Plant Community and White-Tailed Deer Nutritional Carrying Capacity Response to Intercropping Switchgrass in Loblolly Pine Plantations

Mississippi State University Scholars Junction

Theses and Dissertations Theses and Dissertations

1-1-2016

Plant Community and White-tailed Deer Nutritional Carrying Capacity Response to Intercropping Switchgrass in Loblolly Pine Plantations

Ethan Jacob Greene

Follow this and additional works at: https://scholarsjunction.msstate.edu/td

Recommended Citation Greene, Ethan Jacob, "Plant Community and White-tailed Deer Nutritional Carrying Capacity Response to Intercropping Switchgrass in Loblolly Pine Plantations" (2016). Theses and Dissertations. 3547. https://scholarsjunction.msstate.edu/td/3547

This Graduate Thesis - Open Access is brought to you for free and open access by the Theses and Dissertations at Scholars Junction. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholars Junction. For more information, please contact [email protected]. Template A v3.0 (beta): Created by J. Nail 06/2015

Plant community and white-tailed deer nutritional carrying capacity response to

intercropping switchgrass in loblolly pine plantations

By TITLE PAGE Ethan Jacob Greene

A Thesis Submitted to the Faculty of Mississippi State University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Wildlife, Fisheries, and Aquaculture in the Department of Wildlife, Fisheries, and Aquaculture

Mississippi State, Mississippi

May 2016

2016

Plant community and white-tailed deer nutritional carrying capacity response to

intercropping switchgrass in loblolly pine plantations

Ethan Jacob Greene

Approved:

______Stephen Demarais (Major Professor)

______Darren A. Miller (Committee Member)

______Scott A. Rush (Committee Member)

______Kevin M. Hunt (Graduate Coordinator)

______Andy J. Kouba (Department Head)

______George M. Hopper Dean College of Forest Resources

Name: Ethan Jacob Greene ABSTRACT Date of Degree: May 6, 2016

Institution: Mississippi State University

Major Field: Wildlife, Fisheries, and Aquaculture

Major Professor: Stephen Demarais

Title of Study: Plant community and white-tailed deer nutritional carrying capacity response to intercropping switchgrass in loblolly pine plantations

Pages in Study: 90

Candidate for Degree of Master of Science

Switchgrass (Panicum virgatum L.) is a cellulosic feedstock for alternative energy

production that could grow well between planted pines (Pinus spp.). Southeastern planted

pine occupies 15.8 million hectares and thus, switchgrass intercropping could affect biodiversity if broadly implemented. Therefore, I evaluated effects of intercropping switchgrass in loblolly pine (P. taeda L.) plantations on plant community diversity, plant biomass production, and white-tailed deer (Odocoileus virginianus Zimmerman)

nutritional carrying capacity. In a randomized complete block design, I assigned three

treatments (switchgrass intercropped, switchgrass monoculture, and a “control” of

traditional pine management) to 4 replicates of 10-ha experimental units in Kemper

County, Mississippi during 2014-2015. I detected 246 different plant species.

Switchgrass intercropping reduced plant species richness and diversity but maintained

evenness. I observed reduced forb and high-use deer forage biomass but only in

intercropped alleys (interbeds). Soil micronutrient interactions affected forage protein of

deer plants. White-tailed deer nutritional carrying capacity remained unaffected.

DEDICATION

The foresight and design for this research was the product of an exceptional ecologist, conservationist, and colleague. I therefore wish to dedicate this manuscript to

Dr. Sam K. Riffell, his wife Angie, and daughters, Hannah, and Abby. He was a trusted academic advisor and friend. He is sorely missed by all who had the privilege of knowing him.

ACKNOWLEDGEMENTS

It is with great sincerity that I express my appreciation for those individuals who supported me both personally and professionally in the completion of this research. To my wife Rachel, I greatly thank you for helping me accomplish these goals. Without your enduring support and love I simply would not have succeeded. I thank my parents for their many years of love and support, and for their persistence in my education.

It has been an honor to work with an incredible company of colleagues who have provided tremendous support, advice, and reassurance throughout my graduate academic career. I thank Dr. Steve Demarais for his initiative and selflessness as an academic advisor. Without the support and flexibility of Dr. Scott Rush and Dr. Darren Miller this project would have never been completed. For all the humor and advice I thank Brad

Wheat. For help in the field, office, and outside the office, I thank Craig Marshall. Zac

Loman, you considerably helped me to understand the intricacies of data analysis and for that I am grateful. To the numerous others whom I’ve received help in a cacophony of ways, I express to you my sincere thanks. Thank you, Dr. Jeanne Jones for being the inspiring person that you are. Thank you Scott Edwards, Dr. Lisa Wallace, Dr. Bruce

Leopold, Jason Bies, Daniel Sullivan, Jenny Foggia, Tony Francis, and all the technicians who gave tremendous effort for this project; Nick Deutsch, Skylar Keller, Jonathan

Marguet, Jade McCarley, Logan McConnaughey, Samantha Pipo, and Ciera Rhodes.

iii

TABLE OF CONTENTS

DEDICATION ...... ii

ACKNOWLEDGEMENTS ...... iii

LIST OF TABLES ...... vi

LIST OF FIGURES ...... vii

CHAPTER

I. ENERGY AND CONSERVATION IN PINE PLANTATIONS ...... 1

Literature Cited ...... 6

II. PLANT COMMUNITY RESPONSE TO INTERCROPPING SWITCHGRASS IN LOBLOLLY PINE PLANTATIONS ...... 9

Methods ...... 10 Study Area ...... 10 Study Design ...... 11 Statistical Methods ...... 13 Results ...... 14 Evenness ...... 14 Richness ...... 15 Shannon Diversity ...... 16 Summary ...... 17 Discussion ...... 17 Management Implications ...... 20 Literature Cited ...... 29

III. SOIL NUTRIENT AVAILABILITY EFFECT ON NUTRITIONAL VALUE OF THREE WHITE-TAILED DEER FORAGES ...... 33

Methods ...... 36 Study Area ...... 36 Study Design ...... 36 Statistical Methods ...... 40 Results ...... 40 Thoroughwort ...... 40 iv

Blackberry ...... 41 Goldenrod ...... 41 Summary ...... 41 Discussion ...... 42 Management Implications ...... 43 Literature Cited ...... 51

IV. PLANT BIOMASS PRODUCTION AND WHITE-TAILED DEER NUTRITIONAL CARRYING CAPACITY RESPONSE TO INTERCROPPING SWITCHGRASS IN LOBLOLLY PINE PLANTATIONS ...... 54

Methods ...... 56 Study Area ...... 56 Study Design ...... 57 Statistical Methods ...... 60 Results ...... 61 Total Biomass ...... 61 Interbed Biomass ...... 61 Pine Bed Biomass ...... 62 White-Tailed Deer Nutritional Carrying Capacity ...... 63 Summary ...... 64 Discussion ...... 64 Total Biomass ...... 64 Interbed Biomass ...... 65 Pine Bed Biomass ...... 66 White-Tailed Deer Nutritional Carrying Capacity ...... 66 Summary ...... 67 Management Implications ...... 68 Literature Cited ...... 76

