October

2004 in a cultivated field at the Russell E. Larson Research farm. Seed pouches were

exhumed from the soil in early April 2005 prior to any germination and the seeds were

stored at 5°C until the experiment was conducted in late summer, in this way the aged

seed were approximately 12 months older than the “freshly harvested” seed.

As in the previous experiment, both ‘choice’ and ‘no-choice’ treatments were included. The ‘no choice’ test exposed one beetle to 8 seeds of either the fresh or aged giant foxtail. The ‘choice’ test included four aged and four freshly harvested seed. As in the previous experiment, single beetles were confined to a dish for 5 d and the number of

31 consumed seeds was examined daily. Consumption was determined by counting the number of intact seeds remaining and subtracting from the original quantity. This experiment consisted of 10 replications and was repeated 5 times.

Data Analysis. The seed species and seed age preference results were quantified using methods similar to Cardina et al. (1996), by assessing the number of intact seeds that were recovered at the end of the five-day period. Average seed consumption was calculated for each seed treatment. Tests of normality indicated that data transformation was not necessary to satisfy the assumption of ANOVA. Data were analyzed using the general linear models (GLM) procedure of SAS (SAS Institute, 2003). Preference comparisons involving seed species and age were made using Tukey’s multiple comparison test at a significance level of 0.05.

RESULTS AND DISCUSSION

Weed Species Preference. No choice tests evaluating seed species preference established that H. pensylvanicus would consume both giant foxtail and common lambsquarters seeds, but not velvetleaf seeds (Fig 2.1) (P<0.0001). Common lambsquarters and giant foxtail seeds were either intact or completely consumed after the five-day exposure period. At the end of five days, more than 4 seeds or almost half of both giant foxtail and common lambsquarters were consumed by the carabids, whereas all velvetleaf seeds remained uneaten. These results are consistent with Lund and Turpin (1977) and Cardina et al. (1996) who showed that H. pensylvanicus does not

32 consume velvetleaf seeds. Velvetleaf is a dicot species with seeds that are approximately

3 mm long with a hard, dense outer seed coat. Carabid beetles feed by holding seeds

between their tarsal spurs and as a result it may be that velvetleaf is too large to handle.

This seed is also protected by a dense outer seed coat that is harder to manipulate and

feed upon (Larochelle and Lariviere, 2003).

In the choice experiment, H. pensylvanicus preferred the smaller seeded species

(giant foxtail and common lambsquarters) over velvetleaf (Fig 2.1) (P<0.0001) which

agrees with the previous experiment and with Harrison and Schmoll (2003). H.

pensylvanicus did not exhibit a significant feeding preference between the giant foxtail

and the common lambsquarters consuming 1.7 and 2.1 seeds on average, respectively

(total of 3 seeds per species) during the five day period. These results support the findings

of Lund and Turpin (1977) who observed the feeding preference of H. pensylvanicus

among sixteen broadleaf and grass weed species. Lund and Turpin found that H.

pensylvanicus showed no difference in preference between giant foxtail and common

lambsquarters in Indiana cornfields and that velvetleaf was not consumed. Both giant

foxtail and common lambsquarters are small with brittle outer seed coats, this allows for

less energy expenditure and for greater energy consumption. As a result, seed predator

consumption can cause once dominant species to decline and allow more obscure species

to thrive (Zhang et al., 1997).

Weed Seed Age Preference. Results from the no choice seed age experiment showed that H. pensylvanicus would consume both fresh and field aged giant foxtail

seeds, but the number of fresh seeds consumed was greater than aged seed (P<0.0001)

(Fig 2.2). Over a five day period, the carabids consumed over 3.76 fresh seeds and only

33 about 1.2 aged ones (out of 4 seeds). The no choice experimental results indicate that when given a suitable food source, Harpalus will consume old or new seed though they prefer newer seed.

When given a choice, H. pensylvanicus preferred the newly dispersed giant foxtail seeds over the field aged giant foxtail seed (Fig 2.2) (P<0.0001). Over the five day test, an average of 3.8 fresh seeds were consumed while 2.2 aged seed were consumed. More importantly, the beetles tended to consume nearly all the newly dispersed seeds before starting to feed on the field aged seeds (Fig 2.3) (P<0.0001). Most of the fresh seeds were consumed by the fourth day, while less than 40% of the aged seeds had been eaten by this time. There may be several reasons for the differences observed in this experiment.

Even though seed quality was not tested in this study, the field aged seeds may have had lower quality compared with the fresh seeds. Environmental exposure causes changes in permeability in the seed’s plasma membranes resulting in the loss of seed oil, amino acids, sugars and other intracellular components (Golovina et al. 1997). In addition, seed odor, presence of pathogens, toxicity, and overall appearance could have also contributed to the preference of H. pensylvanicus for fresh giant foxtail (Tooley and Brust, 2002).

These results suggest that preference of H. pensylvanicus for newly dispersed seed could directly affect weed recruitment by reducing new seed bank inputs more than aged or the seeds from previous years. H. pensylvanicus may differentially affect the weed seed bank at the time of seed dispersal compared with pre-established weed seeds.

H. pensylvanicus is an opportunistic, omnivorous feeder that has not exhibited a clear preference for seed species (other than having little to no preference for velvetleaf) on the basis of size or chemical composition in previous research (Best and Beegle, 1977; Brust,

34 1994; Brust and House, 1988; and Harrison and Schmoll, 2003). However, H. pensylvanicus may determine preference based on the size and shape of a seed and how they directly affect the ease of handling and opening the seed rather than selection based on textural and chemical characteristics alone (Lund and Turpin, 1977). Our experiments clearly demonstrate a non-preference for velvetleaf and a preference for fresh giant foxtail over field aged seed. Rates of seed predation would be expected to vary with weed species and with the ability and tolerance of the predator. The results observed in these laboratory experiments may not be representative of occurrences in the natural environment where H. pensylvanicus may have a greater or different preference for weed species either included or not included in this study.

LITERATURE CITED

Barney, R. J. and B. C. Pass. 1986. Foraging behavior and feeding preference of ground

beetles (Coleptera:Carabidae) in Kentucky alfalfa. Journal of Economic

Entomology. 79:1334-1337.

Best, R. L. and C. C. Beegle. 1977. Food preference of five species of carabids

commonly found in Iowa cornfields. Environment Entomology. 6:9-12.

Brust, G. E. 1994. Carabids affect the ability of broadleaf weeds to compete.

Agriculture, Ecosystems, and Environment. 48:27-34.

Brust, G. E. and G. J. House. 1988. Weed seed destruction by arthropods and rodents in

low-input soybean agroecosystems. American Journal Alternative Agriculture.

3:19-25.

35 Cardina J., H. M. Norquay, B. R. Stinner and D. A. McCartney. 1996. Postdispersal

predation of velvetleaf (Abutilon theophrasti) seeds. Weed Science. 44:534-539.

Carmona, D. M and D.A. Landis. 1999. Influence of refuge habitats and cover crops on

seasonal activity-density of ground beetle (Coleoptera: Carabidae) in field crops.

Environ. Entomology. 28(6):1145-1153.

Cromar, H. E., S. D. Murphy, and C. J. Swanton. 1999. Influence of tillage and crop

residue on postdispersal predation of weed seeds. Weed Science. 47:184-194.

DeSousa, N., J. T. Griffiths, and C. Swanton. 2003. Predispersal seed predation of

redroot pigweed (Amaranthus retroflexus). Weed Science. 51: 60-68.

Golovina, E. A., A. N. Tikhonov and F. A. Hoekstra. 1997. An electron paramangnetic

response spin-probe study of membrane-permeability changes with seed aging.

Plant Physiology. 114:383-389.

Harrison, S. K. and J. T. Schmoll. 2003. Postdispersal predation on giant ragweed

(Ambrosia trifida) seed in no-tillage corn. Weed Science. 51:955-964.

Larochelle A.and Lariviere M. C. 2003: A natural history of the ground-beetles

(Coleoptera: Carabidae) of America north of Mexico. 583 pp.

Leslie, T. W, G. A. Hoheisel, D. J. Biddinger, J. R. Rohr and S. J. Fleischer. 2007.

Transgenes sustain epigeal insect biodiversity in diversifies vegetable farm

systems. Environmental Entomology. 36:76.

Louda, S. M., M. A. Potvin and S. K. Collinge. 1990. Predispersal seed predation,

postdispersal seed predation and competition in the recruitment of seedlings of a

native thistle in sandhills prairie. American Midland Naturalist, 124:105-113.

36 Lund, R. D. and F. T. Turpin. 1977. Carabid damage to weed seeds found in Indiana

cornfields. Environmental Entomologist. 6:695-698.

Nurse, R. E., B. D. Booth, and C. J. Swanton. 2003. Predispersal seed predation of

Amaranthus retoflexus and Chenopodium album growing in soyabean fields.

Weed Research. 43:260-268.

SAS Institute. 2003. The SAS system for Windows. Release 9.1. SAS Inst. Cary N. C.

Tooley, J and G. E. Brust. 2002. The agroecology of carabid beetles. Intercept Limited.

Andover, Hampshire, UK. 215-229.

White, S. and D. Landis. 2004. Biocontrol agent profile Harpalus pensylvanicus.

Technical Bulletin. Michigan Sate University.

http://www.cips.msu.edu/biocontrol/fact_sheets/H.%20pen.htm

Zhang, J., F. Drummond, M. Liebman, and A. Hartke. 1997. Insect predation of seeds

and plant population dynamics. Technical Bulletin 163. University of Maine.

http://www.umaine.edu/mafes/elec_pubs/techbulletins/tb163.pdf

37 Figure 2-1

100 No Choice n 90 Choice 80 a 70 a 60 A A 50 40 30 20

Percentage consumptio B b 10 0 Giant Foxtail Velvetleaf Common Lambs quarters Seed Type

Figure 2-1: Mean percentage of seeds consumed by Harpalus pensylvanicus over a five day exposure period. Choice experiment provided 3 seeds of each species at the same time, while no choice experiment included 9 seeds of an individual species. Columns within an experiment with the same letter are not significantly different using Tukey’s multiple comparison test at a significance level of 0.05.

38

Figure 2-2 No Choice a 100 90 80 70 60 50 40 b 30 20 10 0

Choice a 100 90 80 Percentage consumption consumption Percentage 70 b 60 50 40 30 20 10 0 Newly Disperesed Field Aged Seed Type

Figure 2-2: Percentage of newly dispersed and aged giant foxtail seeds consumed by

Harpalus pensylvanicus over a five day exposure period. Choice experiment provided 4 fresh and 4 aged seeds at one time, while the no choice experiment included 8 seeds of

one or the other. Columns within an experiment with the same letter are not significantly

different using Tukey’s multiple comparison test at a significance level of 0.05.

39

Figure 2-3

Aged Newly Dispersed 100 e d 90 n 80 70 c 60 E 50 b

Precentage Consumptio Precentage 40 D

30 a 20 BC 10 AB A 0 Day1 Day 2 Day 3 Day 4 Day 5 Feeding Trial Duration

Figure 2-3: Percent of Harpalus pensylvanicus consumption during the duration of the

‘choice’ seed age preference trial. Choice experiment provided 4 fresh and 4 aged seeds at one time. Day and seed species was significant. Columns within an experiment with the same letter are not significantly different using Tukey’s multiple comparison test at a significance level of 0.05.

Chapter 3

Activity Density and Weed Seed Predation by Potential Weed Seed Predators in Five Cropping Systems

ABSTRACT

Successful integration of seed predation into weed management program requires cropping practices that enhance ground beetle abundance at specific times throughout the growing season. Activity density of Amara aenea (DeGeer) and Harpalus pensylvanicus

(DeGeer), two potential weed seed predators, was monitored during the summers of 2004 and 2005 in five cropping systems differing in crop species and level of disturbance.

Activity density was also monitored in sweet corn (Zea mays L.) following a cover or cash crop. Finally, seed predation was assessed from June to September in two of the five cropping systems and in the sweet corn cash crop during the summer of 2005. In

2004, A. aenea peak activity density occurred in early July in a brassica/buckwheat/brassica cover crop rotation. A. aenea was not detected in 2005.

Activity density of H. pensylvanicus peaked in early August in both years. Soil disturbances in the five cropping systems may have influenced beetle activity density in the fall; resulting in H. pensylvanicus abundance being highest in oat/pea-rye/hairy.

Cropping systems with little or no soil disturbance had equal or greater beetle activity densities than frequently disturbed treatments. This suggests that H. pensylvanicus may have less tolerance for one cropping system over another, but may tolerate specific crop types during specific times of the year. Results from the sweet corn (which followed the

41 initial five cropping systems) suggest that the previous year’s crop many not affect activity density in that field the following year. Seed predation rates in the two cropping systems averaged between 38 to 63% and rates in the sweet corn averaged between 38 to

61%. Peak seed predation rates in the two cropping systems occurred in early spring and in August, whereas peak seed predation in the sweet corn peaked towards the end of July.

Open traps did not suffer higher seed losses during the study, suggesting that vertebrate seed predation played a minor role in this experiment.

INTRODUCTION

Seed predation can significantly contribute to weed management. Seed consumption may occur while the seed is attached to the parent plant (predispersal) and/or after seeds are released (postdispersal) (Nurse et al., 2002). Postdispersal seed predation can contribute up to 90% of weed seed bank loss (Cromar et al., 1999).

Rodents, ants, ground beetles and crickets are some of the most important postdispersal weed seed predators. Ground beetles (Carabidae), can consume up to 50% of weed seeds in agroecosystems (Cardina et al., 1996) and are commonly found inhabiting old fields, pastures, cultivated fields and their borders (Larochelle and Lariviere, 2003). These beetles feed on a wide variety of weed seeds; however, they tend to prefer small seeded grass and broadleaved seed (Best and Beegle, 1977).

A number of studies have examined factors that influence seed predation by ground beetles (Carmona and Landis, 1999; Cromar et al., 1999; Gallandt et al., 2005;

Thorbek and Bilde, 2004). House (1989) found tillage reduced soil arthropod density. In

42 addition, Brust and House (1988) observed that seed predation rates in no till soybeans

were 2.3 times higher than in conventionally tilled fields. In a recent study in Maine,

moldboard plowing and rotary tillage reduced Harpalus rufipes (Degeer) activity-density

by 53 and 55%, when compared to no-tillage (Shearin et al., 2007).

Carabid abundance may be influenced by timing and frequency of soil

disturbance. Tillage and mowing operations, as well as the disturbance brought about by

crop harvesting, can drastically decrease insect populations. Immediate population

reductions include direct mortality by injury or burial. Delayed effects of mechanical

management include emigration due to habitat disturbance, and indirect effects consist of

habitat deterioration, removal of plant cover, decrease of prey, alteration of microhabitats

and negative effects on life-history processes (Thorbek and Bilde, 2004). Decreasing the

frequency of soil disturbance, creating refuge strips, and temporal variation in tillage can

result in higher seed predator abundance.

Types of ground cover may also affect the abundance of seed predators. Dense canopy cover provides habitat for seed predators and can increase seed predator populations and the amount of seeds consumed (Cromar et al., 1999; Gallandt et al.,

2005; Kromp, 1999). Higher rates of seed predation have been reported in plots with vegetative cover than those without (Gallandt et al., 2005). Higher beetle abundance is the likely result of several interacting factors including greater food and shelter, a more suitable microclimate and ease of movement. A major influence is also exerted by the phenology of plant canopy development. Summer annual row crops often have a long period of bare soil or low soil cover in early spring resulting in an extreme soil-surface environment with regard to temperature, moisture and exposure. As summer annual row

43 crops mature, the habitat becomes more favorable with a dense plant canopy, cooler

temperatures near the soil surface, higher humidity and greater protection from predators.