APPENDIX

A. PLANT SPECIES DETECTED DURING STUDY IN KEMPER COUNTY, MS ...... 81

LIST OF TABLES

1 Species evenness, richness, and Shannon diversity differences...... 22

2 Species evenness, richness, and Shannon diversity contrasts...... 23

3 Interbed/pine bed evenness, richness, and Shannon diversity contrast...... 25

4 Mean species evenness, richness, and Shannon diversity by sampling period...... 26

5 Mean interbed/pine bed species evenness, richness, and Shannon diversity...... 27

6 Total vegetative biomass mixed model ANOVA results...... 69

7 Total vegetative biomass collected by treatment and year...... 70

8 Mixed model ANOVA results for interbeds...... 71

9 Interbed vegetative biomass collected by treatment and year...... 72

10 Mixed model ANOVA results for pine beds...... 73

11 Pine bed vegetative biomass collected by treatment and year...... 74

12 Plant species detected (2014-2015) in Kemper County MS, USA...... 82

LIST OF FIGURES

1 Ten Hectare Treatment Unit Sampling Design for Diversity Sampling ...... 28

2 Treatment Unit Sampling Design for Soil and Deer Forage Collection ...... 44

3 Thoroughwort NMDS Plot without Treatment ...... 45

4 Thoroughwort NMDS Plot with Similar Treatment ...... 46

5 Blackberry NMDS Plot without Treatment ...... 47

6 Blackberry NMDS Plot with Similar Treatment ...... 48

7 Goldenrod NMDS Plot without Treatment ...... 49

8 Goldenrod NMDS Plot with Similar Treatment ...... 50

9 Ten Hectare Treatment Unit Sampling Design for Biomass Sampling ...... 75

vii

ENERGY AND CONSERVATION IN PINE PLANTATIONS

Humans face a growing energy problem. Conventional oil reserves worldwide are

expected to reach maximum production within the decade (Almeida and Silva 2007,

Benes et al. 2015, Mohr et al. 2015, Owen et al. 2010, Sorrell et al. 2010 and references

therein). Meanwhile, geopolitical influences and global climate change policy are likely

to limit fossil fuel emissions released into the atmosphere to prevent breaching climate

goals (Jakob and Hilaire 2015, and Verbruggen and Van de Graff 2015). Declining liquid

fuel availability, increasing population size, and increasing demand may also lead to economically expedient energy extraction with negative environmental impacts (e.g. mountain top removal coal extraction, artic deep-ocean drilling, and Canadian tar sand extraction) (Farrell and Brandt 2006). Driven by this multifaceted problem, interest has intensified in producing renewable cellulosic feedstocks for human energy needs (Perlack et al. 2005). Switchgrass (Panicum virgatum L.) intercropping is an agroforestry practice

that simultaneously produces a bioenergy feedstock and timber commodity. I explored

switchgrass intercropping and the effect it has on plant communities to meaningfully

contribute to solving these growing issues.

A lternative renewable energies have the potential to alleviate fossil fuel

dependency and thus interest in renewables has grown (Koh & Ghazoul 2008,

McLaughlin et al. 2002, Owen et al. 2010, and Perlack et al. 2005). In the early 1980s

and 1990s, the U.S. Department of Energy (DOE) provided funding to various

institutions to conduct trial studies on potential biofuel feedstocks (Wright and Turhollow

2010). Contracted scientists evaluated thirty-four herbaceous plant species, including several grasses, and a diverse array of non-herbaceous species (Tilman et al. 2006,

Wright and Turhollow 2010). Switchgrass was included in the assessment for its high productivity and sustainability as a cellulosic feedstock (Wright and Turhollow 2010).

Via this evaluation, DOE identified switchgrass as a model bioenergy crop.

Switchgrass stabilized soil, produced sustainable yields, and demanded relatively small amounts of nutrients, water, energy, and chemicals (McLaughlin et al. 2002, Fletcher et al. 2011 and references therein). Proficiency of switchgrass, a C4 plant, as a cellulosic feedstock is a function of its efficient cellular respiration, perennial growth, and ability to grow well in mild climates of southeastern U.S. crop lands (Albaugh et al. 2012, Riffell et al 2012). Furthermore, switchgrass was found to sequester more carbon below ground than many other biofuel feedstocks and could be combined with a stable market (e.g., timber industries) to reduce economic risks associated with entering developing renewable energy markets (Albaugh et al. 2012).

Viability of these renewable energy agroforestry practices are documented. For example, forest managers in the southeastern U.S. investigated dual cropping of loblolly pine (Pinus taeda L.). Dual cropping allows for viable production of timber and biofuel products simultaneously by cultivating two cohorts of evenly spaced pine trees and

harvesting cohorts at different intervals (Scott and Tiarks 2008). Switchgrass

intercropping replaces one cohort of loblolly pine with high yielding switchgrass (Riffell

et al. 2011, Iglay et al. 2012a, and Briones et al. 2013). This method will likely become

more important as the infrastructure for processing sustainable biofuels develops and greenhouse gas emissions, food commodity prices, and human health costs increase

(McLaughlin et al. 2002).

Management intensity of pine plantations will likely increase if a need for

switchgrass biofuels and feedstocks expands (Riffell et al. 2012 and references therein).

There are 15.8 million hectares of planted pine forests and 12 million hectares of

managed loblolly pine and shortleaf pine (Pinus echinata Mill.) forests in the

southeastern U.S., and thus great potential for significant alterations to forest lands

should switchgrass intercropping become widespread (Smith et al. 2009, Wear and Greis

2012). A positive correlation between biodiversity and ecological services (e.g. plant

community diversity and herbivore diets) (Tilman et al. 1996) exists (Cardinale et al.

2006, Cardinale et al. 2011, Flynn et al. 2009). Therefore, an alteration to biodiversity at

the order of magnitude implicated by switchgrass intercropping could affect numerous

biological and ecological processes (Fletcher et al. 2011, Riffell et al. 2012, Ridley et al.

2013).

Several studies have examined avian (Iglay et al. 2012b, Loman et al. 2013),

herpetofaunal (Iglay et al. 2013) and insect (Iglay et al. 2012c) community response to

landscape level disturbances in pine plantations. Avian communities responded

negatively to increased management intensity and switchgrass establishment (Iglay et al.

2012b, Loman et al. 2013). Although, herpetofauna did not respond to treatments of

burning and herbicide application in thinned pine plantations, carabid beetle abundance

declined (Iglay et al. 2012c, 2013). Briones et al. (2013) found that switchgrass

intercropping in pine plantations did not significantly affect the ecological role of white-

footed mouse (Peromyscus leucopus Rafinesque). Homyack et al. (2013) also found no

significant impact on diversity of herpetofauna in North Carolina two years after

switchgrass establishment. Nevertheless, Briones et al. (2013) and Homyack et al. (2013)

suggest long-term studies are needed to fully understand how switchgrass intercropping

will impact herpetofauna diversity.

Two short-term studies examined plant community response to switchgrass

intercropping. Iglay et al. (2012a) demonstrated that species richness, diversity, and

coverage of herbaceous plants increased, while woody plants decreased in response to

intercropping two years after switchgrass establishment. More recently, Wheat (2015)

observed no significant difference in plant diversity between intercropped and pine

control plots two years after the switchgrass establishment period. Again, both authors

recommend long-term studies examining plant community response to switchgrass

intercropping after establishment (Iglay et al. 2012a, Wheat 2015). In summary, it is

critical that an analysis of biodiversity response to switchgrass intercropping be

performed to optimize renewable production while conserving biodiversity (Riffell et al.

2012).

I addressed the need for evaluating longer-term effects (more time after

establishment) of switchgrass intercropping on plant communities by quantifying

biodiversity metrics (i.e. species evenness, species richness, and species diversity) within

different treatments and testing for inherent differences between those treatments. I also

studied the interactions between soil nutrient availability and plant quality for white-

tailed deer to better understand heterogeneity of the herbivorous nutritional environment.

Lastly, I examined how switchgrass intercropped and traditionally managed pine

treatments differed in vegetative productivity and nutritional carrying capacity for white-

tailed deer. White-tailed deer were a model species for my research because of their

ecological importance as keystone herbivores, the strong connection between plant

communities and herbivore nutrition (Waller and Alverson 1997), and their social and

economic importance (Miller 2001). Through this study, a multi-perspective assessment

of plant community response to switchgrass intercropping was gained.

Literature Cited

Albaugh, J. M., E. B. Sucre, Z. H. Leggett, J. Domec, J. S. King. 2012. Evaluation of intercropped switchgrass establishment under a range of experimental site preparation treatments in a forested setting on the lower coastal plain of North Carolina, USA. Biomass Bioenergy 46:673-682.

Almeida, P., and P. Silva. 2007. Peak of oil production- a markets perspective. Proceedings of 3rd International Energy, Exergy and Environment Symposium. University of Evora, Portugal.

Benes, J., M. Chauvet, O. Kamenik, M. Kumhof, D. Laxton, S. Mursula, J. Selody. 2015. The future of oil: geology versus technology. International Journal of Forecasting 31:207-221.

Briones, K. M., J. A. Homyack, D. A. Miller, M. C. Kalcounis-Rueppell. 2013. Intercropping switchgrass with loblolly pine does not influence the functional role of the white-footed mouse (Peromyscus leucopus). Biomass Bioenergy 54:191- 200.

Cardinale, B. J., D. S. Srivastava, J. E. Duffy, J. P. Wright, A. L. Downing, M. Sankaran, C. Jouseau. 2006. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature 443:989-992.

Cardinale, B. J., K. L. Matulich, D.U. Hooper, J. E. Byrnes, E. Duffy, L. Gamfeldt, P. Balvanera, M. I. O’Connor, A. Gonzalez. 2011. The functional role of producer diversity in ecosystems. American Journal of Botany 98:572-592.

Farrell, A. E., and A. R. Brandt. 2006. Risks of the oil transition. Environmental Research Letters 1:1-6.

Fletcher, R. J., B. A. Robertson, J. Evans, P. J. Doran, J.R. Alavalapati, D. W. Schemske. 2011. Biodiversity conservation in the era of biofuels: risks and opportunities. Frontiers in Ecology and the Environment 9:161-168.

Flynn, D.F.B., M. Gogol-Prokurat, T. Nogeire, N. Molinari, B.T. Richers, B.B. Lin, N. Simpson, M.M. Mayfield, F. DeClerck. 2009. Loss of functional diversity under land use intensification across multiple taxa. Ecology Letters 12:22-33.

Homyack, J. A., Z. Aardweg, T. A. Gorman, D. R. Chalcraft. 2013. Initial effects of woody biomass removal and intercropping of switchgrass (Panicum virgatum) on herpetofauna in eastern North Carolina. Wildlife Society Bulletin 37:327-335.

Iglay, R. B., S. K. Riffell, D. A. Miller, and B. D. Leopold. 2012a. Effects of switchgrass intercropping and biomass harvesting on plant communities in intensively managed pine stands. Pages 159-163 in Proceedings from Sun Grant National Conference: Science for Biomass Feedstock Production and Utilization, New Orleans, Louisiana, USA.

Iglay, R. B., S. Demarais, T. B. Wigley, D. A. Miller. 2012b. Bird community dynamics and vegetation relationships among stand establishment practices in intensively managed pine stands. Forest Ecology Management 283:1-9.

Iglay, R. B., D. A. Miller, B. D. Leopold, G. Wang. 2012c. Carabid beetle response to prescribed fire and herbicide in intensively managed, mid-rotation pine stands in Mississippi. Forest Ecology Management 281:41-47.

Iglay, R. B., B. D. Leopold, D. A. Miller. 2013. Summer herpetofaunal response to prescribed fire and herbicide in intensively managed, mid‐rotation pine stands in Mississippi. Wildlife Society Bulletin 38:33-42.

Jakob, M., and J. Hilaire. 2015. Unburnable fossil-fuel reserves. Nature 517:150-151.

Koh, L. P., and J. Ghazoul. 2008. Biofuels, biodiversity, and people: understanding the conflicts and finding opportunities. Biological Conservation 141:2450-2460.

Loman, Z. G., S. K. Riffell, D. A. Miller, J. A. Martin, F. J. Vilella. 2013. Site preparation for switchgrass intercropping in loblolly pine plantations reduces retained trees and snags, but maintains downed woody debris. Forestry 86:353- 360.

McLaughlin, S., D. G. De La Torre Ugarte, C. Garten, L. Lynd, M. Sanderson, V. R. Tolbert, D. Wolf. 2002. High-value renewable energy from prairie grasses. Environmental Science Technology 36:2122-2129.

Miller, K.V. 2001. White-tailed deer. Pages 95-107 in J. G. Dickson, Editor. Wildlife of Southern Forests, Habitat and Management. Hancock House Publishing LTD., Blaine, Washington, USA.

Mohr, S. H., J. Wang, G. Ellem, J. Ward, D. Giurco. 2015. Projection of world fossil fuels by country. Fuel 141:120-135.

Owen, N. A., O. R. Inderwildi, D. A. King. 2010. The status of conventional world oil reserves-hype or cause for concern? Energy Policy 38:4743-4749.

Perlack, R. D., L. L. Wright, A. F. Turhollow, R. L. Graham, B. J. Stokes, D. C. Erbach. 2005. Biomass as feedstock for a bioenergy and bioproducts industry: the technical feasibility of a billion-ton annual supply. U.S. Department of Energy GO-102005-2135.

Ridley, C. E., H. I. Jager, C. M. Clark, R. A. Efroymson, C. Kwit, D. A. Landis, Z. H. Leggett, D. A. Miller. 2013. Debate: can bioenergy be produced in a sustainable manner that protects biodiversity and avoids the risk of invaders? Bulletin Ecological Society of America 94:277-290.

Riffell, S., J. Verschuyl, D. Miller, T. B. Wigley. 2011. Biofuel harvests, coarse woody debris, and biodiversity–a meta-analysis. Forest Ecology Management 261:878- 887.

Riffell, S., J. Verschuyl, D. Miller, T. B. Wigley. 2012. Potential biodiversity response to intercropping herbaceous biomass crops on forest lands. Journal of Forestry 110:42-47.

Scott, D. A., A. Tiarks. 2008. Dual-cropping loblolly pine for biomass energy and conventional wood products. Southern Journal of Applied Sciences Forestry 32:33-37.

Smith, W. B., P. D. Miles, C. H. Perry, and S. A. Pugh. 2009. Forest resources of the United States, 2007: a technical document supporting the forest service 2010 RPA assessment. General Technical Report-USDA Forest Service, WO-78.

Sorrell, S., J. Speirs, R. Bentley, A. Brandt, R. Miller. 2010. Global oil depletion: a review of the evidence. Energy Policy 38:5290-5295.

Tilman, D., D. Wedin, J. Knops. 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379:718-720.

Tilman, D., J. Hill, C. Lehman. 2006. Carbon-negative biofuels from low-input high- diversity grassland biomass. Science 314:1598-1600.

Verbruggen, A., and T. Van de Graaf. 2015. The geopolitics of oil in a carbon- constrained world. International Association for Energy Economics Energy Forum 2:21-24.

Waller, D.M., and W.S. Alverson. 1997. The white-tailed deer: a keystone herbivore. Wildlife Society Bulletin 25:217–226.

Wear, D. N., and J. G. Greis. 2012. The southern forest futures project: Summary report. United States Forest Service, General Technical Report SRS-168, Asheville, North Carolina, USA.

Wheat, B. 2015. Plant community response to intercropping switchgrass in pine plantations of Mississippi. Master’s Thesis. Mississippi State University. Mississippi State, Mississippi, USA.

Wright, L., and A. Turhollow. 2010. Switchgrass selection as a “model” bioenergy crop: A history of the process. Biomass Bioenergy 34:851-868. 8

PLANT COMMUNITY RESPONSE TO INTERCROPPING SWITCHGRASS IN

LOBLOLLY PINE PLANTATIONS

Intensively managed loblolly pine (Pinus taeda L.) in the southeastern U.S.

provides a unique opportunity to combine existing wood production with renewable

energy genesis (Riffell et al. 2012). This opportunity combined with a need for renewable

fuels initiated inquiry into productive capacities of eighteen perennial grass species

(Lewandowski et al. 2003, Tilman et al. 2006, Wright and Turhollow 2010). Switchgrass

(Panicum virgatum L.) performed the best of those researched (Lewandowski et al.

2003). Switchgrass intercropping is an agroforestry practice that grows high yielding

switchgrass feedstocks by cultivating C4 grass between rows of planted pine (Riffell et

al. 2012). Intercropped switchgrass is a versatile energy source that can be used as a

cellulosic feedstock for biofuels or co-fired with other sources of energy (Schmer et al.