Winter cereals on the other hand provide a more favorable environment in early spring by

providing early canopy cover, lower fluctuations in temperature and moisture and

reduced exposure (Krompt, 1999). In the early spring, the fully developed winter cereal

canopy provides protection for ground beetles; while still allowing good mobility as the

ground surface remains open (Melnychuk et al., 2003).

Crop rotation may influence seed predation rates. A study by Nijs et al. (1996)

found that a sugar beet followed by winter wheat increased the abundance of Poecilus

cupreus L. The sugar beet crop allowed the larva and young carabid beetle adults to

develop undisturbed, while the winter wheat crop provided good conditions for

overwintering. However, when winter wheat was followed by sugar beet, beetle activity

densities decreased. After winter wheat harvest, a drastic drop in humidity in mid

summer increased larval mortality which was further promoted by unfavorable

overwintering conditions in the wheat stubble and fallow period before sugar beet

planting.

In Pennsylvania, Harpalus pensylvanicus and Amara aenea are two common granivorous carabid beetles that inhabit agroecosystems. These beetles populate lowlands, mountains, forests, grasslands, pastures, cultivated fields and their borders and prefer open ground and dry soil covered with moderately dense vegetation. H. pensylvanicus is abundant from January to December in the northeastern U.S. and has peak activity density in early September (Larochelle and Lariviere, 2003). This beetle is

omnivorous in its feeding habits, consuming pollen, fungi, seeds and other insect eggs,

44 larvae and adults (Larochelle and Lariviere, 2003). The lifecycle of Amara aenea is less

understood; however, it is abundant from January to December. Although the larval diet

for A. aenea is unknown, adult beetles feed on various insect larvae, pupa and adults

along with foliage and seeds from several plant species. Previous research demonstrated

that A. aenea preferred seeds less than 3 mm in length (Holland, 2002).

Understanding how cover crops and soil disturbance influence resident seed predators is critical to the success of conservation biocontrol. Although previous research suggests that vegetative ground cover and lower disturbance levels can aid in increasing seed predators, the conditions and practices by which these two ground beetle species can be conserved in agricultural settings is poorly understood. For these reasons, the first objective of this study was to monitor A. aenea and H. pensylvanicus activity density in five cover cropping systems that differed in plant species and timing of soil disturbance. The second objective quantified prevalence of ecosystem seed predation in these systems with an added fallow treatment maintained with tillage. This information will help us evaluate the system level effect and the effect that specific crops may have on conserving weed seed predators. Utilizing seed predation by ground beetles in conjunction with other weed control measures in an integrated approach could make a valuable contribution to reducing the weed seed bank.

45 MATERIALS AND METHODS

Cropping Systems. A field study evaluating the effect of five cropping systems on

weed seed bank dynamics was conducted at Penn State’s Russell E. Larson Agricultural

Research Farm in Centre County (40o44’ N, 77o57’ W). Within this cropping system

study, ground beetle activity density was monitored and a seed predation experiment was

conducted during the summers of 2004 and 2005. The five cropping systems included

(1) bare fallow where weeds were controlled with tillage and no crop was grown, (2)

organically managed soybean (Glycine max Merr.) that relied on mechanical weed

control, (3) yellow mustard, (Sinapis alba L. ‘Idagold’) rotated to buckwheat

(Fagopyrum esculentum Moench), and then winter canola, (Brassica napus L. `Dwarf

Essex’) (brassica/buckwheat/brassica). In the third system, the three different cover

crops were planted in sequence as green manure crops and the yellow mustard and

buckwheat were flail mowed at full flower and mechanically incorporated with a rotary

cultivator followed by a finishing tool (cultimulched) prior to seeding the winter canola.

The fourth system (4) included spring oat (Avena sativa L. ‘Ogle’) plus field pea (Pisum sativum L. ‘Maxum’) followed by cereal rye (Secale cereale L. ‘Aroostook’) plus hairy vetch (Vicia villosa Roth). The oat/pea mixture was managed as a green manure as described previously for system 3 and the fifth system (5) included spring oat plus red

clover (Trifolium pretense L. ‘Mammoth’). The spring oats were rotary mowed at full

panicle emergence and prior to seed set and the red clover was rotary mowed twice about

30 days apart to prevent weed seed production. A more detailed description of field

operations is provided in Table 3.1. It was intended that these five cropping systems

46 represent a wide range of disturbance frequencies and disturbance types (tillage and mowing) as well as crop species and dry matter production. Individual plots were 12.2 by 12.2 m and the experiment was designed as a randomized complete block with four replications.

All five cropping system treatments were planted to sweet corn (Zea mays L.

‘Kandy Kwik’) the following year. For the sweet corn crop, main plots were split into no- till and conventional tillage subplots (split plot = 6.1 by 12.2 m) to examine the effects on incorporated vs. surface residues. Cover crops were either tilled or treated with a herbicide in mid May. In the no-till subplot, 0.84 kg ae/ha glyphosate plus 0.28 kg ae/ha

2,4-D ester were applied for control of the cover crops and any emerged weeds approximately 10 days prior to corn planting. For the conventional tillage treatments, the cover crops were first flail mowed, rotary cultivated, and then cultimulched prior to corn planting. These mechanical treatments were conducted at 3 day intervals. Weeds emerging after crop planting were managed with a single in-row cultivation in the conventional tillage plots and with a postemergence herbicide (35 g ai/ha nicosulfuron plus 104 g ai/ha mesotrione plus 1% crop oil concentrate) in the no-tillage plots. Both the cultivation and the postemergence herbicide application occurred about four weeks after corn planting.

Activity Density. Experiments assessed A. aenea and H. pensylvanicus activity density in the different cropping systems and in a sweet corn crop. Beetle activity density was defined as the average number of beetles captured per pitfall during the 72 h sampling period. Pitfall traps consisted of a 950 ml plastic container (height 10.9 cm x diameter 11.4 cm) with a 240 ml polystyrene cup filled about a third full with ethylene

47 glycol (killing agent) placed in the bottom of the plastic container. The specimen collection cup allowed the traps to be checked without removing the entire trap from the soil. The top half of a 2000 ml plastic soda bottle was inverted placed in the 950 ml plastic container to direct invertebrate organisms into the specimen cup (Figure 3.1).

Three pitfall traps were placed flush with the soil surface in each cropping system main plot or each tillage subplot in the sweet corn year and arranged every 3 m along a horizontal transect in the center of each plot. Pitfall captures were based on a 72 h collection period. When not in use, traps were covered to prevent the unnecessary capture of beetles and to keep out debris and water. Pitfall trapping took place two times each month from July to October in 2004 (total of 4 times) and from May to September in

2005 (total of 6 times). Specimens were identified and the activity density was determined for these two carabids. Activity density was assessed in the sweet corn crop only in 2005.

Weed Seed Predation. In 2005 only, seed predation was assessed in the (1) fallow and in the (4) oat/pea followed by rye/hairy vetch systems (as described previously) and sweet corn plots that were fallow and oat/pea followed by rye/ hairy vetch in 2004. These two cropping systems were selected due to extreme differences in vegetative cover (little or none in fallow system and up to 100% in the cover crop system) during the period of the experiment. The experimental design was a split plot replicated four times with cropping system as the main plot factor and cage type was the subplot factor. In addition, the effect of tillage (no-till vs. conventional as previously described) was also examined in the sweet corn crop.

48 Six cages were positioned randomly along a horizontal transect in each plot.

Three of the six cages were closed traps (Figure 3.2) consisting of galvanized metal hardware cloth (3.3 by 3.3 cm openings) which prevented access by small mammals and birds, but allowed most invertebrates. The closed traps were cylinders, 14 cm diameter by

9 cm tall, capped with a 14 cm diameter clear lid to repel rain. Open traps (Figure 3.3) were constructed without metal hardware cloth which permitted access to vertebrate and invertebrate seed predators. These traps consisted of three 18 cm bolts suspending a 14 cm diameter clear lid to keep out rainfall. The open traps were accessible by both ground dwelling vertebrates and invertebrates.

Within each trap a 10 cm Petri dish was placed bottom side up and buried so that the dish was flush with the soil surface. The surface of the Petri dish was covered with double-sided indoor/outdoor carpet tape and 100 giant foxtail (Setaria faberi Herrm.) seeds were randomly sprinkled on the tape surface. In preliminary studies we found that

H. pensylvanicus will readily consume giant foxtail seeds (Murray et al., 2005).

Approximately 1 to 2 g of field soil was sprinkled on the tape to remove stickiness and better mimic the soil surface (Figure 3.4).

Weed seed predation rate was based on the removal of giant foxtail seed over a

14-day period repeated five times during the summer beginning in late May and ending in early September. During the course of the summer, traps were moved then returned to their original position when agronomic management practices were preformed

Statistical Analysis. The activity density and seed predation experiments included observations with repeated measures over time arranged in a completely randomized design. Because of the complexity and dissimilarity of the five cropping

49 systems, direct comparison of disturbance timing and type is difficult to make.

Appropriate covariance structures that took into consideration the degree of heterogeneity of variance and covariance between repeated measurements were selected. Activity density sampling in the five cropping systems did not occur on the same dates each year or within the same period of time between sampling periods because of differences in timing of field activities and rainfall. For this reason, activity density data could not be pooled across years and was analyzed separately each year. Beetle activity density was analyzed using a univariate analysis of variance (ANOVA) with the mixed procedure of

SAS (SAS Institute Inc., 2004). This procedure is designed for models that have both fixed and random effects. Treatment comparisons were made using Tukey’s multiple comparison, with significance (P-value) set at less than or equal to 0.05. Seed predation data were analyzed as described for the activity density experiment with the exception of cage type being included as a second independent variable.

RESULTS AND DISCUSSION

Weather conditions varied substantially in 2004 and 2005. Total rainfall were near normal for May and June of 2004, but was two to three times above average from July through September (Table 3.2) causing flooding and standing water in the field experiment during some of the sampling times. In contrast, crops were under drought stress for much of the season in 2005, with severe drought conditions in August and

50 September. Air temperatures during summer 2004 were near normal and slightly elevated throughout the summer in 2005 (Table 3.3).

Activity Density-Cropping Systems. The activity density of Amara aenea during

2004 across the five cropping systems averaged 1.5 beetles/pitfall (total number captured

= 89). A. aenea had peak activity density in the beginning of July (P< 0.0001) decreasing in early August and no beetles were observed in September and October 2004 (Figure

3.5). More beetles were captured in the brassica/buckwheat/brassica crop system in than the other four cover cropping systems (P< 0.0006) (Figure 3.6). In this cover cropping system, a lack of soil disturbance along with a dense canopy cover of the flowering yellow mustard in early July may have provided good habit. A. aenea was almost absent in 2005 in both the cover crop treatments and the sweet corn. Previous research shows that A. aenea is most active in early spring and the larval stage dominates in late summer

(Larochelle and Lariviere, 2003). We can only speculate, but the heavy rains and flooding in late summer 2004, may have caused high mortality for the soil dwelling larvae resulting in the lower activity density in 2005. Due to the relatively low activity density observed for A. aenea in both 2004 and 2005, no strong conclusions can be made with regard to the systems tested. We conclude that it is unlikely that A. aenea plays a significant role in weed seed predation in our study system.

H. pensylvanicus was the more abundant of the two ground-dwelling invertebrates. In 2004, the average activity density was 1.9 beetles/pitfall (total number captured = 115). In 2005 beetle numbers increased greatly with almost 6600 beetles captured. In 2005, activity density over the six sampling dates averaged 15 beetles/pitfall. Higher rates of H. pensylvanicus activity density in 2005 may have been

51 the result of the drier, warmer climate compared to the wet conditions, particularly during

July through September in 2004. In addition, although the experiment was repeated over years, different fields on the same farm were used each season also potentially affecting beetle activity density. H. pensylvanicus had peak activity density in the five cropping systems beginning in August for both the 2004 (P< 0.0489) and 2005 (P< 0.0001) (Figure

3.7). Beetle activity density increased from the earliest sampling dates through early

August in both years after which beetle numbers declined. In October 2004, no beetles were captured; this may be a consequence of the heavy rains and cooler temperatures. In

2005, beetles activity density was declining in September and sampling did not take place in October.

Although cover crop type was not significant, an interaction was observed between sampling date and cropping system for both 2004 (P< 0.0209) and 2005 (P<

0.0346) (Figure 3.8). Although there was no apparent preference or tolerance for cropping system through mid-July either year, from mid-July to September, activity density increased with peak activity density occurring in the oat-pea followed by rye- hairy vetch system in early August both in 2004 and 2005. The oat-pea cover crop was flail mowed during the first half of July and allowed to lay fallow for over a month. In early August, the dead cover crop residue covering the surface with no disturbance appeared to favor beetle activity density. By September, a small increase in activity density was observed in the fallow treatment and the oat/red clover systems (in 2005 only) and decreased in the brassica/buckwheat/brassica (in both 2004 and 2005) and oat- pea followed by rye-hairy vetch systems (in both 2004 and 2005). The fallow treatment contained a few immature weeds with little residue and the last tillage event occurred in

52 mid August, in contrast the red clover offered a dense legume stand. The reduction in activity density in the brassica/buckwheat/brassica and oat-pea followed by rye-hairy vetch treatments could be the result of early September tillage and a small amount of surface residue in the mid September when H. pensylvanicus densities were peaking in other treatments.

The changes in activity density indicate that H. pensylvanicus may not tolerate one specific crop throughout the growing season, but may tolerate certain crops at different times throughout the growing season or during the insect’s life cycle. In the spring and early summer, H. pensylvanicus showed little preference or tolerance between cropping systems, while in late summer and early fall, activity density tended to be greatest in the cropping systems with little soil disturbance. These results correspond to when adult carabids are in one of their most vulnerable stages (breeding and reproducing) and probably tolerate habitats with fewer disturbances (Thorbek and Bilde, 2004).

Activity Density-Sweet Corn. In the sweet corn in 2005, H. pensylvanicus activity density averaged 13.5 beetles/pitfall and neither the previous cropping system

(P=0.9195) nor tillage treatment (P=0.2424) affected activity density. As in the cropping systems study, peak activity density in the sweet corn occurred in early to mid August and began to decline in September (P< 0.0001) (Figure 3.9). Although the cropping systems and sweet corn experiments were in different fields about 0.5 km apart, beetle numbers in the sweet corn were similar to those found in the cropping systems study being collected at the same time. In the sweet corn, cover crops were either mechanically incorporated with tillage or killed with herbicides in mid May, so large differences in the amount of plant residue on the soil surface were present early in the

53 season depending on the system. The tilled plots were again disturbed in early July with row cultivation and the no-till plots received a post-emergent herbicide application.

Although early summer cover crop and weed management activities were quite different in the no-till and tilled treatments, these differences did not influence beetle activity density. By mid July, sweet corn was actively growing and some weeds were present in both the tilled and no-till systems, potentially providing suitable habitat for H. pensylvanicus.

The lack of effect in the no-tillage treatments in the sweet corn could have also been due to promotion of greater insect predator populations along with increased seed predators potentially masking differences. A more favorable predator habitat and greater biological diversity in no-till plots could increase seed predator populations, but may have also increased predator populations and food source diversity (Cardina et al., 1996).

However, we did not monitor for beetle predators. These results suggest that beetles may not favor familiar locations and may migrate to more tolerable habitats over time. This supports Gallant et al. (2005) findings that all organisms have survival instincts, and even though they are familiar with a certain area they will migrate to new locations that promise greater opportunities for survival.