2008). The Energy Independence and Security Act (EISA) of 2007 requires cellulosic

biofuel production to increase to 60.6 billion liters a year and forest biomass collected to

rise to 190508.8 million dry kilograms by 2022 (Boundy et al. 2011, Perlack et al. 2011).

Thus, potential demand for switchgrass intercropping may impact biodiversity and

ecosystem functioning across the approximately 15.8 million hectares of planted pine

forests and 12 million hectares of managed loblolly pine and shortleaf pine (Pinus

echinata Mill.) forest in the southeastern U.S., (Smith et al. 2009, Wear and Greis 2012).

Terrestrial biodiversity relies upon these expansive managed landscapes

(Brockerhoff et al. 2008, Miller et al. 2009, Pawson et al. 2013 and references therein) and intensification of management in these areas may reduce this biodiversity (Flynn et al. 2009, Fletcher et al. 2011). To better understand ecological effects of switchgrass

intercropping on biodiversity, I examined the plant community response to these practices by quantifying plant species evenness, richness, and diversity in switchgrass

intercropped pine stands and non-switchgrass intercropped pine stands in Mississippi

during 2014 and 2015.

Methods

Study Area

As part of its landholdings, Weyerhaeuser Company managed a 25,000 hectare

block for pine timber production in Kemper County, Mississippi. Four recently clear-cut

experimental stands were selected from this area in 2009. During 2009-2010, a bulldozer

fitted with v-blade plow, subsoil ripper, and bedding plow cleared woody debris and

prepared raised beds for tree planting. In each stand loblolly pine (Pinus taeda L.)

seedlings were planted in raised soil beds (hereafter “pine beds”) 1.5 meters apart. Pine

beds were spaced 6.1 meters apart leaving “interbeds” between pine rows for planting

switchgrass. Wide interbeds provided adequate light for switchgrass growth and space for

machinery to operate during switchgrass establishment, maintenance, and harvest (Riffell

et al. 2012). A mixed tank banded application of sulfometuron methyl and imazapyr

herbicides were sprayed in pine beds during the first growing season to temporarily

reduce competition.

These experimental stands, established as part of a study with Catchlight Energy,

LLC, a Weyerhaeuser Company and Chevron joint venture, were surrounded by intensively managed pine forest (70%), mature pine-hardwood (17%), hardwood/riparian areas (10%), and non-forested areas (3%) (Loman et al. 2013). The overall landscape had an annual average minimum and maximum temperature of 10.1 ºC - 23.5 ºC and received

140.2 cm of precipitation annually (National Oceanic and Atmospheric Administration

2015).

Study Design

I used a complete randomized block design to assign one of three treatments to three 10 hectare treatment units, distanced ≥ 50 meters from each other, and within each of the four experimental stands. Treatments included: traditionally managed pine control

(PINE), switchgrass intercropped (IC), and switchgrass monoculture (MONO).

Treatment descriptions are presented as follows. (1) (PINE): Treatment units selected for

PINE treatments underwent routine site preparation activities previously described. PINE treatment units were not harvested for timber or cellulosic feedstocks. (2) (IC): IC treatment units were prepared similarly to PINE treatment units with several important exceptions. In IC treatment units, most of the coarse woody debris (CWD) in interbeds was displaced into pine beds and glyphosate was applied to interbeds to reduce

competition to switchgrass prior to disking and seeding. Alamo switchgrass was

broadcast seeded in interbeds from late May to early June, 2012. Alamo was selected

because of its propensity for efficient growth in the southeastern U.S. (Sanderson et al.

1996). Switchgrass cut and baled switchgrass during December, 2013 and again in

December, 2014. Broadcast urea coated fertilizers were applied in interbeds to supply 11

approximately 60 kg/ha of nitrogen in April each year. A banded herbicide treatment of triclopyr (Garlon 4 Ultra®), metsulfuron methyl, and chlorsulfuron (Cimmaron Plus®)

was applied to interbeds in June each year to control woody encroachment and broadleaf

weeds. (3) (MONO): In MONO treatment units, practically all coarse woody debris was

displaced out of the treatment unit and glyphosate was administered before disking or

planting operations. Alamo switchgrass was broadcast seeded from late May to early

June, 2012. Switchgrass was cut and baled during December, 2013 and again in

December, 2014. Broadcast urea coated fertilizers was applied annually which supplied

approximately 60 kg/ha of nitrogen in April each year. A broadcast herbicide application

of triclopyr (Garlon 4 Ultra®), metsulfuron methyl, and chlorsulfuron (Cimmaron Plus®)

was applied in June each year to control woody encroachment and broadleaf plants.

I used the diagonal distances from the upper (NW and NE) and lower corners

(SW and SE) of treatment units to position three interior midpoints 100 meters apart using geospatial software ArcGIS (Figure 1). I used a random point spatial analyst function in ArcGIS with ≤ 50 meters programmed separation distance around each of the three interior midpoints and five meter separation distance between individual points to establish random sampling points (Roberts-Pichette and Gillespie 1999, Wheat 2015). I did this twice for each year (i.e. 2014, 2015), once for a May sampling period and again for a July sampling period. Two temporally distinct sampling periods made it possible to detect floristic changes between early spring and late summer. Thirty random points and twelve random points per treatment unit were generated for May and July sampling, respectively.

In PINE and IC treatment units, I used each geo-located random point to place a

1m2 sampling frame in the center of both the pine bed and interbed, equaling 60 frames per treatment unit (n=60) in May and 24 frames per treatment unit (n=24) in July. In

MONO treatment units, I placed only one frame at each random point because of

treatment homogeneity, equaling 30 frames per treatment unit (n=30) in May and 12

frames per treatment unit (n=12) in July. In PINE and IC treatment units, I sampled the

closest pine bed and interbed to each random point. I positioned frames in pine beds or interbeds opposite from my direction of entry to standardize choice and remove bias. I

identified all individual plant specimens detected in sampling frames to the species level.

Statistical Methods

I compiled detected species across treatment units and totaled to calculate plant species richness (hereafter “richness”), plant species evenness (hereafter “evenness”), and

Shannon Diversity Index (hereafter “diversity”) for plants. I used Shannon’s index because of its sensitivity and equitability over other indices (Peet 1974, Wheat 2015).

Using a log linear species accumulation curve in program R, I projected MONO species richness data upwards to n=60 to eliminate unequal sampling intensity between MONO

(n=30) and PINE/IC (n=60) treatment units (Colwell and Coddington 1994, R Core Team

2014). I used projected richness to calculate evenness and diversity for MONO treatment

units.

I used richness, evenness, and diversity to test for differences among treatments

by sampling period (i.e. May, July) and across years (i.e. 2014, 2015). Treatment (i.e.

PINE, IC, and MONO) and year were categorical explanatory variables. Using program

R, I fitted richness, evenness, and diversity responses to generalized linear models (R 13

Core Team 2014). I included year and treatment x year covariates within linear models. I tested for differences among treatments, years, and treatment x year using analysis of variance. In addition, I tested for differences between treatments, years, and interbed/pine

bed x year to detect incongruences between pine beds and interbeds. I calculated least

square means for each model and performed post-hoc pairwise comparisons of the

adjusted means to determine which treatments were significantly different. I used Šidák

correction of post-hoc analyses to avoid type-I error (Šidák 1967). I considered results

significant at alpha=0.05.

Results

I detected 246 species from 67 families across 1,470 1m2 sampling frames in all

four survey periods (Table 12). I sampled 630 frames in 2014 and 840 frames in 2015. Of

62 unknown plant specimens, 15 were recurrently detected whereas 46 were detected once. Seven of the 15 unknown recurrent species belonged to family Asteraceae, seven

belonged to family Poaceae, and one was unknown.

Evenness

During the May sampling period, I detected a significant difference in evenness

among treatments (F2, 29 = 18.59, P = < 0.001) but not years (F1, 29 = 0.17, P = 0.68) or

treatment x years (F2, 29 = 0.36, P = 0.70) (Table 1). I detected no difference between IC

and PINE evenness (t = -1.16, P = 0.99) in May, 2014 nor between IC and PINE

evenness (t = -0.10, P = 1.00) in May, 2015 (Table 2). I did, however, detect a significant

difference in evenness between IC and MONO (t = 3.30, P = 0.04) in May, 2014 and

again (t = 3.98, P = 0.01) in May, 2015 (Table 2). Likewise, I detected a significant

difference between PINE and MONO evenness (t = 4.25, P = 0.01) in May, 2014 and

2015 (t = 4.07, P = 0.01) (Table 2). There were no detectable differences between the

interbeds and pine beds of PINE and IC treatments (Table 3).

During July, again there was a significant difference in evenness among treatments (F2, 29 = 66.19, P = <0.001) but not years (F1, 29 = 2.07, P = 0.16) or treatment

x years (F2, 29 = 0.31, P = 0.73) (Table 1). Similar to my results in May, I detected no

difference between IC and PINE evenness (t = -0.95, P = 0.99) in 2014 nor between IC

and PINE evenness (t = -1.59, P = 0.86) in 2015 (Table 2). I detected a significant

difference between IC and MONO evenness (t = 5.91, P = < 0.001) in July, 2014 and

again (t = 7.63, P = < 0.001) in 2015 (Table 2). I also detected a significant difference

between PINE and MONO evenness (t = 6.68, P = < 0.001) in July, 2014 and 2015 (t =

8.92, P = < 0.001) (Table 2). There were no detectable differences between the interbeds

and pine beds of PINE and IC treatments (Table 3).

Overall, May and July evenness for both years shared identical treatment

responses with no detectable interactions between IC and PINE and detectable significant

interactions between IC/PINE and MONO. IC and PINE least square mean evenness was

higher than MONO treatments (Table 4). Least square means of interbed/pine bed

evenness are shown in Table 5.

Richness

I detected a significant difference among treatments (F2, 29 = 5.40, P = 0.01) but

not years (F1, 29 = 0.45, P = 0.51) or treatment x years (F2, 29 = 0.50, P = 0.61) in May

(Table 1). There were no significant differences in richness among any of the treatments

for May, 2014 or 2015 (Table 2). As before, I detected no differences in richness between

the interbeds and pine beds of PINE and IC treatments (Table 3).

In July, I detected significant richness differences among treatments (F2, 29 =

11.76, P = <0.001) but not years (F1, 29 = 3.39, P = 0.08) or treatment x years (F2, 29 =

0.19, P = 0.83) (Table 1). I detected a significant difference between IC and PINE

richness (t = -3.77, P = 0.01) in 2015 (Table 2). There were no differences between IC

and MONO or PINE and MONO treatments in July, 2014 or 2015 (Table 2). This time,

both richness of interbeds and pine beds in IC and PINE treatments were significantly

different from each other for both years (t = -5.52, P = <0.001) (Table 3).

In summary, IC and PINE treatments differed in richness in July, 2015. The least square mean in July, 2015 for IC was 39.8 species whereas the least square mean in July,

2015 PINE was 57.9 species, a difference of 18.1 species (Table 4).

Shannon Diversity

I detected a significant difference in diversity among treatments (F2, 29 = 10.05, P

= <0.001) but not years (F1, 29 = 0.39, P = 0.54) or treatment x years (F2, 29 = 0.02, P =

0.98) for the May sampling period (Table 1). There were no significant differences among any of the treatments for May, 2014 or 2015 except between PINE and MONO (t

= 3.45, P = 0.03) in 2015 (Table 2). No diversity differences between the interbeds and

pine beds of PINE and IC treatments were detected (Table 3).

In July, diversity differed among treatments (F2, 29 = 15.52, P = <0.001) but there

was not a treatment x year interactive effect (F2, 29 = 0.30, P = 0.74) (Table 1). In July,

2014 there was no difference between IC and PINE (t = -2.34, P = 0.33) or in IC and

PINE treatments contrasted with MONO; but in July, 2015 there was a significant 16

difference between IC and PINE (t = -3.49, P = 0.02) (Table 2). Additionally, there was a

difference in diversity between PINE and MONO treatments in July, 2015 (t = 4.32, P =

0.01) (Table 2). Like richness in July, both the diversity of interbeds and pine beds in IC and PINE treatments were significantly different from each other for both years (t = -

4.93, P = 0.01) (Table 3).

Summary

To summarize all results, there was greater evenness in IC/PINE treatments than

MONO and no differences between IC and PINE (Table 4). Richness was greater in

PINE treatments than in IC treatments in July, 2015 (Table 4). Similarly, diversity was greater in PINE treatments than both IC and MONO treatments across all significant

interactions (Table 4).

Discussion

Previous studies demonstrate that switchgrass intercropping alters species

richness, evenness, and diversity. Iglay et al. (2012) demonstrated that during years two

and three after establishment, switchgrass intercropping promoted diversification of plant

communities in response to increases in light and nutrient availability. As part of the

current study, Wheat (2015) demonstrated similar results with some important

exceptions. Following poor switchgrass establishment in the first year, a subsequent

replanting occurred one year later (Wheat 2015). Diversity and richness declined after

replanting because of an additional herbicide treatment and mechanical disturbance

(Wheat 2015). However, one year post establishment, there were no differences between

PINE and IC treatments (Wheat 2015). Overall, additional site preparation was

implicated as the source of richness and diversity losses during establishment (Swindel et

al. 1989, Miller et al. 1995, Jones et al. 2009, Lane et al. 2011, Grace et al. 2011, Wheat

2015).

My research followed Wheat’s (2015) work for two additional years and found a

similar result in year three (2014) as did Wheat in year two (2013). Similarities in our

findings are likely explained by the lack of additional disturbances and consistencies in

treatment activities from 2013-2014 with a minimal change in succession. However,

successional changes paired with cumulative treatment effects may explain significant

differences between IC and PINE treatments in July, 2015. Richness and diversity decline

with decreasing light and nutrients as vegetative communities mature and canopy closure

is completed (Campbell et al. in press). Multiple broadcast herbicide treatments in pine

plantations also inhibit early successional species establishment beyond optimal

conditions for the plant assemblage to flourish (Jones et al. 2012). Thus, succession and

repeated herbicide applications due to failure of the initial switchgrass planting provide a

likely explanation for the changes observed between IC and PINE mean richness and

diversity from 2013 to 2015. Wheat’s (2015) mean estimates for richness and diversity in

IC and PINE treatments changed from being higher in IC than PINE in 2012 to being

higher in PINE than IC in 2013, with several exceptions. This trend stabilized during

2014 and intensified in 2015 suggesting synergistic impacts on IC and interbed areas

driven by succession and treatment activities (Table 4 and 5). The gradual separation of

similarities between IC and PINE became significant in July, 2015 with differences in

richness and diversity.

To understand this distinction, it is important to recognize that significant

reductions in identifiable forbs in interbeds of IC treatments followed June herbicide

applications. Interbed species assemblages were reduced to little more than switchgrass

and other graminoids (i.e. grasses, sedges, and rushes) approximately one week post

application. Forb species detected were typically shade tolerant and buried underneath

dominating switchgrass and graminoids out of reach of direct foliar contact. Triclopyr

must reach target foliage to be most effective and is not bound tightly to soil particles,

limiting residual soil activity (Wigley et al. 2002). Triclopyr controls broadleaf species

while releasing graminoids (Miller et al. 2010). Metsulfuron methyl and chlorsulfuron

suppress herbaceous plants (Rhodes and Phillips 2012). From my results, it is clear that

these herbicides temporarily eliminated broadleaf species, released switchgrass in

interbeds, suppressed richness and diversity in interbeds of IC treatments, and ultimately

reduced plant community biodiversity.