Weed Seed Predation. In 2005, average giant foxtail seed predation in the fallow and oat-pea followed by rye-hairy vetch system varied through the growing season from

38 to 63% (Figure 3.10). Similar rates of seed predation were also observed in the sweet corn (38 to 61%) (Figure 3.10). Trap type did not affect predation in the cropping system

(P=0.2003) or sweet corn (P=0.7733) suggesting that vertebrates played a minor role in seed predation in this study. Cromar et al. (1999) found that invertebrates were

54 responsible for 80 to 90% of seed predation in long-term tillage and crop rotations studies

that included corn, soybean and wheat. They attributed 10 to 22% of total seed predation

to vertebrates.

Cropping system type did not influence seed predation either during the cover

crop sequence (P=0.8678) or in the sweet corn (P=0.0712). However, as with beetle

activity density, sampling date (P< 0.0001) affected rates of predation in both systems

(Figure 3.10). Rates of seed predation were similar for three of five sampling periods in

the fallow and cover crop treatment. Predation rates dropped in the late July and

September sampling periods. During July, the fallow treatment was cultivated to kill

emerged weeds, the brassica treatment was cultivated and seeded to buckwheat and the

oat/pea cover crop was mowed. Tillage or low surface residue conditions also existed in

three of five systems during September. All of the situations could likely reduce insect

foraging. In addition, comparatively high predation rates in June suggest that predators

other than H. pensylvanicus contribute to seed losses, since H. pensylvanicus activity

density in June was relatively low (Figure 3.7). Also the abundance of other food sources

during the sampling periods could have influenced giant foxtail seed predation.

In the sweet corn, seed predation rates peaked in late July (Figure 3.10). The higher

predation rates in early August are in agreement with H. pensylvanicus beetle activity density for the sweet corn (Figure 3.9), although other predators certainly may have contributed to the giant foxtail seed losses. In addition, an interaction between cover crop system and tillage occurred in the sweet corn (P< 0.0005) where the tilled treatment in the oat/pea followed by rye/hairy vetch had less predation, while the no-till and tilled fallow plots and no-till cover crop treatment were similar to one another (Figure 3.11).

55 Because the tilled cover crop treatment received the most disturbance (i.e. mowed,

cultivated, and cultimulched) prior to planting sweet corn, we hypothesize that

disturbances could have reduced some seed predator populations. The tilled fallow

treatment had similar management prior to sweet corn planting minus mowing. As

mentioned previously, H. pensylvanicus activity density was not different across tillage

treatments and predation rates ranged from 50 to 65% regardless of treatment, suggesting

that differences due to tillage were relatively small. Although no differences in cage type

were observed, predators other than invertebrates could have also played a role.

High levels of cumulative seed predation by invertebrates in late summer and fall correspond to the time of maximum foraging activity by some seed-eating carabid adults

(Cromar et al., 1999; Harrison and Schmoll, 2003; Lund and Turpin, 1977). In addition, the life cycle of ground dwelling seed predators may have affected weed seed predation.

Seasonal peaks in beetle abundance occur in late May and June and in late August and

September, depending on the species and the climate zone (Cromar et al., 1999). In this research, we observed at least 60% giant foxtail predation throughout the summer in three of five sampling periods in the cover crop treatments and 38 to 61% predation in sweet corn. Distinct early and late summer peaks did not occur in our research further suggesting that several seed predators likely contributed to seed losses in this research.

However, rates of seed loss can also be influenced by rain, wind, or movement of non- predatory organisms, so the results of this and other studies should be interpreted with caution (Brust and House, 1988; Cardina et al., 1996).

Observations during this experiment suggest that higher frequencies of tillage could decrease the level of H. pensylvanicus activity density in certain crops. However,

56 these results were not consistent, implying that this carabid may not prefer one cropping

system throughout the entire growing season but may tolerate a specific crop type during

a specific time in the season. Results from the sweet corn (which followed the cover

cropping systems) suggest that the previous year’s crop may not greatly affect activity

density in that field the following year. Shearin et al. (2007) found Harpalus rufipes was

twice as likely to travel to “more suitable”habitats. In addition, peak seed predation rates

in the two cropping systems occurred in early spring and in August while peak seed

predation in the sweet corn peaked towards the end of July. This suggests that seed

predators may immigrate according to seed abundance.

Future research should focus on the effect of food preference and tillage timing on

H. pensylvanicus and other potential seed predators. Weed seed predation is an important component of a farmer’s weed management routine. More time should be spent evaluating specific crops and rotations that create “suitable habitats” for these beneficial organisms. Also, plot size, crop arrangement and experimental design should be monitored to clearly test the mechanisms that create “suitable habitats”. By creating

“suitable habitats,” farmers could maximize the effectiveness of indigenous seed predators.

LITERATURE CITED

Brust, G. E. and G. J. House. 1988. Weed seed destruction by arthropods and rodents in

low input soybean agroecosystems. American Journal of Alternative Agriculture.

3:19-25.

57 Best, R. L. and C. C. Beegle. 1977. Food preference of five species of carabids

commonly found in Iowa cornfields. Environmental Entomology. 6:9-12.

Cardina, J., H. M. Norquay, B. R. Stinner, and D. A. McCartney. 1996. Post-disperal

predation of velvetleaf (Abutilon theophrasti) seeds. Weed Science. 44:534-539.

Carmona, D. M and D.A. Landis. 1999. Influence of refuge habitats and cover crops on

seasonal activity-density of ground beetle (Coleoptera: Carabidae) in field crops.

Environmental Entomology. 28:1145-1153.

Cromar, H.E., S.D. Murphy, and C. J. Swanton. 1999. Influence of tillage and crop

residue on postdispersal predation of weed seeds. Weed Science. 47:184-194.

Gallant, E. R, T. Molloy, R. P. Lynch. 2005. Effect of cover cropping systems on

invertebrate weed seed predation. Weed Science. 53:69-76.

Harrison, S. K. and J. T. Schmoll. 2003. Postdispersal predation on giant ragweed

(Ambrosia trifida) seed in no-tillage corn. Weed Science. 515:955-964.

House, G. J. 1989. Soil arthropods from weed and crop roots of an agroecosystem in a

wheat-soybean-corn rotation: impact of tillage and herbicides. Agriculture,

Ecosystems, and Environment. 25:233-244.

Holland, J. M. 2002. The agroecology of carabid beetles. Intercept Limited. Andover,

Hampshire, UK. 1-30.

Krompt, B. 1999. Carabid beetles in sustainable agriculture: a review on pest control

efficacy, cultivation impacts and enhancement. Agriculture, Ecosystems, and

Environment. 74:187-228.

Larochelle A. and M. C. Lariviere. 2003: A natural history of the ground-beetles

(Coleoptera: Carabidae) of America north of Mexico. 583 pp.

58 Lund, R. D. and F. T. Turpin. 1977. Carabid damage to weed seeds found in Indiana

cornfields. Environmental Entomology. 6:695-698.

Melnychuk, N. A., O. Olfert, B. Young and C. Gillott. 2003. Abundance and diversity of

Carabidae (Coleoptera) in different farming systems. Agriculture, Ecosystems,

and Environment. 95:69-72.

Murray, M. J., W. S. Curran, D. A. Mortensen, and M. E. Barbercheck. 2005. Ecological

Weed Management: The Role of Ground Beetles in Weed Seed Predation.

Proceedings of the Northeast Weed Science Society 59.

Nurse, R. E., B. D. Booth, and C. J. Swanton. 2002. Predispersal seed predation of

Amaranthus retoflexus and Chenopodium album growing in soyabean fields.

Weed Research. 43:260-268.

SAS Institute, Inc. 2004. SAS/STAT 9.1 user’s guide. SAS Institute, Inc., Cary, NC.

Shearin, A., C. B. Reberg-Horton and E. R. Gallandt. 2007. Cover crop effects on the

activity density of the weed seed predatory Harplaus rufipes. M. S. Thesis. Orono,

ME: University of Maine.

Thorbek, P. and T. Bilde. 2004. Reduced number of generalist arthropod predators after

crop management. Journal of Applied Ecology. 41:526-538.

59

Table 3-1: Time line of management events that occurred in the five cropping systems during 2004 and 2005. Table 3-1 System Field 1-2004 Field 2-2005 Management Fallow 5/12-5/24 5/27 Field cultivation (2x) 6/15 6/29 Field cultivation (2x) 7/9 7/20 Field cultivation (2x) 8/17 8/15 Field cultivation (2x)

B-BW-B 5/12-5/24 5/5 Field cultivation (2x) 5/24 5/6 Plant cover crop 7/9 6/29 Mow 7/17 7/7 Field cultivation (2x) 7/17 7/7 Plant buckwheat cover crop 9/5 9/15 Mow with flail mower 9/11 9/17 Field cultivation (2x) 9/14 9/18 Plant canola cover crop

O/P- R/HV 4/28-4/30 5/5 Field cultivation (2x) 4/28-4/30 5/6 Plant O/P cover crops 7/9 7/7 Flail mow 8/18 8/24 Field cultivation (2x) 8/18 8/29 Plant R/HV cover crop

O/RC 4/28-4/29 5/5 Field cultivation (2x) 4/28-4/29 5/6 Plant O/RC cover crop 7/9 7/7 Mow O at boot stage 7/22, 8/02, 8/16 7/19 Mow tops of clover (2-3x)

Soybean 5/12-5/24 5/9-5/20 Field cultivation (2x) 5/24 5/20 Plant crop 5/28, 6/5 5/27, 6/3 Tine weed 6/20, 7/1 6/17 Row cultivation 11/1 10/31 Harvest grain

60

Table 3-2: Monthly precipitation at Russell E. Larson Research farm during 2004 and 2005 growing seasons. Table 3-2

61

Table 3-3: Monthly average temperatures at Russell E. Larson Research farm during 2004 and 2005 growing seasons. Table 3-3

62 Figure 3-1

Figure 3-1: Pitfall construction

Figure 3-2

Figure 3-2: Closed seed predator trap construction.

63 Figure 3-3

Figure 3-3: Open seed predator trap construction.

Figure 3-4

Figure 3-4: Seed card construction for seed predator traps.

64

Figure 3-5

3.5 a

3

2.5

2

1.5

1 Activitydensity (# beetles/pitfall

0.5 b b b 0 7/9 8/9 9/24 10/24 Sampling date

Figure 3-5: Average Amara aenea activity density over time in five cropping systems in

2004. Sampling occurred in 72 h periods and started on the sampling dates. Activity density within a sampling date with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05).

65

Figure 3-6

l 6 b 5

4

3

2

1 a a a a 0 Beetle activitydensity (#beetles/pitfal F S OP/RHV B/BW/B O/RC Cover crop

Figure 3-6: Average Amara aenea activity density (number of beetles/pitfall) in five cropping systems in 2004. Activity density is the number of beetles captured per pitfall over a 72h period. Beetle activity density values within a cropping system with the same letter are not significantly different according to the Tukey test for mean separation (P <

0.05). (F, fallow; S, soybean; OP/RHV, oat-pea/ rye-hairy vetch; B/BW/B, brassica- buckwheat-brassica; O/RC, oat/red clover)

66

Figure 3-7 2004

3 b 2.5

a ) 2 a 1.5 itfall p 1

0.5 a 0 #beetles/ ( 7/9 8/9 9/24 10/24 y

densit 2005 y 50 45 b 40 35 30 c

Beetle activit 25 20 15 a a 10 5 a a 0 6/10 6/24 7/15 7/29 8/12 9/15 Sampling date Figure 3-7: Average Harpalus pensylvanicus activity density in five cropping systems over time in 2004 and 2005. Activity density is the number of beetles captured per pitfall

over a 72h period. The five cropping systems observed were fallow, soybean, oat-pa/rye-

hairy vetch, brassica-buckwheat-brassica, and oat/red clover. Mean beetle activity density

(#beetles/pitfall) within sampling dates with the same letter are not significantly different

according to the Tukey test for mean separation (P < 0.05).

67

Figure 3-8

2004

5 Fallow 4 Soybean 3 OP/RHV 2 B/BW/B 1 O/RC

0 7/9 8/9 9/24 10/24 -1

-2 Sampling date

2005 80 70 60 50 Fallow

Beetle activity density (#beetles/pitfall) 40 Soybean 30 OP/RHV 20 B/BW/B 10 O/RC 0 -10 6/10/05 6/24/05 7/15/05 7/29/05 8/12/05 9/15/05 -20 Sampling Date

Figure 3-8: Average Harpalus pensylvanicus activity density in five cropping systems over time in 2004 and 2005. Activity density is the number of beetles captured per pitfall over a 72h period. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation. Bars represent standard error of the estimate. (F, fallow; S, soybean;

OP/RHV, oat-pea/rye-hairy vetch; B/BW/B, brassica-buckwheat-brassica; O/RC, oat/red clover)

68

Figure 3-9

50 c 45 40 35 30 c 25 20 15 b b 10

Activity density (# beeltes/pitfall 5 a a 0 6/10 6/24 7/15 7/29 8/12 9/15 Sampling date

Figure 3-9: Average Harpalus pensylvanicus activity density over time in sweet corn in

2005. Sampling occurred in 72 h periods and started on the sampling dates. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation

69

Figure 3-10 Cover crops

70 a a a 60

50 b b 40

30

20

10

0

Sweet corn

70 c 60 b 50 b

Percentage of weed seed consumed a a 40 30 20 10 0 6/15-6/29 7/15-7/29 7/29-8/11 8/11-8/24 9/1-9/15 Sampling date

Figure 3-10: Average percentage of weed seed consumed in two cropping systems

(fallow and oat-pea/ rye-hairy vetch) and in sweet corn in 2005. The dates indicated are the start dates for the 14 day sampling period. Percentage weed seed consumed within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation.

70

Figure 3-11 70 a F 60 a OP/RHV d a 50 b 40

30

20 Percentage consume Percentage 10

0 No-Till Till Tillage type

Figure 3-11: Effect of tillage type on giant foxtail seed predation in fallow and oat/pea- rye/hairy vetch rotated into sweet corn in 2005. Averaged percent weed seed loss was observed over a 14 day sampling period. Percentage weed seed consumed tillage-type with the same letter are not significantly different according to the Tukey test for mean separation.

Chapter 4

Occurrence and abundance of Harpalus pensylvanicus relative to giant foxtail seed rain

ABSTRACT

Harpalus pensylvanicus (DeGeer) is a common granivorous carabid beetle found throughout North America and is abundant from January to December (Larochelle and

Lariviere, 2003). Several common weed seeds that H. pensylvanicus is known to consume are foxtail species (Setaria spp.), common ragweed (Ambrosia artemisiifolia), and pigweed species (Chenopodium album ) (Cromar et al., 1996). Previous research suggests that specific species of carabid beetles have co-evolved with major food sources and that the lifecycle of H. pensylvanicus has coincided with the ripening of some summer annual grassy weeds (Brust, 1994; Brust and Tooley, 2002).

Experiments were conducted at two locations in 2005 and a single location in

2006 at Russell E. Larson Research farm in Centre County, PA. Giant foxtail was seeded on four dates approximately 10 days apart from mid May to mid June. Giant foxtail seed rain was determined by collecting shed seeds from August through October. H. pensylvanicus activity density was monitored using pitfall traps in foxtail plots over a 72h period. Sampling occurred from June to October.

In 2005, giant foxtail seed rain began in mid August and ended in late October.

The total seed collected over this period averaged between 12,000 and 18,000 seeds/trap.

Peak seed rain occurred in early October at the two locations. In 2006, seed rain began in

72 late August and did not peak until the October 19 collection, then quickly declined. H.

pensylvanicus was most abundant in August. In the foxtail plots in 2006, beetle activity

density was more constant in August and September with a spike in late August with

fewer beetles captured in October and none in November.

H. pensylvanicus activity density does not appear to coincide with the production of giant foxtail seed in Pennsylvania. Other research has suggested large crabgrass

(Digitaria sanguinalis), fall panicum (Panicum dichotomiflorum) or yellow foxtail (S. glauca) (Curran, unpublished) may be better candidates (Brust, 1994).These findings suggest that giant foxtail seed may not be a key food source for the survival of H. pensylvanicus in the Northeast, especially in tilled cropping systems. However, H. pensylvanicus may prefer and may have synchronized with other annual weed lifecycles.

Additional research should be focused on the phenology of weeds and the timing of weed seed rain in relationship to this and other potential weed seed predators

INTRODUCTION

H. pensylvanicus (DeGeer) is a common omnivorous carabid beetle found throughout North America including Pennsylvania. H. pensylvanicus is abundant in farm fields and field edges from January to December (Larochelle and Lariviere, 2003).

Although only one generation is produced per year, adults can survive through the winter and emerge in early June. The overwintered females begin ovipositing in early August

(White and Landis, 2004) and eggs are laid singly beneath the soil surface, approximately

73 5 to 15 cm deep. First to third instar larvae overwinter in burrows that are approximately

30 cm deep (White and Landis, 2004). The following year, the majority of adults emerge

from pupae from early July to mid August having peak activity density in early

September (Larochelle and Lariviere, 2003).

H. pensylvanicus larvae consume seeds, other insect larvae and insect adults.

Adult beetles consume seeds, plant tissues, pollen, fungi, insect eggs, larva and other insect adults. Several common weed seeds including foxtail species, common ragweed

(Ambrosia artemisiifolia L.), pigweed species (Amaranthus spp., (Larochelle and

Lariviere, 2003) and common lambsquarters (Chenopodium album L.) (Cromar et al.,

1996) can be a part of their diet.

Cardina et al. (1996) reported that carabid beetles are responsible for approximately 50% of all weed seed predation that occurs in agroecosystems. Other studies have concluded that carabid beetles consume up to 90% of weed seed in long- term tillage and crop rotation studies that included corn, soybean and wheat (Cromar et al., 1999). Cromar et al. (1999) further reported that H. pensylvanicus reduced foxtail species emergence by 67%, although a laboratory experiment showed that it preferred seeds of broadleaved species when they are mixed with grass seeds (Holland, 2002).

Results from other experiments indicate that H. pensylvanicus can locate seeds buried up to 9.5 mm deep in the soil as easily as it can locate seeds on the surface (White and

Landis, 2004).

Giant foxtail is a common grassy weed in Pennsylvania farm fields and a potential food source for H. pensylvanicus (Brust and Tooley, 2002). Giant foxtail was first introduced in

North America in the 1920’s near New York City, NY (Warwick, 1990). In 1931, giant

74 foxtail was found in Philadelphia and as far west as Missouri by 1932 (Dekker, 2003).

After World War II, this species quickly spread across the North American corn and

cereal-growing regions (Slife, 1954). The life history traits important to the success of

weedy Setaria spp. include seed dormancy, long-lived seed pools, long duration in the

timing of seedling emergence and the induction of secondary (summer) dormancy

regulated by varying soil oxygen, water, and temperature signals (Dekker, 2003).

Giant foxtail takes approximately 45 to75 days to advance from flowering to seed production (Dekker, 2003). Once the first seeds reach maturity, the time of seed rain is continuous for several weeks up to months, with the seeds falling from the top of the panicle first and continuing down the tiller (Dekker et al., 1996). The amount of seed produced by giant foxtail is highly variable and ranges from 1 to 12,000 seeds plant-1

(Peters et al. 1963; Rominger 1962; Slife 1954; Steel et al. 1983; Vanden Born 1971).

Haar (1998) reported that giant foxtail seed number per panicle varied from 165 to 2,127

depending on geographic location, and panicle size. Attempts to provide more exact seed

productivity estimates (e.g., relationship between panicle length and seed number per unit

length) have produced variable results (Barbour and Forcella 1993; Defelice et al. 1989;

Fausey et al. 1997; Haar 1998).

Previous research suggests that specific species of carabid beetles have co-

evolved with major food sources, and that the lifecycle of H. pensylvanicus is

synchronized with the ripening of some summer annual grassy weeds (Brust 1994; Brust

and Tooley, 2002;). Brust (1994) reported that H. pensylvanicus adult activity density in

August to October was synchronized with large crabgrass (Digatara sanguinalis (L.)

Scop) and fall panicum (Panicum dichotoflorum Michx.) seed set in Indiana. Another

75 example of life system synchronization was documented by Sundell and Ylonen (2008)

who found that weasels, which are usually generalist predators, will specialize their diet

to correspond with rising populations of boreal moles in Finland. They concluded from

their research that this interspecific synchronization may be driven by the least amount of

exertion for prey at the net gain of the predator. The coevolution of predator and prey

certainly suggest that some food sources may be essential for predator survival. In

addition, the knowledge of specific relationships between predators and prey could

enhance our ability to better manage certain weeds using biocontrol tactics.

The overall objective for this study was to determine if H. pensylvanicus activity

density has coincided with giant foxtail seed rain in central Pennsylvania. To accomplish

this, an experiment examined H. pensylvanicus activity density and giant foxtail seed rain in 2005 and 2006. A component of this study examined giant foxtail seed rain as influenced by the date of weed emergence. This study could help assess the biocontrol potential for this seed predator and an important summer annual weed.

METHODS

A field study was conducted at Penn State University Russell E. Larson

Agricultural Research Center in Centre County (RELARC) at two locations approximately 0.4 km apart in 2005 and at a single location in 2006. Field site locations were historically in corn-soybean rotations. In 2005, one site utilized a naturally severe giant foxtail seed bank that was established prior to conducting this study, whereas the

76 second site relied on amending the seed bank by over-seeding the target weed. The experiment was repeated in the supplemented seed bank field location in 2006.

Each experiment location was designed as a randomized complete block with four replications. In the amended plots, two rows of giant foxtail were seeded about 0.5 m apart using an Earthway garden seeder1 in plots approximately 1 m by 1 m. Giant foxtail was seeded on four different dates about 10 days apart starting in mid May and ending in mid June. The seeding rate was intended to produce about 100 giant foxtail plants/m linear row. The giant foxtail seed was collected from a local population at the RELARC the previous year. The plot areas were tilled using rear-tine tiller2 just prior to seeding the giant foxtail at each establishment date. The study areas were irrigated twice, providing about 2 cm of water each time during the establishment period. At the natural seed bank location, individual plots were tilled as previously described, but not amended with additional weed seed. Approximately 2 weeks after the final establishment date,

0.28 kg ae/ha dicamba was applied postemergence to control broadleaved weeds.

Giant Foxtail Seed Rain. Seed rain was assessed by placing a collection tray consisting of an aluminum pan (45.7 cm long x 30.5 cm wide x 7.6 cm deep) covered with galvanized metal hardware cloth (1.25 cm by 0.5 cm openings) placed in the middle of each plot, similar to that used by Westerman et al. (2003). The pan was used to capture the giant foxtail seed, whereas the hardware cloth prevented organisms from entering and

1 Model 1001B, Earthway Products, 1009 Maple St., Bristol, IN 46507 2 Troy-Bilt Big Red Horse Tiller, MTD Products, Inc., P. O. Box 368022, Cleveland, OH 44136-9722

77 eating the seed. Every few days or weekly from August through early November, seeds were collected and counted using an electronic seed counter3 .

H. pensylvanicus Activity Density. Beetle activity density was monitored using pitfall traps at each location. Pitfall traps consisted of a 0.95 l plastic container (height

10.9 cm x diameter 11.4 cm) with a 0.24 l polystyrene cup filled about a third full with ethylene glycol (killing agent) placed in the bottom of the plastic container for specimen collection (see revised methods from chapter 3). The top half of a 2 l plastic soda bottle was placed in the opening of the plastic container to act as a funnel and direct invertebrate organisms into the ethylene glycol filled cup. Pitfall captures were based on a 72 hour period. When not in use, traps were covered with plastic lids to prevent the unnecessary capture of beetles and to keep debris and rain out of the traps. In 2005, one pitfall trap was placed in each of the foxtail plots (total number of pitfalls per sampling

=16) to determine beetle activity density. Pitfall trapping took place at three times (late

August and early and mid September). In 2006, a single pitfall trap per replication was utilized (pitfalls per sampling = 4) and pitfall trapping occurred weekly starting in early

August and ending in early November. Captured H. pensylvanicus beetle specimens were identified and enumerated to determine activity density.

RESULTS AND DISCUSSION

Giant Foxtail Seed Rain. In 2005, giant foxtail seed rain began in late August, occurring over a six to eight week period, and intensity varied over time. Total seed

3 Electronic Counter Model 8503, The Old Mill Company, Savage, MD 20863.

78 production over this period averaged 12,251 seeds/trap in the naturally established plots

and 18,127 seeds/trap in the amended plots. Peak giant foxtail seed rain occurred in the

beginning of October at both locations (P< 0.0001, P< 0.0001; Figure 4.1); the naturally

established plots averaged 5,000 seeds/trap with approximately 6,500 seeds/trap in the

amended plots during this weekly sampling period. The period of peak seed rain

comprised 36 to 40% of the total seed rain for the two locations. In 2006, seed rain also

began in late August and but did not peak until the October 19th collection date

(P<0.0001; Figure 4.2). Seed rain during this peak comprised 41% of the total weekly

seed rain. Seed rain quickly declined to less than 5% by November 1. Seed production

was less in 2006 than in 2005 with the total amount averaging about 2,900 seeds/trap.

Less successful giant foxtail stand establishment and differences in mid and late summer

rainfall (2006 had higher rainfall accumulation) may have contributed to the difference in

seed production between years.

Date of foxtail establishment did not influence timing of seed maturation

(P=0.2453; Figures 4.3 and 4.4) or numbers of seed produced (P=0.4393) at either

location in 2005. In 2006, seed production varied by planting date (P=0.0035) with the

largest amount of seed (average of 2056/trap) being produced in the last giant foxtail

establishment date on June 10 and the fewest in the first establishment date on May 10

(average of 1436/seed) and the second and third planting date was intermediate to the

first and fourth in terms of seed production. These differences likely occurred because of varying rainfall producing differences in subsequent panicle length or numbers at the establishment periods. Giant foxtail reproduction is controlled by photoperiod; flowering and anthesis timing is in response to the shortening of day length and panicle

79 development usually occurs more rapidly under shorter day photoperiods and with higher

air temperatures (Dekker, 2003). Apparently the 3 to 4 week delay in foxtail

establishment that occurred in this study was insufficient to delay flowering and seed set.

These results support the concept that later emerging seedlings will develop faster and

flower with earlier emerging cohorts.

H. pensylvanicus Activity Density. Average activity density across three sampling

dates and two locations in 2005 were 13 beetles/pitfall over a 72 h period. Activity

density was highest in the naturally established giant foxtail location at the first sampling

date in late August (P< 0. 0001; Figure 4.4) with a slight decline by mid September.

There was no difference in activity density at the three sampling dates in the amended

location with an average density of 12 beetles/pitfall. In 2005, activity density was

highest in June 9 foxtail establishment plot (forth seeding date) compared to some earlier

dates at the amended site (P<0.0126). However, no differences in activity density were

observed between establishment dates at the naturally established site and these data were

not collected in 2006. In 2006, sampling date did not affect beetle activity density which

was relatively constant in August and September with a small spike in late August and

fewer beetles being captured in October and none in November (Figure 4.5; P=0.0659).

H. pensylvanicus was most abundant in August of 2006, beetle activity density averaged

6.3 beetles/pitfall. Although this experiment was primarily intended to measure giant foxtail seed production and seed rain as influenced by establishment date, the beetle activity density data collected in the experiment and adjacent study suggest that beetle activity density was independent of giant foxtail seed rain.

80 H. pensylvanicus Coincidence with Giant Foxtail Seed Rain. In 2005, H. pensylvanicus activity density either slightly decreased or remained unchanged between late August and mid September. Although only three dates were sampled in 2005, other research in an adjacent location the same year showed a decline in H. pensylvanicus activity density between early August and mid September (Murray et al., 2007). In 2006,

H. pensylvanicus activity density in the foxtail plots was relatively constant in August and September with a spike in late August and gradual decline into late October. Brust

(1994) also found that carabid species in Indiana, and especially H. pensylvanicus, increase in numbers in late summer and early autumn coinciding with summer annual weed maturity. The importance of seeds within the carabid diet can be directly linked to beetle fecundity as Jorgensen and Toft (1997) showed that female H. rufipes laid significantly more eggs when fed mixed seed diets compared to an insect diet.

Giant foxtail seed rain first occurred in low numbers around mid August and peaked in early to mid October in both 2005 and 2006. Giant foxtail continued to shed seed through much of October. This suggests that H. pensylvanicus activity density may not be well synchronized with the annual production of giant foxtail in central

Pennsylvania (Figure 4.6). Brust’s (1994) observations of H. pensylvanicus lifecycle synchronization coinciding with the ripening of some summer annual grasses in Indiana did not include giant foxtail, but rather large crabgrass and fall panicum. These two summer annual grasses generally flower 2 to 4 weeks prior to giant foxtail (Doll, 2001) which would be better timed with peak H. pensylvanicus activity density.

These findings suggest that fresh giant foxtail seed may not be a key food source for the survival of H. pensylvanicus in the Northeast, especially in tilled cropping

81 systems. However, even though H. pensylvanicus is not well synchronized with giant

foxtail ripening, it could be a food source for the following year (Ward, 2008

unpublished). H. pensylvanicus may prefer and may have synchronized with other annual

weed lifecycles. Additional research should be focused on the phenology of weeds and

the timing of weed seed rain in relationship to this and other potential weed seed

predators. Prevention of weed infestations should not focus primarily on eliminating

seedlings but on managing the problem before it starts. This can be done by focusing

more on the timing of seed rain to prevent further addition to the weed seed bank.

LITERATURE CITED

Barbour, J. C. and F. Forcella. 1993. Predicting seed production by foxtails

(Setaria spp.). Proceedings of North Central Weed Science Society 48:100.

Brust, G. E. 1994. Carabids affect the ability of broadleaf weeds to compete. Agriculture,

Ecosystems, and Environment. 48:27-34.

Brust, G. E. and G. J. House. 1988. Weed seed destruction by arthropods and rodents in

low input soybean agroecosystems. American Journal of Alternative Agriculture.

3:19-25.

Brust, G. E. and J. Tooley. 2002. The agroecology of carabid beetles. Intercept Limited.

Andover, Hampshire, UK. 215-229.

Cromar, H.E., S.D. Murphy, and C. J. Swanton. 1999. Influence of tillage and crop

residue on postdispersal predation of weed seeds. Weed Science. 47:184-194.

Defelice, M. S., W. B. Brown, R. J. Aldrich, B. D. Sims, D. T. Judy, and D. R. Guethle.

82 1989. Weed control in soybeans (Glycine max) with reduced rates of

postemergence herbicides. Weed Science. 37:365–374.

Dekker, J. 2003. The foxtail (Setaria) species-group. Weed Science. 51:641-656.