In MONO treatments, the plant community responded in a way akin to IC interbed

plant assemblages. Again, herbicide eliminated broadleaf species, released switchgrass, and suppressed diversity. Evenness of MONO treatments likely differed from evenness of

PINE and IC treatments because of the homogeneity of switchgrass and graminoids in

MONO treatments. Essentially, switchgrass and other graminoids occurred far more frequently and thus disrupted the balance of species occupancy skewing evenness in

MONO treatments. It is also likely that the lack of treatment differences Wheat (2015) found in evenness in 2013 and subsequent differences detected in evenness both May and

July, 2014-2015 for MONO treatments can be primarily explained by the same phytotoxic selectivity of the herbicides. Synergistic lingering effects and immediate

short-term effects from broadcast herbicide applications may explain differences between

PINE and MONO diversity May, 2015 and July, 2015 (Table 2).

In conclusion, intercropping switchgrass as a process does not improve or necessarily negatively impact plant community evenness, richness, or diversity during the first 3 years after establishment. However, synergistic effects attributable to switchgrass

intercropping did negatively impact plant communities during this time. Herbicides were

likely the predominant negative influence on plant diversity and, when coupled with seral

changes, negated potential benefits of wider planting distance and delay of canopy

closure characteristic of switchgrass intercropping (Andreu et al. 2008, Riffell et al.

2012). Between interbed management disturbance and unabated succession in pine beds

it will be difficult for plant communities to persist under current management strategies.

Given the clearing, seeding, possible reseeding, and annual herbicide + switchgrass

harvest, it is also likely that plant communities will be unable to maintain pre-

intercropping evenness, richness, and diversity levels on a longer temporal scale.

Management Implications

Land managers must balance switchgrass establishment, weedy competition, and

plant community diversity. Switchgrass establishment can be difficult and incomplete

during establishment (Schmer et al. 2006, Garland 2008, Albaugh et al. 2012). Despite

this difficulty, incomplete switchgrass dominance must be tolerated to maintain plant

diversity. Using alternative herbicide applications that do not eliminate broadleaf species

may be a viable way of protecting plant diversity over time (Hartley 2002). I suggest that managers explore alternative establishment techniques to minimize diversity loss and reach production goals. I also recommend that studies of plant community response to 20

switchgrass intercropping be continued until canopy closure to follow these trends and to

test alternatives. While Wheat (2015) suggests that switchgrass intercropping may only

have temporary effects on plant communities, my results suggest that repeated temporary

influences had a perpetual effect.

Table 1 Species evenness, richness, and Shannon diversity differences.

Response Month Effect MS F-statistic P-value Evenness May Treatment 0.0062207 18.59 < 0.001* Year 0.0000569 0.17 0.68 Treatment: Year 0.0001191 0.36 0.70 July Treatment 0.0200634 66.20 < 0.001* Year 0.0006279 2.07 0.16 Treatment: Year 0.0000949 0.31 0.73 Richness May Treatment 353.08 5.40 0.01* Year 29.49 0.45 0.51 Treatment: Year 32.98 0.50 0.61 July Treatment 1089.95 11.76 < 0.001* Year 314.23 3.39 0.08 Treatment: Year 17.76 0.19 0.83 Diversity May Treatment 0.220339 10.05 < 0.001* Year 0.008499 0.39 0.54 Treatment: Year 0.000357 0.02 0.98 July Treatment 0.87517 15.52 < 0.001* Year 0.2004 3.55 0.07 Treatment: Year 0.01692 0.30 0.74 Differences among treatments, years, and treatment x year interactions for community metrics (Richness, Evenness, Diversity = Shannon diversity) during study in Kemper Co., MS 2014–2015. Mean square (MS) values are included for reference and reverse calculation. P-values* represents significant interactions at α = 0.05.

Table 2 Species evenness, richness, and Shannon diversity contrasts.

Response Month Year Contrast Estimate T-Statistic P-value Evenness May 2014 IC-PINE -0.01 (0.01) -1.16 0.99 IC-MONO 0.04 (0.01) 3.30 0.04* PINE-MONO 0.06 (0.01) 4.25 0.01* 2015 IC-PINE -0.001 (0.00915) -0.11 1.00 IC-MONO 0.0446 (0.0112) 3.98 0.01* PINE-MONO 0.0455 (0.0112) 4.07 0.01* July 2014 IC-PINE -0.00951 (0.010052) -0.95 0.99 IC-MONO 0.0727 (0.0123) 5.91 < 0.001* PINE-MONO 0.0822 (0.0123) 6.68 < 0.001* 2015 IC-PINE -0.0138 (0.008705) -1.59 0.86 IC-MONO 0.0813 (0.01066) 7.63 < 0.001* PINE-MONO 0.0951 (0.01066) 8.92 < 0.001* Richness May 2014 IC-PINE -6.5 (4.671) -1.39 0.94 IC-MONO -2.206 (5.721) -0.39 1.00 PINE-MONO 4.293 (5.721) 0.75 0.99 2015 IC-PINE -12.25 (4.04506) -3.03 0.07 IC-MONO -2.481 (4.954) -0.50 1.00 PINE-MONO 9.769 (4.954) 1.97 0.59 July 2014 IC-PINE -16.5 (5.558) -2.97 0.09 IC-MONO -7.872 (6.80706) -1.16 0.99 PINE-MONO 8.628 (6.80706) 1.27 0.97 2015 IC-PINE -18.125 (4.813) -3.77 0.01* IC-MONO -3.931 (5.895) -0.67 1.00 PINE-MONO 14.194 (5.895) 2.41 0.29

Table 2 (Continued)

Diversity May 2014 IC-PINE -0.138 (0.0855) -1.62 0.84 IC-MONO 0.149 (0.1047) 1.43 0.93 PINE-MONO 0.288 (0.1047) 2.75 0.14 2015 IC-PINE -0.148 (0.07402) -2.00 0.57 IC-MONO 0.165 (0.09066) 1.82 0.71 PINE-MONO 0.312 (0.09066) 3.45 0.03* July 2014 IC-PINE -0.3209 (0.137) -2.34 0.33 IC-MONO 0.143 (0.168) 0.85 1.00 PINE-MONO 0.464 (0.168) 2.77 0.14 2015 IC-PINE -0.414 (0.119) -3.49 0.02* IC-MONO 0.215 (0.145) 1.48 0.91 PINE-MONO 0.629 (0.145) 4.32 0.01* Treatment contrasts for each sampling period and community metrics (Richness, Evenness, Diversity = Shannon Diversity) during study in Kemper Co., MS 2014–2015. Treatments are: IC = switchgrass intercropped, MONO = switchgrass monoculture, PINE = traditionally managed control. Estimates are least square means with standard error and were considered significant at α = 0.05. P-values* represents significant interactions.

Table 3 Interbed/pine bed evenness, richness, and Shannon diversity contrast.

Response Month Year Contrast Estimate T-Statistic P-value Evenness May 2014 I, IC - PINE -0.01 (0.01) -2.11 0.74 B, IC - PINE -0.01 (0.01) -2.11 0.74 2015 I, IC - PINE -0.01 (0.01) -2.11 0.74 B, IC - PINE -0.01 (0.01) -2.11 0.74 July 2014 I, IC - PINE -0.01 (0.01) -2.11 0.74 B, IC - PINE -0.01 (0.01) -2.11 0.74 2015 I, IC - PINE -0.01 (0.01) -2.11 0.74 B, IC - PINE -0.01 (0.01) -2.11 0.74 Richness May 2014 I, IC - PINE -9.79 (3.12) -3.14 0.12 B, IC - PINE -9.79 (3.12) -3.14 0.12 2015 I, IC - PINE -9.79 (3.12) -3.14 0.12 B, IC - PINE -9.79 (3.12) -3.14 0.12 July 2014 I, IC - PINE -17.43 (3.16) -5.52 <0.001* B, IC - PINE -17.43 (3.16) -5.52 <0.001* 2015 I, IC - PINE -17.43 (3.16) -5.52 <0.001* B, IC - PINE -17.43 (3.16) -5.52 <0.001* Diversity May 2014 I, IC - PINE -0.14 (0.06) -2.48 0.44 B, IC - PINE -0.14 (0.06) -2.48 0.44 2015 I, IC - PINE -0.14 (0.06) -2.48 0.44 B, IC - PINE -0.14 (0.06) -2.48 0.44 July 2014 I, IC - PINE -0.37 (0.08) -4.93 0.01* B, IC - PINE -0.37 (0.08) -4.93 0.01* 2015 I, IC - PINE -0.37 (0.08) -4.93 0.01* B, IC - PINE -0.37 (0.08) -4.93 0.01* Interbed/pine bed treatment contrasts for each sampling period and community metrics (Richness, Evenness, Diversity = Shannon Diversity) during study in Kemper Co., MS 2014–2015. Treatments are: IC = switchgrass intercropped, PINE = traditionally managed control. Letter “I” = interbed and letter “B” = pine bed. Estimates are least square means with standard error. Results were considered significant at α = 0.05. P- values* represents significant interactions.