Dekker, J., B. I. Dekker, H. Hilhorst, and C. Karssen. 1996. Weedy adaptation in Setaria

spp.: IV. changes in the germinative capacity of S. faberii embryos with

development from anthesis to after abscission. American Journal of Botany.

83:979–991.

Doll, J. D. 2001. Knowing when to look for what: weed emergence and flowering

sequence in Wisconsin. Wisconsim State University Extension IPM.

Fausey, J. C., J. J. Kells, S. M. Swinton, and K. A. Renner. 1997. Giant foxtail (Setaria

faberi) interference in non-irrigated corn (Zea mays). Weed Science. 45:256–260.

Haar, J. M. 1998. Characterization of foxtail (Setaria spp.) seed production and giant

foxtail (Setaria faberii) seed dormancy at abscission. Ph.D. dissertation. Iowa

State University, Ames, IA.

Jorgenson, H. B and S. Toft 1997. Food preference, diet dependent fecundity and larval

development in Harpalus rufipes (Coleoptera:Carabidae).Pedobiologia. 41: 307-

315.

Kirk, V. M. 1972. Seed-catching by larvae of two ground beetles, Harpalus

pensylvanicus and H. erraticus. Annuals of the Entomological Society of

America. 65: 1426-1428.

Landis, D. and S. White. 2004. Biocontrol agent profile Harpalus pensylvanicus.

Technical Bulletin. Michigan Sate University.

http://www.cips.msu.edu/biocontrol/fact_sheets/H.%20pen.htm

83 Larochelle A. and M. C.Lariviere. 2003. A natural history of the ground-beetles

(Coleoptera: Carabidae) of America north of Mexico. 583 pp.

Peters, R. A., J. A. Meade, and P. W. Santelmann. 1963. Life history studies as related

to weed control in the Northeast. 2. Yellow Foxtail and Giant Foxtail. Kingston,

RI: University of Rhode Island Agricultural Experiment Station, p. 18.

Rominger, J. M. 1962. Taxonomy of Setaria (Gramineae) in North America. Illinois

Biology of Monographs. 29.

Slife, F. W. 1954. A new Setaria species in Illinois. Proceedings of North Central Weed

Control Conference. 11:6–7.

Steel, M. G., P. B. Cavers, and S. M. Lee. 1983. The biology of Canadian weeds. 59.

Setaria glauca (L.) Beauv. and S. verticillata (L.) Beauv. Canadian Journal of

Plant Science. 63:711–725.

Sundell J. and H. Ylonen. 2008. Specialist predator in a multi-species prey

community:boreal voles and weasels. Integrative Zoology. 3:51-63.

Vanden Born, W. H. 1971. Green foxtail: seed dormancy, germination and growth.

Canadian Journal of Plant Science. 51:53–59.

Warwick, S. I. 1990. Allozyme and life history variation in five northwardly colonizing

North American weed species. Plant Syst. Evol. 169:41–54.

Westerman, P. R., J. S. Wes, M. J. Kroff and W. Van Der Werf. 2003. Annual losses of

weed seeds due to predation in organic cereal fields. J Ecol. 40:824-836.

Zhang, J., F. Drummond, M. Liebman, and A. Hartke. 1997. Insect predation of seeds

and plant population dynamics. Technical Bulletin 163. University of Maine.

http://www.umaine.edu/mafes/elec_pubs/techbulletins/tb163.pdf.

84

Figure 4-1

7000 d Naturally Established Amended 6000 y 5000 c E

4000

3000 b D D 2000 ef e

Number of Seeds/Tra C F f 1000 BC ab B A a 0 8/26 9/8 9/13 9/22 10/3 10/10 10/17 10/20 Sampling Date

Figure 4-1: Average giant foxtail (Setaria faberi Herrm.) seed rain over time in 2005.

Two sites were sampled including a naturally established giant foxtail plot and an amended giant foxtail location. Average seed rain within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P <

0.05).

85

Figure 4-2

1400 f 1200 y 1000 800 e 600 cd d 400 bc a a ab Number of seeds/tra 200 a a a ab 0

8 2 9 3 9 2 2 9/ 0/1 /1 8/10 8/17 8/25 8/31 9/15 9/ 9/ 10/6 1 10 11/9 Sampling Date

Figure 4-2: Average giant foxtail (Setaria faberi Herrm.) seed rain over time in the amended plot in 2006. Averaged seed rain within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05).

86

Figure 4-3 7000 5/12/2005 6000 5/26/2005

5000 6/7/2005 6/13/2005 4000

3000

2000

1000

0 8/26 9/8 9/13 9/22 10/3 10/10 10/17 10/20 Sample date

Number of 8000 5/12/2005 5/26/2005 7000 6/7/2005 6000 6/13/2005 5000 4000 3000 2000 1000 0 8/26 9/8 9/13 9/22 10/3 10/10 10/17 10/20 Sampling date

Figure 4-3: Figure 4-3a Average influence of disturbance date and sampling date on naturally established giant foxtail seed rain in 2005. The naturally established location had a severe giant foxtail seed bank that had established in a corn-soybean rotation prior

to this study. The disturbance date indicates the date at which the giant foxtail plots were

tilled to kill emerged plants and stimulate the foxtail germination. Figure 4-3b Average

influence of disturbance date and sampling date on amended giant foxtail seed rain in

2005. The amended location consisted of amending the seed bank by over-seeding the

target weed species. 87

Figure 4-4

) 18 a 16 14 a b

12

10 8 6 4

Activity Denisty (# beetles/pitfall 2 0 8/26 9/8 9/15 Date of Sampling

Figure 4-4: Average Harpalus pensylvanicus Herrm. activity density in naturally established giant foxtail in 2005. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P < 0.05).

88

Figure 4-5

40 a 35

30

25

20

15 a a 10 a a a

Beetle activty density (#beeltes/pitfall) density activty Beetle a a 5 a a a a a a 0

8 8 8 8 8 9 9 9 9 1 1 1 1 1 /3 /1 /1 /2 /3 /7 /1 /2 /2 0 0 0 0 1 1 7 4 1 4 1 9 /5 /1 /1 /2 /2 2 9 6 Sampling Date

Figure 4-5: Average Harpalus pensylvanicus Herrm. activity density in giant foxtail in

2006. Mean beetle activity density (#beetles/pitfall) within sampling dates with the same letter are not significantly different according to the Tukey test for mean separation (P <

0.0659).

89

Figure 4-6

H. Gian

pensylvanicus t foxtail seed

Figure 4-6: Hypothesized Harpalus pensylvanicus activity density synchronization with giant foxtail (Setaria faberii Herrm.) seed rain in central Pennsylvania. Although there is some overlap, these two organisms are not well synchronized.

Appendix A

Fact Sheets

      

Weed Seed Predators: Potential

Contributors to Weed Control By Meredith Murray, William Curran, and Dave Mortensen, Department of Crop & Soil Science, Mary Barbercheck, Department of Entomology, Pennsylvania State Weeds of Ontario University, Eric Gallandt, Department Weed Ecology, University of Maine

Competition between weeds and prevent crop yield loss, • Seed predation can be responsible for weeds and crops in the U.S. for but combining this tactic with up to 90% of seed loss in agroecosys- limited resources causes an other cultural, mechanical or overall yield loss of 12% annu- chemical management tactics tems ally and this da mage and loss could have a greater positive im- • Seed predators include rodents, ants, costs U.S. farmers at least $23 pact than any single tactic alone. crickets and ground beetles. billion annually (Pimentel et Numerous insects consume al., 2000). To control weeds in weed seeds. Some of the most • Ways to increase weed seed predation major crops, most farmers have common and most promising bio- ♦ relied mainly on herbicides. In control agents include rodents, Reduce pesticide use 2002, almost 194 billion acres ants, crickets and ground beetles. ♦ Reduce tillage frequency of cropland across the U.S. Weed seed predation rates will ♦ were treated with herbicides vary depending on predator popu- Delay tillage events (USDA, 2002) and U.S. farmers lations, the food supply (weed ♦ Incorporate cover crops spent almost $5 billion on seeds), and how the field is man- ♦ chemical weed control making aged. In an Iowa study, seed pre- Plant fencerows, filter strips and them the most widely used crop dation rates from May to Novem- refuge strips protection chemica ls (Kiely et ber ranged from 7 to 22% per day al 2004). A number of negative depending on crop type (Figure 1) consequences result fro m herbi- (Westerman et al., 2005). Higher cide use including ground and predation rates were observed in 25 surface water contamination small grains and alfalfa compared (Barbash and Resek 1996; Lar- to corn and soybean. The rate of 20 son et al. 1997) and the evolu- seed predation typically increases 15 tion of herbicide resistant as the crop canopy develops 10 weeds (Heap 2006). The nega- within a field. Spring planted corn tive impacts that herbicides can and soybean crops provide little 5 have on the environment neces- protection for seed predators early 0 sitate the need to seek alterna- in the growing season compared to Corn Soybean Small Grain Alfalf a tive weed management prac- small grains or established alfalfa. Crop Type tices that better promote inte- In another study, predation of gi- grated and ecologically-based ant foxtail seeds in wheat was in- Figure 1. Seed predation rates in five cropping pest management principles. creased by over seeding wheat systems. Westernman et al., 2005 Conservation biological with red clover in the spring control may imp rove our ability (Davis and Liebman, 2003). Seed predation>25% x 8 predation<25% to manage weeds using less predators likely seek habitats that -2 herbicide. Enhancing popula- provide adequate cover for their 6 tions of neutral enemies can protection as well as a plentiful ) help reduce pest populations. food source. 5 4 10 Creating habitats that are attrac- Identifying specific p redator 2 tive to pest suppressive organ- organisms and the tactics that pro-

Giant Foxtail Seed 0 isms at a time when weed popu- mote their conservation and use is Bank Density (#m lations are vulnerable could an exciting area of study with a 135791113151719 increase the effectiveness of growing body of knowledge. Time (years) biological control. The sole use Farming practices of biological control will not be Figure 2. Effect of seed predation on giant foxtail effective enough to suppress seed bank. Hartzler et al., 2006

PAGE 1 Weed Seed Predation Continued can be altered to increase populations Creating refuge stripes of peren- in the fall can cause certain seed preda- of weed suppressive organisms. As an nial grasses around the boundary of tor populations to flourish. Consider example, integrating a legume cover crops can create an ideal overwintering ways to increase or conserve potential crop after small grains in a farming site for beneficial ground beetles, fungi beneficial seed predators in your farm- rotation, may enhance predation by and nematodes. Increasing plant resi- ing operation. providing protection for seed predators. due and decreasing tillage, especially

Promising Weed Seed Predators

Rodents

Mice are opportunistic feeders, seeded broadleaf weed species consuming the easiest available high (Cardina et al., 1996) such as velvet- density food source in an area. As a leaf, giant ragweed, jimsonweed, and result, their primary food is seeds morningglory, then shift to smaller (Zhang et al., 1997). Mice can con- seeds after the bigger seeds have been Image 1. A mouse consuming seeds. sume up to 100% of available weed collected (Abramsky 1983). Mice are bbc.co.uk seeds in a 12 hour period (Image 1). also one of the few weed seed preda- High predation rates are the result of tors that feed consistently on hard Table 1. Efficiency of removal of buried rodents finding a high density food shelled seeds (Brust & House, 1988). seeds by rodents. Abramsky, 1983 source, filling their mouth with as Unfortunately, while mice can be pro- many seeds as possible, caching or lific weed seed predators, they can Depth Number % Re move d of Seeds storing the seed and making repeated also be problematic in some cropping (in) trips back to collect additional seeds systems, feeding on desirable crop Removed (Abramsky, 1983). seeds and/or disrupting irrigation 2.54 102 79% Rodents locate seeds by scent and equipment, plastic mulch, and other 7.62 104 80% tools and equipment used in the agri- can even find seeds buried under the 12.7 70 70% soil surface (Table 1) (Abramsky, cultural production. 1983). Rodents feed first on larger 38.1 47 47%

Ants

Ants are diurnal insects that abundant non-preferred seeds. In spend the day actively foraging contrast, non-preferred seeds are (Image 2) and feeding. Ants have taken much more readily if they been found to feed on small seeded occur with abundant preferred seeds weed species (Brust & House, (Zhang et al., 1997). Ants also tend 1988) such as redroot pigweed and to colonize in agricultural fie lds in common lambsquarters. They can high numbers; however, their activ- remove up to 43% of small weed ity is reduced by tillage and possi- seeds over a 20 day interval and 2 bly by higher levels of crop residue to 30% of annual ryegrass seeds or stubble (Jacob et al., 2006). As a within 24 hours in pastures (Jacob result, ants could be more signifi- et al., 2006). Feeding preference cant seed predators in row crops studies have shown that the amount after inter-row cultivation is com- of each seed type removed by ants plete. Image 2. Ant seed predation. Uni- was strongly influenced by the versity of Würzburg, Germany amount and kinds of other seeds in the immediate area. A strongly pre- ferred seed is removed less fre- quently if it occurs in the midst of

WEED SEED PREDATORS: POTENTIAL CONTRIBUTORS TO WEED P AGE 1 Crickets When in large numbers, crickets sume over 200 redroot pigweed seeds are considered an economic pest in per day (Carmona et al., 1999). new seedlings of no-till alfalfa and clover. However, they can also be important weed seed predators. Crickets are nocturnal omnivores (generalists) that consume dead and living insects, broadleaf plants, Image 3. A common method of measuring seed grasses and seeds. They emerge in predation involves lightly attaching seeds to early August and have peak activity sandpaper or a similar material and placing in the middle of September with populations decreasing in October 80 (Carmona et al., 1999). Field obser- 70 vations (Image 3) and laboratory 60 studies showed that they consume 50 40 common agricultural weed seeds period) 30 such as velvetleaf, common lamb- 20 squarters, redroot pigweed, large 10 crabgrass, common ragweed, and % Consum ed hr feeding (24 0 Foxtail Lambsquarters Velvetleaf Waterhemp giant foxtail. Crickets can cause Weed Species greater than 76% weed seed removal in 24 hours (Figure 3) and a single Figure 3. Weed seed predation by crickets. female northern field cricket can con- Carmona et al, 1999 Ground Beetles

y it 50 s ) vania, giant foxtail seed predation n l 45 e a D tf 40 by insects including ground beetles y i t /p 35 i s iv e t lt 30 ranged fro m 37% to 62% in a 14- c e A e 25 e b day period from June through early tl # 20 e ( e September in a sweet corn crop B 15 10 (Figure 5). A single ground beetle 5 can consume up to 11 seeds daily 0 June July August Septe mber and seed removal can be as high as Sampling date -2 -1 Image 4: Ground beetle, Harpalus 120 to 130 seeds ft day (Honek et pensylvanicus. Murray, 2006 al., 2003). Unlike rodents, ground beetles do not survive periods of Figure 4. Harpalus pensylvanicus activity over Ground beetles, also known as intense disturbance such as fall or time in Central PA. Murray, 2006 carabid beetles, are common through- spring plowing. Fortunately, many out North America in agroecosys- ground beetles are fairly mobile and tems. Harpalus pensylvanicus a com- they can abandon fields in autumn mon carabid found in Pennsylvania and overwinter in refuge strips, and a known seed predator (Image 4), fence rows and filter strips. They 70 overwinters as an adult and has the do not necessarily prefer one crop 60 greatest abundance from July through over another, but instead may pre- 50 September in Pennsylvania (Figure 4) fer different crop types throughout 40 (Murray and Curran, 2006). Adults the growing season. Decreasing or 30 20 consume plant tissues, pollen, fungi, eliminating soil disturbance espe- 10

insects and seed; they prefer smaller cially in the late summer when Predation Seed Percent 0 sized broadleaf and grass seeds (Best beetles are feeding, mating and 6/15/05 7/15/05 7/29/05 8/11/05 9/1/05 & Beegle, 1977). Ground beetles can reproducing can increase ground Sampling Date be responsible for up to 90% of weed beetle activity. Figure 5. Weed seed predation rates at seed predation in some agroecosys- tems. At the Penn State University Pennsylvania State University Rock Springs Research Farm in Central Pennsyl- Research Farm. Murray, 2006