Table 4 Mean species evenness, richness, and Shannon diversity by sampling period.

Response Month Year Treatment IC PINE MONO x̄ LCL UCL x̄ LCL UCL x̄ LCL UCL Evenness May 2014 0.87A 0.86 0.89 0.89A 0.87 0.90 0.83B 0.81 0.85 July 2014 0.91A 0.90 0.93 0.92A 0.91 0.93 0.84B 0.82 0.86 May 2015 0.88A 0.87 0.90 0.88A 0.87 0.90 0.84B 0.82 0.86 July 2015 0.90A 0.89 0.91 0.92A 0.90 0.93 0.82B 0.80 0.84 Richness May 2014 70.5A 63.7 77.3 77.0A 70.2 83.8 72.7A 63.2 82.3 July 2014 45.7A 37.6 53.7 62.2A 54.1 70.2 53.5A 42.2 64.9 May 2015 70.0A 64.2 75.8 82.3A 76.4 88.1 72.5A 64.2 80.8 July 2015 39.8A 32.8 46.7 57.9B 50.9 64.8 43.7AB 33.8 53.5 Diversity May 2014 3.71A 3.59 3.84 3.85A 3.73 3.98 3.56A 3.39 3.74 July 2014 3.47A 3.27 3.67 3.79A 3.59 3.99 3.33A 3.05 3.61 May 2015 3.74AB 3.64 3.85 3.89A 3.79 4.00 3.58B 3.43 3.73 July 2015 3.30A 3.12 3.47 3.71B 3.54 3.88 3.08A 2.84 3.32 Least square mean estimates for each sampling period and community metrics (Richness, Evenness, Diversity = Shannon Diversity) during study in Kemper Co., MS 2014–2015. Treatments are: IC = switchgrass intercropped, MONO = switchgrass monoculture, PINE = traditionally managed control. Differences among treatments means are designated by different letters within rows. Differences were considered significant at α = 0.05.

Table 5 Mean interbed/pine bed species evenness, richness, and Shannon diversity.

Response Month Year Location Treatment IC PINE x̄ LCL UCL x̄ LCL UCL Evenness May 2014 Interbed 0.87A 0.85 0.88 0.87A 0.86 0.89 Pine Bed 0.89A 0.88 0.90 0.90A 0.88 0.91 July 2014 Interbed 0.91A 0.89 0.92 0.92A 0.90 0.93 Pine Bed 0.91A 0.90 0.93 0.93A 0.91 0.94 May 2015 Interbed 0.88A 0.86 0.89 0.88A 0.87 0.90 Pine Bed 0.88A 0.87 0.90 0.89A 0.88 0.90 July 2015 Interbed 0.89A 0.88 0.90 0.90A 0.89 0.92 Pine Bed 0.92A 0.90 0.93 0.93A 0.92 0.94 Richness May 2014 Interbed 69.1A 61.4 76.8 78.9A 71.2 86.6 Pine Bed 68.6A 60.9 76.3 78.4A 70.7 86.1 July 2014 Interbed 45.3A 37.5 53.1 62.7B 54.9 70.5 Pine Bed 45.1A 37.4 52.9 62.6B 54.8 70.3 May 2015 Interbed 70.1A 63.3 76.9 79.9A 73.1 86.7 Pine Bed 72.4A 65.5 79.2 82.1A 75.3 89.0 July 2015 Interbed 34.4A 27.5 41.3 51.8B 44.9 58.8 Pine Bed 45.8A 38.9 52.7 63.2B 56.3 70.1 Diversity May 2014 Interbed 3.67A 3.52 3.81 3.81A 3.67 3.95 Pine Bed 3.76A 3.61 3.90 3.90A 3.76 4.04 July 2014 Interbed 3.42A 3.23 3.61 3.80B 3.61 3.98 Pine Bed 3.47A 3.29 3.66 3.85B 3.66 4.03 May 2015 Interbed 3.72A 3.59 3.85 3.86A 3.74 3.99 Pine Bed 3.77A 3.65 3.90 3.92A 3.79 4.05 July 2015 Interbed 3.14A 2.97 3.31 3.51B 3.35 3.68 Pine Bed 3.49A 3.33 3.66 3.87B 3.70 4.03 Least square mean estimates of community metrics (Richness, Evenness, Diversity = Shannon diversity) for interbeds/pine beds during each sampling period for study in Kemper Co., MS 2014–2015. Treatments are: IC = switchgrass intercropped, PINE = traditionally managed control. Differences between treatments means are designated by different letters within rows. Results were considered significant at α = 0.05.

Figure 1 Ten Hectare Treatment Unit Sampling Design for Diversity Sampling

Example of 10 hectare treatment unit containing 50 m buffer area, midpoints, and 3 ≤ 50 m circular separation areas with (n=60) sampling frames at random locations. Interbeds and pine beds were oriented west to east.

Literature Cited

Albaugh, J. M., E. B. Sucre, Z. H. Leggett, J-C. Domec, J. S. King. 2012. Evaluation of intercropped switchgrass establishment under a range of experimental site preparation treatments in a forested setting on the lower coastal plain of North Carolina, USA. Biomass and Bioenergy 46:673-682.

Andreu, M., K. Zobrist, T. Hinkley. 2008. Management practices to support increased biodiversity in managed loblolly pine plantations. Florida Cooperative Extension Service Publication FOR183. 1-8.

Boundy, B., S. W. Diegel, L. Wright, S. C. Davis. 2011. Biomass energy data book edition 4. ORNL/TM-2011/446. Oak Ridge National Laboratory: Oak Ridge, Tennessee, USA.

Brockerhoff, E. G., H. Jactel, J. A. Parrotta, C. P. Quine, J. Sayer. 2008. Plantation forests and biodiversity: oxymoron or opportunity? Biodiversity and Conservation 17:925-951.

Campbell, T. N., P. D. Jones, S. Demarais, and A. W. Ezell. In press. Vegetation communities in intensively established loblolly pine plantations at crown closure. Journal of Forestry.

Colwell, R. K., and J. A. Coddington. 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transaction: Biological Sciences 345:101-118.

Fletcher, R. J., B. A. Robertson, J. Evans, P. J. Doran, J. R. R. Alavalapati, D. W. Schemske. 2011. Biodiversity conservation in the era of biofuels: risks and opportunities. Frontiers in Ecology and the Environment 9:161-168.

Garland, C.D. 2008. Growing and harvesting switchgrass for ethanol production in Tennessee. Publication no. SP701-A. University of Tennessee Extension, Knoxville, Tennessee, USA.

Grace, J. L., M. J. Chamberlain, D. A. Miller, and P. C. Stouffer. 2011. Effects of row spacing and debris distribution on vegetative communities in newly established loblolly pine plantations of Louisiana. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 65:38-44.

Hartley, M. J., 2002. Rationale and methods for conserving biodiversity in plantation forests. Forest Ecology and Management 155:81-95.

Iglay, R. B., S. K. Riffell, D. A. Miller, and B. D. Leopold. 2012. Effects of switchgrass intercropping and biomass harvesting on plant communities in intensively managed pine stands. Pages 159-163 in Proceedings from Sun Grant National Conference: Science for Biomass Feedstock Production and Utilization, New Orleans, Louisiana, USA.