Ways to Increase Weed Seed Predation

• Reduce pesticide use and especially insecti- • Delay tillage. Weed seed predation occurs mostly cides. Insecticides often on the soil surface. It is important kill not only target insects Beneficial insects require: to keep weed seeds on, or near, the but also beneficial in- •Food soil surface. Delaying or eliminat- sects. When you do need ing tillage in the fall leaves seeds to apply a pesticide, use •Water on the soil surface longer resulting lower toxicity/softer pes- •Overwinter habitats in higher predation rates by insects ticides that are not persis- and mammals. tent in the environment. •Shelter from adverse conditions Do not apply pesticides to fencerows, refuge strips Agricultural practices such as pesticide appli- • Incorporate cover crops into or border alleys that serve cations, cultivation, tillage and harvesting your cropping system. Cover crops reduce available habitats for beneficial in- as refuges for seed preda- help hold the soil in place and cre- tors. ate better habitats for seed preda- Suitable habitats can be found in: tors and other organisms. • Promote and maintain diverse Reduce tillage events and •Herbaceous strips use conservation tillage fencerows, filter strips, and refuge and no-till practices. •Fencerows habitats that allow overwintering sites and protection for ground These practices create •Hedgerows “preferable” habitats for beetles, rodents, crickets and •Woodlots weed seed predators by other seed predators. decreasing disturbance •Uncultivated field areas events and destruction of beneficial habitat. M ichigan State University Extension

      

Ground Beetles: Potential for Biological Control of Weeds By: Meredith Murray and William Curran, Department of Crop & Soil Science, Weeds of Ontario Pennsylvania State University Conservational Biological Control of Weeds

• De finition- Conservation biological control of Seed Predation in Sweet Corn Over Time weeds is achieved by managing a site in such a 70 way that favors naturally occurring weed sup- 60 50 pressive organisms. 40 • There are numerous weed seed predators, some 30 20 of the most common include birds, mice, ants, 10

crickets, insect larvae and ground beetles. Percent Seed Predation 0 6/15/05 7/15/05 7/29/05 8/11/05 9/1/05 • At Pennsylvania State University Rock Springs Sampling Date Research Farm weed seed predation rates were found to range from 37% to 62% over a 14 day Figure 1. Weed seed predation rates at Penn- period (Figure 1). sylvania State University Rock Springs Re- search Farm. Ward, 2006 Weed Seed Predation by Ground Beetles

• De finition– The act of a predator capturing prey. Seed predation refers to the removal of seeds Harpalus pensylvanicus Activity Density in Five from the seed bank. Cropping Systems • Ground beetles can be responsible for up to 90% 50 of weed seed predation in agroecosystems 40 (Image 1). 30 2004 20 2005

• A ground beetles, also known as a carabid bee- 10 (# beetles/pitfall) (#

tles, can consume up to 11 seeds daily and seed Density Avtivity Beetle 0 removal can be as high as 120-130 seeds ft-2day-1 June July August September October (Honek et al., 2003). Sampling Date • Harpalus pensylvanicus has peak activit y in the Figure 2. Harpalus pensylvanicus acti vity over beginning of August in Central PA (Figure 2) time . Ward, 2006 (Murray and Curran, 2006), when they are ac- tively feeding, mating, and laying eggs (White & Landis, 2004). • Amara aenea is a weed seed predator that com- monly found throughout Pennsylvania. A. aenea adults mate from May to June and emerge from pupae between July and September. Adults over- winter and emerge in spring, being most abun- dant during April and May (Landis & White, 2004). • The key to using carabid beetles as a biocontrol agent of weeds is to understand their preference Image 1. Harpalus pensylvanicus con- for certain weed seeds and the factors influencing suming weed seed. Ward, 2006 population abundance.

PAGE 1

Common Ground Beetle Weed Seed Predators

Amara aenea Harpalus pensylvanicus • 5/16-3/8 inch long. • 1/2- 5/8 inch long, • Narrow with paral- nonmetallic body. le l sides. • Black to blackish • Rather flat with brown in color. very fine lines on • All appendages are wingcase. light brown in color. • Copperish to black in color. Weed Seed Predator Monitoring If you would like to gauge how common some weed seed predators are on your farm or in a particular field or plot area, construct and place a pitfall trap in the area of interest. This device will collect insects that crawl along the soil surface that could be potential weed seed predators. Warning: this device will also collect spi- ders, caterpillars, and other insects that are not involved with weeds. The following instructions explain how to build a pitfall trap.

Materials: • Deli container with lid (or bottom of 2-liter bottle)  Styrofoam cup with lid • Funnel created by cutting a 2-liter bottle  Antifreeze Methods: • Place deli container in the field so that the lip of the container is flush with soil surface. • Fill sytrofoam cup with 1 inch of antifreeze (killing agent). • Leave traps open in the field for 72 hours. • Count the number of Harpalus pensylvanicus, Amara aenea, crickets and ants captured in the field, this number will give you a general idea of how abundant seed predators are at that location.

What to Expect:

• Seed predator numbers will be greater in areas with dense canopy cover, areas surrounded by greenways and with less soil disturbance such as tillage. • The duration of sampling will influence how many insects you collect. In our trials, we collected as few as 1 or 2 beetles per pitfall to over 100 in a 3 day period. • You may want to trap in several areas to determine how different habitats influence predator numbers. • Sampling in the spring will result in higher numbers of Amara aenea while late summer samplings will have higher numbers of Harpalus pensylvanicus.

G ROUND BEETLES: POT ENT IAL FOR BIOLOGICAL CONTROL OF WEEDS P AGE 1 Weed Seed Predation Monitoring Pitfall trapping will help determine what potential insect weed seed predators you have in your area. Weed seed cards will help you determine whether these insects are in fact consuming your weed seeds. The follow- ing instructions explain how to construct and use weed seed cards. Materials: • 1/2 inch hardware mesh cloth  100 mm Petri dishes  Giant foxtail weed seeds • 150 mm Petri dishes  Soil  Nuts and bolts • Indoor/outdoor carpenter tape Methods: 1. Seed Card: Inverted 100mm Petri dish covered with indoor/outdoor carpenter trap. The tape is covered with weed seeds and soil. Randomly scatter a known number of seeds (about 100) on the tape. Lightly cover the tape with soil or sand to eliminate exposed adhesive (see photo 1). 2. Closed Trap: 1/2 inch hardware mesh cloth molded into cylinder cone (approximately 9 inches tall), capped with 150 mm Petri dish. The closed trap specifically tests for insect seed predators and blocks out birds and mammals (see photo 2). 3. Open Trap: Three holes drilled in a 150 mm Petri dish, suspended by three bolts (about 9 inches long) with nuts. The open trap allows access by both insects and mammals such as mice (see photo 3). 4. Exclosure Trap: 1/2 inch hardware mesh cloth and window screen molded into cylinder cone (approximately 9 inches high), capped with 150 mm Petri dish. (This trap keeps all organisms out and is not necessary unless you want to include a control treatment)

A seed card is placed in the middle of each trap so that the seed card is level with the soil level and looks like part of the “natural” soil surface. The traps are left out for 14 days, however; monitoring of the traps should take place every two days to determine how fast seeds are being removed in the desired area. After the 14 day duration is complete count the number of seeds remaining on the seed card and subtract from the starting num- ber of seeds. This will help determine which seeds are being consumed by seed predators and how many seeds are being removed from the desired area.

1. 2. 3.

What to Expect:

• Areas with high activity density (#beetles/pitfall) will have higher numbers of small seed removal (i.e. grasses, common lambsquaters and pigweed) compared to larger seeded weed species such as velvetleaf or ragweed. • Higher amount of seed removal during late summer when more weed seed predators are present.

Ways to Increase Weed Seed Predation

• Reduce pesticide use and especially insecticides. Insecticides often kill not only target insects but also beneficial insects. When you do need to apply a pesticide, use lower toxicity/softer pesticides that are not persistent in the environment. Do not apply pesticides to fencerows, refuge strips or border alleys that serve as refuges for seed predators. • Reduce tillage events and use conservation tillage and no-till practices. These practices create “preferable” habitats for weed seed predators by decreasing disturbance events and destruction of benefi- cial habitat. • Delay tillage. Weed seed predation occurs mostly on the soil surface. It is important to keep weed seeds on, or near, the soil surface. Delaying or eliminating tillage in the fall leaves seeds on the soil surface longer resulting in higher predation rates by insects and mammals. • Incorporate cover crops into your cropping system. Cover crops decrease tillage and herbicide use and create better habitats for seed predators. • Promote and maintain diverse fencerows, filter strips, and refuge habitats that allow overwintering sites and protection for ground beetles, rodents, crickets and other seed predators.

Appendix B

Lesson Plans

Meredith Ward

Weed Identification

Description: Learn the basic biology and how to identify ten weeds common to the northeastern United States

Subject: Weed Identification

Duration: 2 periods 40 minutes in length

Grade Level: Introduction to Advance 12th grade classes

Pennsylvania 4.4.12. A.1 Academic Define the component of an agriculture system that would result in a Standards/Pennsylvania minimal waste of resources. Ag Education: Objective: 1. Students will be able to identify ten weeds commonly found throughout the northeastern United States 2. Students will describe the basic biology of the ten identified weeds. Focus: Write on the chalkboard, “what is a weed?” How is a weed different then a flower in a flower garden and a corn plant in a corn field. Ask students to share their first impressions that come to their minds. Discuss their responses. Have students explain why weeds can be viewed as a threat in agricultural and nonagricultural systems (i.e. natural areas, lawns, gardens, parks). Mention the loss of crop yield, quality, and economic effects. Teach: 1. Ask students why is it important to properly identify weeds? Discuss their responses. Are their reasons mostly related to weed control? Discuss how there are specific control methods for specific weeds and how misidentification of weeds can result in no weed control. Do they mention economic effect? Applying the wrong herbicide to control weeds can result in the application of additional herbicides. 2. Show students the two example weed species. Point out the unique characteristics of the weeds and explain their basic biology. 3. Have students examine each of the ten weeds, making drawings of each and filling in the Weed Identification Activity Sheet provided. 4. At the end of the time period get students to think about what they have just learned. Ask them: o Which weed do they think would be the easiest to control with tillage? (broadleaves, leaves catch on tines of tillage equipment, easier to hand weed. etc.) o Which weeds are spread via seeds and underground tubers? (Yellow nutsedge) o What weeds are most likely to “hitch a ride” on animals and humans and why? (Burdock because of the burs) Assessment: 1. Have students group weeds based on plant lifecycles, reproduction, and plant type (study sheets for exam).

2. Administer a Weed Identification exam. Set up ten stations throughout the room. Each station should consist of one of the ten weeds and one or two questions on the Weed Identification Activity Sheet that correspond to that specific weed.

Resources: Uva. R. H., J. C. Neal and J. M. DiTomoaso. 1997. Weeds of the Northeast. Cornell University Press. Ithaca. NY.

Liebman, M., C. L. Mohler. And C. P. Staver. 2001. Ecological Management of Agricultural Weeds. Cambridge University Press.

Westbrook, F. E. and L. C. Gibbs. 1978. Common Weed Seedlings of the United States and Canada. University of Georgia Press.

Davis, A., K. Renner, C. Sprague, L Dyer, and D. Mutch. 2005. Integrated Weed Management “One Year’s Seedling..”. Michigan State University Extension Bulletin E-2931.

Hartwig, N. L. 1996. Introduction to Weeds and Herbicides. Pennsylvania State University Extension Circular 365.

Lanin, T. and M. Wertz. 2001. Weed Management. Pennsylvania State University. http://weeds.cas.psu.edu/psuweedfactsheets.html.

Teacher Sheet

Name:

Date:

Class:

Weed Identification Lab

Weeds are defined as any plant growing where it is not wanted. Virtually any plant depending on location is considered a weed. Weeds are naturally strong competitors. There are approximately 8,000 species of plants that behave as weeds; of those only 200 to 250 are major problems in cropping systems world wide. Weeds must have a unique biology in order for them to become a major pest. Most agricultural weeds have been found to possess one or more of the following characteristics: abundant seed production, presence of vegetative reproductive structures, and the ability to become spread easily and in large numbers. Weed management in cropping systems is a constant challenge faced by farmers. Competition between weeds and crops for limited resources causes an overall yield loss of about 12% annually. In the United States this yield loss costs producers over $15 billion. Proper weed identification is essential for maximum weed control. There are many different forms of weed control from chemical herbicides to tillage and the use of invertebrate weed seed predators. No matter which weed control tactic you choose, correct weed identification is needed. Some chemical herbicides that kill grassy weeds, for example, Poast (sethoxydim) will not control yellow nutsedge which is often mistaken for a grass.

Materials: • 36 pots (three plants per weed species) • Potting soil • Water • Print out fact sheets for 10 weed species • Seeds from each of the 10 weed species o Yellow nutsedge o Giant foxtail o Large crabgrass o Velvetleaf o Common lambsquarters o Common cocklebur o Common ragweed o Wild carrot o Common burdock o Sheperd’s-purse

Procedure: 1. Show the students the two example weed species provided. Point out unique characteristics of each that students should use to identify these weeds. 2. One station will be setup for each of the weed species. Students will be able to examine vegetative, reproduction and flowering plant parts. Each station will also have a basic biology information card that will assist the students in answering the Weed Identification Activity Sheet. 3. Students will move from station to station examining each weed species identifying unique characteristics and making drawings of important characteristics.

Analyzing Results: Students should complete the Weed Identification Activity Sheet

Student Sheet

Name:

Date:

Class:

Weed Identification Lab

Weeds are defined as any plant growing where it is not wanted. Virtually any plant depending on location is considered a weed. Weeds are naturally strong competitors. There are approximately 8,000 species of plants that behave as weeds; of those only 200 to 250 are major problems in cropping systems world wide. Weeds must have a unique biology in order for them to become a major pest. Most agricultural weeds have been found to possess one or more of the following characteristics: abundant seed production, presence of vegetative reproductive structures, and the ability to become spread easily and in large numbers. Weed management in cropping systems is a constant challenge faced by farmers. Competition between weeds and crops for limited resources causes an overall yield loss of about 12% annually. In the United States this yield loss costs producers over $15 billion. Proper weed identification is essential for maximum weed control. There are many different forms of weed control from chemical herbicides to tillage and the use of invertebrate weed seed predators. No matter which weed control tactic you choose, correct weed identification is needed. Some chemical herbicides that kill grassy weeds, for example, Poast (sethoxydim) will not control yellow nutsedge which is often mistaken for a grass.

Materials: • Fact sheets for 10 weed species • Seeds from each of the 10 weed species o Yellow nutsedge o Giant foxtail o Large crabgrass o Velvetleaf o Common lambsquarters o Common cocklebur o Common ragweed o Wild carrot o Common burdock o Sheperd’s-purse

Procedure: 1. One station is setup for each of the 10 weed species being observed. Examine vegetative, reproduction and flowering organs for each weed. 2. Make individual drawings of each weed, noting unique characteristics that will aid in proper identification later on. 3. Each station also has a basic biology information card, use this information card to fill in the appropriate boxes on the Weed Identification Activity Sheet.

Analyzing Results: Students should complete the Weed Identification Activity Sheet

Weed Identification Activity Sheet

Name:

Date:

Class:

Common Scientific Unique Origin Life Cycle Plant Method of Spread Problems to Name Name Characteristics (annual, Type Reproduction (vehicles, Agriculture biennial or (broadleaf (seed, animal, wind, perennial) or grass) underground water, birds, roots or shoots, foot traffic, or both) manure, other) 1. Cyperus • Triangular Native to Perennial Grass-like • Produces • Rhizom Rice, Yellow esculentus shaped stem North up to 90,000 e growth peanuts, Nutsedge • V-shaped America seeds from tubers corn, cotton, (Pull one plant leaf blades and • Undergrou • Seeds: soybeans, up so students • Shiny or Eurasia nd tubers: spread by and potatoes can see nutlets) waxy upper nutlets (most animals, foot surface important) traffic, • Yellow vehicles spiky flower

2. Setaria faberii • Ligule is China Summer Grass Seeds Animals, water, Row crops: Giant Foxtail fringed Annual foot traffic, soybean, (Point out • Upper vehicles corn, fringed ligule surface covered and small hairs with fine hairs on upper • Large, fuzzy surface of arching panicle blades) (Spike)

Common Scientific Unique Origin Life Cycle Plant Method of Spread Problems to Name Name Characteristics (annual, Type Reproduction (vehicles, Agriculture biennial or (broadleaf (seed, animal, wind, perennial) or grass) underground water, birds, roots or shoots, foot traffic, or both) manure, other) 3. Digitaria • Leaf sheath Native to Summer Grass Tillers and seeds Animals, water, Sorghum and Large sanguinalis tinted purple and Southern Annual (up to 150,000) foot traffic, soybean Crabgrass covered with Europe vehicles (Point out large long hairs number of • Ligule hairs) membranous • Flowers look like segmented fingers from a hand 4. Abutilon • Heart India Summer Broadleaf Seeds (up to Animals, Row crops Velvetleaf theophrasti shaped velvety Annual 8,000) water, manure especially (Point out heart leaves corn and shaped leaves • Yellow soybeans and taproot) flower • Heart shaped seeds • Taproot 5. Chenopodium • Leaves have Native to Summer Broadleaf Seeds Drops off Row crops Common album white coating North Annual parent plant lambsquarters • Triangular America and form (Point out shaped leaves however patches around white coating • Older leaves may also parent plant, on leaves) have toothed be from dispersal by margins Europe or manure and • Flower Asia harvesting clusters located equipment towards end of stems/branches

Common Scientific Unique Origin Life Cycle Plant Method of Spread Problems to Name Name Characteristics (annual, Type Reproduction (vehicles, animal, Agriculture biennial or (broadleaf or (seed, underground wind, water, birds, perennial) grass) roots or shoots, both) foot traffic, or manure, other) 6. Xanthium • Leaves are Germany Summer Broadleaf Seeds Wind, water, Competitive Common pensylvanicu alternate and Annual animals, birds, with Cocklebur m toothed with 3-5 foot traffic, soybeans and (Point out lobes vehicles other field sharply toothed • Upper crops leaves and surface of rosette) leaves is darker then lower surfaces • Visible midvien • Stiff short hair on leaves • Seed pods are burs 7. Ambrosia • Deep clefts Native to Summer Broadleaf Seeds (up to Water, birds, Row crops Common artemisiifolia in margin, North Annual 62,000) animals, foot and cereals Ragweed forming lobes America traffic (Point out that are deeply lobed rounded and leaves with slightly pointed rounded edges) at the ends • Dense hairs on lower surfaces of leaves

Common Scientific Unique Origin Life Cycle Plant Method of Spread Problems to Name Name Characteristics (annual, Type Reproduction (vehicles, Agriculture biennial or (broadleaf (seed, animal, wind, perennial) or grass) underground water, birds, roots or shoots, foot traffic, or both) manure, other) 9. Common Arctium minus • Leaves: Europe Biennial Broadleaf Seeds (up to Animals, birds, Not a crop Burdock rosette, 15,000) human traffic weed ( point out triangular or problem in fleshy leaves oval, fleshy, tilled field with stiff hairs coarsely veined, growth, on top and notched at tip mostly found whitish wooly • Stiff hairs in old fields hairs on on surface and along borders bottom) lower surfaces and fence are whitish and rows. woolly No till • 2nd year: problem. flower (burs) stalk emerges from center of rosette 10. Sheperd’s Capsella bura- • Rosette Europe Winter Broadleaf Seeds (up to Wind Problem of Purse pastoris • Leaf Annual or 50,000) vegetable (point out margins are Biennial crops and deeply toothed or wavy winter indented • Older cereals can leaves) leaves are be alternative indented more host for than halfway to diseases and the midvein can harbor • 2nd year: viruses tiny white flowers, seed pods heart- shaped

Weed Identification Activity Sheet

Name:

Date:

Calss:

Common Name Scientific Unique Characteristics Life Cycle Plant Type Method of Problems to Name (annual, (broadleaf or Reproduction Agriculture biennial or grass) (seed, perennial) underground roots or shoots, both) 1.

2.

3.

Common Name Scientific Unique Characteristics Life Cycle Plant Type Method of Problems to Name (annual, (broadleaf or Reproduction Agriculture biennial or grass) (seed, perennial) underground roots or shoots, both) 4.

5.

6.

7.

Common Name Scientific Unique Characteristics Life Cycle Plant Type Method of Problems to Name (annual, (broadleaf or Reproduction Agriculture biennial or grass) (seed, perennial) underground roots or shoots, both) 8.

9.

10.

Mechanical Weed Control: White Thread Stage

Description: Learn about early weed control’s affect on weed vitality

Subject: Mechanical Weed Management

Duration: o Plant weed seeds - 30 minutes o Rake trays (after 5 days, 10 days and 15 days) - 10 minutes o Take observations 30 days after planting - 20 minutes Grade Level: Introduction to Advance 12th grade classes

Pennsylvania 4.4.12. A.1 Academic Define the component of an agriculture system that would result in a Standards/Penns minimal waste of resources. ylvania Ag Education:

Objectives: 1. Identify plant life stages. 2. Discuss mechanical weed control practices and susceptibility of seedlings to control. 3. Discuss timing of mechanical weed control. Focus Write on the chalk board timing of weed control. Ask students why they think timing of control practices is important to weed control. Discuss their answers. Do they mention the amount of plant death, repeated control practices, economics. Teach: 1. Discuss the plant life stages. Mention the white thread stag of seedlings. 2. Discuss tillage as a form of weed control. Ask students to make a list of benefits and downfalls of tillage. 3. Have students observe the effect of tillage timing on plant vitality. 4. At the end of the lesson ask students the following questions: o When is it most beneficial time to till a field? When the weeds are in the white tread stage (the earliest time possible).

o Why are the plants vulnerable at this stage? Used most of their stored energy (seed) to germinate and begin growth.cDamage to plant structures has a higher mortality rate.

o If tillage after 20 days did not have a high mortality rate what can you do to manage the remaining weeds? Till again or use herbicides. Assessment: 1. Students will be evaluated based their completion of the weed control lab exercise.

2. Observations from this lab will be incorporated into a lab exam. Resources: Davis, A., K. Renner, C. Sprague, L Dyer, and D. Mutch. 2005. Integrated Weed Management “One Year’s Seedling..”. Michigan State University Extension Bulletin E-2931.

Liebman, M., C. L. Mohler. And C. P. Staver. 2001. Ecological Management of Agricultural Weeds. Cambridge University Press.

Taiz, L. and E. Zeiger. 2002. Plant physiology. Sinauer Associates, Inc. Sunderland. MA.

Teacher Sheet Mechanical Weed Control: White Thread Stage

Name:

Date:

Class:

Properly timed tillage events can drastically suppress weed populations. Timing of tillage is based primarily on the weeds life cycle. The most effective time to control annual weeds is when they are at their most fragile life stage, the white thread stage. During this time, tillage can prevent weed problems by weakening the plants after they have invested a large amount of energy in germination, stem elongation, and leaf and root production. A plant's life cycle describes how long a plant lives. There are four main life cycle stages: seed, growth, reproduction and death. Most plants start out as a seed. That seed develops using energy stored in the seed to germinate, push through the soil, and develop stems, roots and leaves becoming a young seedling. The seedling continues growing; however, instead of using the stored energy in the seed, the young plant can now perform photosynthesis to make food using the sun’s energy. As the plant matures it starts putting more energy into reproduction, either by creating flowers and seeds or vegetative structures (i.e. stolons, rhizomes and tubers). Once these plants have produced seed and completed reproduction annual plants die and the white thread stage will begin again with the germination and development of the next generation. Possible Annual Broadleaf Seeds Materials: Weeds • Seeds Common Lambsquarters, Redroot Pigweed, Velvetleaf • Soil Crops • Buckwheat, Soybean, Canola Planting Trays Possible Annual Grass Seeds • Water Weeds • Hand Rake Foxtail, Large Crabgrass, Witch grass, Fall Panicum Crops Japanese Millet, Rye, Wheat, Barley

Procedure: 1. Separate students into 4 groups. 2. Choose one annual broadleaf seed and one annual grass seed. Assign each group a seed type, two groups using the same broadleaf seed and the other two groups using the same grass seed. 3. Direct the students to plant 100 seeds randomly throughout the planting trays. 4. Water planting trays when needed.

5. Five days after planting direct the students to select one planting tray and use the hand rake to rake the entire planting tray once. Show the students the small developing white thread plants that are exposed with the tillage (photo 1). 6. Have students make observations 30 days later, counting the number of thriving plants. Record results on lab exercise. 7. Direct students to repeat numbers five and six. The second plating tray should be raked ten days after planting and observed 30 days later. The third planting tray should be raked 20 days after planting and observed 30 days later. 8. Have students write their results on the chalk board and compare results with the rest of the class. From the results students will determine when the highest mortality occurred for each plant species thereby, determining the best time to control each weed species.

Photo 1. White thread stage. Results: Data Table1: Annual Broadleaf Tillage Time After Group #1 Group # 2 Average Number Planting Number of Thriving Number of Thriving of Thriving Plants Plants Plants

5 days

10 days

20 days

Date Table 2: Annual Grasses Tillage Time After Group #3 Group # 4 Average Number of Planting Number of Thriving Number of Thriving Thriving Plants Plants Plants

5 days

10 days

20 days

1. What tillage timing had the largest amount of broadleaf mortality and why? Based on student results. – small seeded broadleaves will be impacted the most (pigweed vs. ragweed). Broadleaf weeds or dicots are generally more susceptible to this mechanical control because the growing point or meristem is above ground at the apex of the plant.

2. What tillage timing had the largest amount of grass mortality and why? Based on student results. – grasses are less effectively controlled at this time. The development of grasses allows the growing point to remain below the soil surface helping them to survive.

3. When would be the best time for a farmer to till his fields if he/she had a combination of all weed types and why? “Blind” weeding is used within the first two or three weeks of planting crops and is generally first used within the first week after planting. Farmers should till as soon as the white thread stage of weeds can be seen and repeated until the weeds are visible above the surface of the soil. Weed seedlings in the white thread stage have used up a majority of energy reserves and have a higher mortality rate when injured.

4. If a tillage event had a low weed mortality rate due to poor timing of tillage what could the farmer do to reduce the weed infestation? The farmer could use a different type of tillage such as row cultivation or consider using a rescue herbicide if allowed in their farming operation.

Student Sheet

Mechanical Weed Control: White Thread Stage

Name:

Date:

Class:

Properly timed tillage events can drastically suppress weed populations. Timing of tillage is based primarily on the weeds life cycle. The most effective time to control annual weeds is when they are at their most fragile life stage, the white thread stage. During this time tillage can prevent weed problems by weakening the plants after they have invested a large amount of energy into germination, stem elongation, and leaf and root production. A plant's life cycle describes how long a plant lives. There are four main life cycle stages: seed, growth, reproduction and death. Most plants start out as a seed. That seed develops using energy stored in the seed to germinate, push through the soil, and develop stems, roots and leaves becoming a young seedling. The seedling continues growing; however, instead of using the stored energy in the seed, the young plant can now perform photosynthesis to make food using the sun’s energy. As the plant matures it starts putting more energy into reproduction, either by creating flowers and seeds or vegetative structures (i.e. stolons, rhizomes and tubers). Once these plants have produced seed and completed reproduction, annual plants die and the white thread stage will begin again with the germination and development of the next generation.

Materials: • Seeds • Soil • Planting Trays • Water • Hand Rake

Procedure: 1. Plant all of the assigned weed seeds in the planting tray. 2. Water planting trays as necessary. 3. Five days after planting select a planting tray and hand rake the entire planting tray once or twice. 4. Repeat number 3 at 10 and 15 days after seeding in separate trays. The second plating tray should be raked ten days after planting – again observe and the impact on the developing seedlings and the third planting tray should be raked 15 days after planting and again evaluated for the impact on the seedling weeds. On all trays, make observations 30 days after seeding by counting the number of surviving thriving plants. Record results on lab exercise. 5. Write your group’s results on the chalk board and compare your results with those of your class. Determine when the highest mortality rate occurred for each plant species thereby, determining the best time to control each annual weed species.

Photo 1. White thread stage. Results: Data Table1: Annual Broadleaf Tillage Time After Group #1 Group # 2 Average Number Planting Number of Thriving Number of Thriving of Thriving Plants Plants Plants

5 days

10 days

15 days

Date Table 2: Annual Grasses Tillage Time After Group #3 Group # 4 Average Number of Planting Number of Thriving Number of Thriving Thriving Plants Plants Plants

5 days

10 days

15 days

1. What tillage timing had the largest amount of broadleaf mortality and why?

2. What tillage timing had the largest amount of grass mortality and why?

3. When would be the best time for a farmer to mow his fields if he/she had a combination of all weed types and why?

4. If a tillage event had a low weed mortality rate due to poor timing of tillage what could the farmer do to reduce the weed infestation?

Weed Seed Predation

Description: Investigate weed seed predation by ground beetles

Subject: Biological Weed Management

Duration: Setup - 40 min Analysis – 60 min

Grade Level: Introduction to Advanced 10th grade classes.

Pennsylvania Academic 4.5.10.B. 2 Standards/Pennsylvania Analyze health benefits and risks associated with integrated pest Ag Education: management. • Asses levels of control within different integrated pest management practices for weed control

Objectives: 1. Students will describe alternative weed management methods. 2. Students will identify four common invertebrate weeds seed predators. AND/OR 3. Students will approximate weed seed predation by invertebrates and invertebrates+vertebrates. Focus: Write the words agricultural weed management on the chalkboard. Discuss different forms of weed control used in agriculture. The most common answers are chemical herbicides and mechanical management such as tillage and mowing. Now write Integrated Pest Management on the board. Have student think and talk about different ways that insects may help control weeds in agricultural ecosystems. Discuss their answers and the feasibility of those answers. Teach: 1. Discuss the many different ways of reducing weed infestations. List the positives and negatives of each. 2. Have students discuss and list the benefits and downfalls of using insects as a weed management method. 3. Discuss agricultural events and cropping systems that may increase a biocontrol agent’s activity density. 4. Identify five common weed seed predators (rodents, birds, ants, crickets, and beetles). 5. Each group will average the activity density for their three pitfall traps according to seed predator and write it on the chalk board AND/OR Each group will find the average the % seed predation for each trap type and write it on the chalk board.

6. At the end of the period ask students the following questions: o What could a farmer do to increase invertebrate seed predator populations? (Reduce tillage, use cover crops, decrease pesticide use) o Why are organic and no-till farmers interested in weed seed predation? (organic can not use herbicides and no-till farmers use a lot of cover crops in their farming rotations) Assessment: 1. Have students complete the Weed Seed Predation Activity Sheet. 2a. Have students determine the impact that invertebrates and vertebrates have on seed predation in the locations sampled. AND/OR 2b. Have students determine the amount of predation occurring in the landscape. 3. Each group will pick one of the five weed seed predators discussed and do a presentation on that organism for the class. Resources: Liebman, M., C. L. Mohler. And C. P. Staver. 2001. Ecological Management of Agricultural Weeds. Cambridge University Press.

White, S. and D. Landis. 2004. Biocontrol Agent Profile Harpalus pensylvanicus. Technical Bulletin. Michigan Sate University. http://www.cips.msu.edu/biocontrol/fact_sheets/H.%20pen.htm

Zhang, J., F. Drummond, M. Liebman, and A. Hartke. 1997. Insect predation of seeds and plant population dynamics. Technical Bulletin 163. University of Maine. http://www.umaine.edu/mafes/elec_pubs/techbulletins/tb163.p df

Tooley, J and G. E. Brust. 2002. The agroecology of carabid beetles. Intercrpt Limited. Andover, Hampshire, UK. 215-229.

Mortenson, D. A., L. Bastiaans and M. Sattin. 1999. The role of ecology in the development of weed management systems: an outlook. Weed Research. 40:49-62.

Teacher Sheet Weed Seed Predator Identification and Activity Density

Name:

Date:

Class:

Introduction:

Weed seed predation is the removal of potentially viable seeds from the weed seedbank. Numerous organisms have been found to be weed seed predators. Some of the most common are rodents, birds, ants, crickets, and carabid beetles. The amount of seed predation occurring in a cropping system is directly related to the activity density of that seed predator. For the purpose of this demonstration activity density is the number of seed predators per pitfall; this number represents the number of seed predators that are searching for food on the soil surface. In this experiment you will be determining the activity density of two ground beetles Harpalus pensylvanicus and Amara aenea. This information will approximate the amount of seed predation that may be occurring by Harpalus pensylvanicus and Amara aenea. Activity density can vary greatly between cropping systems. Annual cropping systems have unique characteristics because of the high frequency and intensity of both physical and chemical disturbances required to maintain them. Intensive management of agricultural landscapes such as frequent cultivation and pesticide application can negatively affect abundance, diversity and efficiency of ground beetles and other weed seed predators. Multiple studies have shown that seed predators have a higher activity density in cropping systems that require fewer soil disturbances and produce a denser canopy cover. Many seed predators live in the soil, frequent soil disturbances can cause habitat deterioration, emigrantion to more preferable habitats and in some cases can cause death. Cropping systems with dense canopy covers provide these biocontrol agents with much needed protection from other predators.

Materials Needed Per Group: • 6- 2L soda bottles cut into two so that top forms a funnel • 6- 24 oz deli containers or 6 2L soda bottle bottoms • 6- 8 oz Styrofoam cups with lids • Antifreeze • 6 aluminum pie pans • Tweezers • Seed Predator Identification Guide

Procedure: • Assign students to five groups. • Pitfalls are made by cutting a 2-liter bottle in half. A Styrofoam cup is placed in the bottom of the 2-liter bottle and the nozzle is inverted to form a funnel. • 3 pitfalls are randomly placed throughout a field for each of the five groups.

• Pitfalls are placed/buried in the field so that the lip of the 2-liter bottle is flush with the soil surface (See Photo 3). • Once the pitfall traps are placed, pour one inch of antifreeze into the Styrofoam cup, this acts as the killing agent. • Pitfalls are left out/open in the field for 72 hours. • Samples can then be taken into the classroom and the contents of the cups are poured into aluminum pie pans and the common invertebrate seed predators are identified and counted (disregard other insects). • Common invertebrate seed predators consist of Harpalus pensylvanicus, Amara aenea, ants and crickets.

Photo 1: Materials for pitfalls Photo 2: Pitfall Photo 3: Pitfall placement in field

Analyzing Data: Data Table: Your Group’s Results Pitfall Number Ants Crickets Harpalus Amara aenea pensylvanicus 1

2

3

Data Table: Averaged Results for Class Group Ants Crickets Harpalus Amara aenea Number pensylvanicus 1

2

3

4

5

1. Which seed predator had the highest activity density and which seed predator had the lowest activity density?

Based on student results

2. Did the seed predator’s activity density vary according to location? If so, which seed predator activity density varied and why?

Based on student results

3. Do you think that they activity density of these seed predators will change throughout the summer? Explain.

Yes, seed predator activity density will change. Activity density will be the lowest in the beginning of summer until all eggs hatch and adults emergefrom hibernation, activity density will be lowest towards the end of the summer when generations start to die off.

4. What habits do you think these seed predators prefer in regards to soil disturbances and canopy cover and why?

Seed predators like habits with a dense canopy cover with provides protection from other predators such as birds. They also prefer habits with low numbers of soil disturbances (i.e. tillage); soil disturbances can cause mortality.

Student Sheet

Weed Seed Predator Identification and Activity Density

Name:

Date:

Class:

Introduction: Weed seed predation is the removal of potentially viable seeds from the weed seedbank. Numerous organisms have been found to be weed seed predators. Some of the most common are rodents, birds, ants, crickets, and carabid beetles. The amount of seed predation occurring in a cropping system is directly related to the activity density of that seed predator. For the purpose of this demonstration activity density is the number of seed predators per pitfall; this number represents the number of seed predators that are searching for food on the soil surface. In this experiment you will be determining the activity density of two ground beetles Harpalus pensylvanicus and Amara aenea. This information will approximate the amount of seed predation that may be occurring by Harpalus pensylvanicus and Amara aenea. Activity density can vary greatly between cropping systems. Annual cropping systems have unique characteristics because of the high frequency and intensity of both physical and chemical disturbances required to maintain them. Intensive management of agricultural landscapes such as frequent cultivation and pesticide application can negatively affect abundance, diversity and efficiency of ground beetles and other weed seed predators. Multiple studies have shown that seed predators have a higher activity density in cropping systems that require fewer soil disturbances and produce a denser canopy cover. Many seed predators live in the soil, frequent soil disturbances can cause habitat deterioration, emigrantion to more preferable habitats and in some cases can cause death. Cropping systems with dense canopy covers provide these biocontrol agents with much needed protection from other predators.

Materials Needed Per Group: • 6- 2L soda bottles cut into two so that top forms a funnel • 6- 24 oz deli containers or 6 2L soda bottle bottoms • 6- 8 oz Styrofoam cups with lids • Antifreeze • 6 aluminum pie pans • Tweezers • Seed Predator Identification Guide

Procedure: • Place 3 pitfalls randomly throughout the field burying them so that the lip of the 2-liter bottle is flush with the soil surface ( See photo 3). • Once the pitfall traps are in place pour one inch of antifreeze into the Styrofoam cup, this acts as the killing agent.

• Pitfalls are left out/open in the field for 72 hours. • Collect samples and bring them into the classroom • Pour pitfall contents into aluminum pie pans • Separate contents into 5 piles, one for each of the weed seed predators mentioned earlier (Harpalus pensylvanicus, Amara aenea, ants and crickets) and another pile for other insects (disregard). • Counted the number of each seed predator found for each pitfall and fill data table • Find the average activity density for each of the seed predators over the three pitfalls for your group and write it on the chalk. • Using the class results on the chalk board fill in data table 2.

Photo 1: Pitfall materials Photo 2: Pitfall Photo 3: Pitfall in field

Analyzing Data: Data Table 1. : Your Group’s Results Pitfall Number Ants Crickets Harpalus Amara aenea pensylvanicus

1

2

3

Data Table 2. : Averaged Results for Class Group Ants Crickets Harpalus Amara aenea Number pensylvanicus 1

2

3

4

5

1. Which seed predator had the highest activity density and which seed predator had the lowest activity density?

2. Did the seed predator’s activity density vary according to location? If so, which seed predator activity density varied and why?

3. Do you think that they activity density of these seed predators will change throughout the summer? Explain.

4. What habits do you think these seed predators prefer in regards to soil disturbances and canopy cover and why?

Weed Seed Predator Identification Guide

Crickets Harpalus •2 3/10 inches pensylvanicus long • ½- 5/8 inch long •Brown or black • Long black body body • Li ght brown •Large barbed appendages hind legs •Actual size • Actual size

Meredi th Murray Uni versity of Missouri

Ants Amara aenea • 1/20 to ½ inch • 5/16-3/8 inch • Body color can long be red, brown • Copperish or black metallic body •6 legs • Footbal l shaped • 3 body • Relatively flat segments •Actual size •Actual size

Wikipedi a L’ encycl opedia li bre Meredith Murray

Teacher Sheet Weed Seed Predation

Name:

Date:

Class:

Introduction: Weed seed predation is the removal of potentially viable seeds from the weed seedbank. Numerous organisms have been found to be weed seed predators. Some of the most common are rodents, birds, ants, crickets, and ground beetles. Seed predation has been found to be responsible for up to80-90% of weed seed bank depletion. Seed predation can vary greatly between cropping systems. Annual cropping systems have unique characteristics because of the high frequency and intensity of both physical and chemical disturbances required to maintain them. Intensive management of agricultural landscapes such as frequent cultivation and pesticide application can negatively affect abundance, diversity and efficiency of ground beetles and other weed seed predators. Multiple studies have shown that seed predation is greater in cropping systems that require fewer soil disturbances and produce a denser canopy cover.

Materials per Group: • 6 small Petri dishes • 6 Large Petri dishes • Double sided carpenter tape • 0.5 in Medal wire mesh • 18- 6 in bolts • Sand • 600 Giant Foxtail seeds or similar weed/crop seeds • Duct Tape

Procedure: 1. Assign the students to 5 groups. 2. Have students remove paper off of six small Petri dishes with double sided tape, then sprinkle one packet of seeds (100 giant foxtail seeds) and sand onto the sticky surface. The seed cards will be used to determine the rates of

predation [(subtract the number of remaining seeds on the seed card from 100)x100]= % seed predation. (Photo 1) 3. Each group selects a location in the landscape and randomly places three open traps and three closed traps. Open traps represents predation by invertebrates and vertebrates. This trap consists of three bolts being duct tapped to an inverted large Petri dish. The bolts are then pressed into the soil about an inch deep over top of the seed card. (Photo 2) The closed traps represents invertebrate feeding only. This trap consists of 0.5 inch metal wire mesh being cut and made into a cylinder with a large Petri dish fitting securely on top. (Photo 3) The trap bolts are pushed about an inch into the soil surface and has a seed card placed inside it. 4. The seed cards are placed so that they are flush with the soil surface directly under the large Petri dish that makes up the top of each of the traps. 5. The seed cards are left out in the field for up to 14 days at which time the seed cards are brought into the classroom, the number of remaining seeds counted and subtracted from 100 to get the percent predation occurring in the field. 6. Record and compare your results to the result of the class to determine the average amount of seed predation occurring in the field for invertebrates (closed trap) and invertebrates+vertebrates (open trap).

Seed Card (1) Open Trap (2) Closed Trap (3)

Analyzing Results: Date Table Your Group’s Data Trap Number Number of Seeds % Seed Number of Seeds % Seed Predation Remaining Open Predation Open Remaining Closed Closed Trap Trap Trap Trap 1

2

3

Average

Class Data Averaged Over Traps Group Number Number of % Seed Predation Number of Seeds % Seed Predation Seeds Open Trap Remaining Closed Trap Remaining Closed Trap Open Trap

1

2

3

4

5 Average

1. Which trap had the greatest percent seed predation? Why do you think this occurred?

More predation may occur in the open traps because both invertebrates and vertebrates have access to the seed card.

2. Did seed predation rates vary according to location? Why or why not?

There may have been more predation in areas with more ground cover and less disturbance.

3. Do you think the rate of seed predation you saw occurring would have a large impact on next year’s weed population in that location?

Seed predation rates may have been high however seed predation rates vary by location. Seed predation alone may not significantly decrease crop yield loss; additional weed management methods such as crop rotation, selective herbicides and tillage are generally needed.

4. What weed control practices would you use in ADDITION to weed seed predation to help reduce weed population in that location?

Farmers can use crop rotation, intercropping, cover crops, tillage, low toxicity herbicides…..

Student Sheet Weed Seed Predation

Name:

Date:

Class:

Introduction: Weed seed predation is the removal of potentially viable seeds from the weed seedbank. Numerous organisms have been found to be weed seed predators. Some of the most common are rodents, birds, ants, crickets, and ground beetles. Seed predation has been found to be responsible for up to 80-90% of weed seed bank depletion. Seed predation can vary greatly between cropping systems. Annual cropping systems have unique characteristics because of the high frequency and intensity of both physical and chemical disturbances required to maintain them. Intensive management of agricultural landscapes such as frequent cultivation and pesticide application can negatively affect abundance, diversity and efficiency of ground beetles and other weed seed predators. Multiple studies have shown that seed predation is greater in cropping systems that require fewer soil disturbances and produce a denser canopy cover.

Materials per Group: • 6 small Petri dishes • 6 Large Petri dishes • Double sided carpenter tape • 0.5 in Medal wire mesh • 18- 6 in bolts • Sand • 600 Giant Foxtail seeds or other similar weed/crop seeds • Duct Tape

Procedure: 1. Remove paper off of 6 small Petri dishes with double sided tape, then sprinkle 1 packet of seeds (100 giant foxtail seeds) and sand onto the sticky surface. The seed cards will be used to determine the rates of predation [(subtract the number

of remaining seeds on the seed card from 100) x 100]= % seed predation. (Photo 1) 2. Select a location in the landscape and randomly places three open traps and three closed traps. Open traps represents predation by invertebrates and vertebrates. This trap consists of three bolts being duct tapped to an inverted large Petri dish. Press the bolts into the soil about an inch deep. (Photo 2) 3. The closed traps represents invertebrate feeding only. This trap consists of 0.5 inch metal wire mesh being cut and made into a cylinder with a large Petri dish fitting snugly on top. Push the trap about an inch into the soil surface. (Photo 3) 4. Place a seed card so that it is flush with the soil surface directly under the large Petri dish that makes up the top of each of the traps. 5. Leave the seed cards out in the field for 14 days. 6. Bring seed cards back into the classroom and count the number of remaining seeds. 7. Subtract the number of remaining seeds from 100 to get the percent predation occurring in the field. 8. Record and compare your results to the result of the class to determine the average amount of seed predation occurring in the field for invertebrates (closed trap) and invertebrates+vertrebrates (open trap).

Seed Card (1) Open Trap (2) Closed Trap (3)

Analyzing Results: Date Table Your Group’s Data Trap Number of Seeds % Seed Number of Seeds % Seed Predation Number Remaining Open Predation Remaining Closed Trap Trap Open Trap Closed Trap 1

2

3

Average

Classes Data Averaged Over Traps Group Number of Seeds % Seed Number of % Seed Predation Number Remaining Open Predation Open Seeds Closed Trap Trap Trap Remaining Closed Trap

1

2

3

4

5 Average

1. Which trap had the greatest percent seed predation? Why do you think this occurred?

2. Did seed predation rates vary according to location? Why or why not?

3. Do you think the rate of seed predation you saw occurring would have a large impact on next year’s weed population in that location?

4. What weed control practices would you use in ADDITION to weed seed predation to help reduce weed population in that location?