Jones, P. D., S. L. Edwards, S. D. Demarais, and A. W. Ezell. 2009. Vegetation community responses to different establishment regimes in loblolly pine (Pinus taeda) plantations in southern MS, USA. Forest Ecology and Management 257:553-560.

Jones, P. D., S. Demarais, and A. W. Ezell. 2012. Successional trajectory of loblolly pine (Pinus taeda) plantations established using intensive management in Southern Mississippi, USA. Forest Ecology and Management 265:116-123.

Lane, V. R., K. V. Miller, S. B. Castleberry, D. A. Miller, T. B. Wigley, G. M. Marsh, and R. L. Mihalco. 2011. Plant community responses to a gradient of site preparation intensities in pine plantations in the Coastal Plain of North Carolina. Forest Ecology and Management 262:370-378.

Lewandowski, I., J.M.O. Scurlock, E. Lindvall, M. Christou. 2003. The development and current status of perennial rhizomatous grasses as energy crops in the US and Europe. Biomass and Bioenergy 25:335-361.

Miller, J. H., B. R. Zutter, S. M. Zedaker, M. B. Edwards, and R. A. Newbold. 1995. Early plant succession in loblolly pine plantations as affected by vegetation management. Southern Journal of Applied Forestry 19:109-126.

Miller, D. A., T. B. Wigley, K. V. Miller. 2009. Managed forests and conservation of terrestrial biodiversity in the southern United States. Journal of Forestry 107:197- 203.

Miller, J. H., S. T. Manning, S. F. Enloe. 2010. A management guide for invasive plants in southern forests. United States Department of Agriculture Forest Service Gen. Technical Report SRS-13I. 120 p.

National Oceanic and Atmospheric Administration. 2010. Southern Regional Climate Center, climate normals. Available from http://www.srcc.lsu.edu/climate_normals.html. Accessed 22 October 2015.

Pawson, S. M., A. Brin, E. G. Brockerhoff, D. Lamb, T.W. Payn, A. Paquette, J. A. Parrotta. 2013. Plantation forests, climate change and biodiversity. Biodiversity and Conservation 22:1203-1227.

Peet, R. K. 1974. The measurement of species diversity. Annual Review of Ecology and Systematics 5:285-307.

Perlack, R. D., and B.J. Stokes (leads), U.S. Department of Energy. 2011. U.S. billion-ton update: biomass supply for a bioenergy and bioproducts industry. ORNL/TM- 2011/224. Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA.

R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Rhodes, G. N. Jr., W. P. Phillips Jr. 2012. PB 1801 Weed Management in Pastures and Hay Crops. University of Tennessee Extension Publications, Knoxville, Tennessee, USA.

Riffell, S. K., J. Verschuyl, D. Miller, T. B. Wigley Jr. 2012. Potential biodiversity response to intercropping herbaceous biomass crops on forest lands. Journal of Forestry 110:42-47.

Roberts-Pichette, P., and L. J. Gillespie. 1999. Terrestrial vegetation biodiversity monitoring protocols. EMAN Occasional Paper Series Report No. 9, Ontario, Canada.

Sanderson, M. A., R. L. Reed, S. B. McLaughlin, S. D. Wullschleger, B. V. Conger, D. J. Parrish, D. D. Wolf, C. Taliaferro, A. A. Hopkins, W. R. Ocumpaugh, M. A. Hussey, J. C. Read, C. R. Tischler. 1996. Switchgrass as a sustainable bioenergy crop. Bioresource Technology 56:83-93.

Schmer, M. R., K. P. Vogel, R. B. Mitchell, L. E. Moser, K. M. Eskridge, R. K. Perrin. 2006. Establishment stand thresholds for switchgrass grown as a bioenergy crop. Crop Science 46:157-161.

Schmer, M. R., K. P. Vogel, R. B. Mitchell, R. K. Perrin. 2008. Net energy of cellulosic ethanol from switchgrass. Proceedings of the National Academy of Sciences 105: 464-469.

Šidák, Z. 1967. Rectangular confidence regions for the means of multivariate normal distributions. Journal of the American Statistical Association 62:626-633.

Swindel, B. F., J. E. Smith, D. G. Neary, and N. B. Comerford. 1989. Recent research indicates plant community responses to intensive treatment including chemical amendments. Southern Journal of Applied Forestry 13:152-156.

Tilman, D., J. Hill, C. Lehman. 2006. Carbon-negative biofuels from low-input high diversity grassland biomass. Science 314:1598-1600.

Wear, D. N., and J. G. Greis. 2012. The southern forest futures project: Summary report. United States Forest Service, General Technical Report SRS-168, Asheville, North Carolina, USA.

Wheat, B. 2015. Plant community response to intercropping switchgrass in pine plantations of Mississippi. Master’s Thesis. Mississippi State University. Mississippi State. Mississippi, USA.

Wigley, T. B., K. V. Miller, D. S. deCalesta, M. W. Thomas. 2002. Herbicides as an alternative to prescribed burning for achieving wildlife management objectives. In: Ford, M. W., K. R. Russell, C. E. Moorman, (editors). Proceedings: the role of fire for nongame wildlife management and community restoration: traditional uses and new directions. Gen. Tech. Rep. NE-288. Newtown Square, PA: U.S. Dept. of Agriculture, Forest Service, Northeastern Research Station: 124-138.

Wright, L., and A. Turhollow. 2010. Switchgrass selection as a “model” bioenergy crop: A history of the process. Biomass Bioenergy 34:851-868.

SOIL NUTRIENT AVAILABILITY EFFECT ON NUTRITIONAL VALUE OF

THREE WHITE-TAILED DEER FORAGES

Understanding soil nutrient availability within pine (Pinus spp. L. ) stands

intercropped with switchgrass (Panicum virgatum L.) may foster a greater understanding of plant community nutritional environments. Switchgrass intercropping is an

agroforestry practice that facilitates energy feedstock and wood production simultaneously by growing switchgrass between planted rows of loblolly pine (P. taeda

L.) (Riffell et al. 2012). In a complex system, like switchgrass intercropped pine

plantations, there is considerable natural variability. For example, Long Term Ecological

Research (LTER) demonstrated that variable local level decomposition rates do not

accurately indicate appropriate rate of decomposition at a global scale, as was previously

thought (Kratz et al. 2003). Recognizing nuances of decomposition provided a better estimate of carbon storage for analyses examining effects of climate change (Kratz et al.

2003). Another study showed that seemingly insignificant variability in precipitation timing and duration, irrespective of total accumulation, greatly altered plant community dynamics (Knapp et al. 2002). Cambardella et al. (1994) found that information about edaphic variables such as organic carbon, power of hydrogen, and total nitrogen could

not be extrapolated from one field to another for herbicide applications, planting, and

fertilization because of spatial variability. As part of this study, Wheat (2015) uncovered

variability in crude protein of composite samples constituted by identical plant species. It is clear from these examples that ignoring variability can result in inconclusive research findings and improper management recommendations.

The nutritional environment for white-tailed deer (Odocoileus virginianus

Zimmerman) (hereafter “deer”) epitomizes ecological variability. Sunlight availability,

sunlight quality, temperature, soil nutrients, plant maturation, interspecific interactions,

and other factors work together to change what, where, and how deer feed (Oelberg 1956,

Jones et al., 2008, Jones et al. 2009). Seasonal changes in nutritional value of forages

create more nutritious environments during warm seasons than cool seasons (White

1983). Drought can drive deer to select the most nutritious forages from fewer available

high-quality forages (Lashley and Harper 2012). Soil characteristics of the Mississippi

Delta region foster higher deer forage quality and deer body mass than non-delta regions

(Jones et al. 2008). Varying rates of plant maturation reduces digestible energy and

protein later in the growing season (Blaser 1964, Blair and Epps 1967, Buaphan et al.

2011). Increased shade intensity causes reductions in nutritional value of forages (Blair et

al. 1983). Even commensal relationships between rhizobia and leguminous plant species

increase forage quality and nutritional variability (Kempel et al. 2009). Ignoring these

variables could result in an incorrect analysis of nutritional environments. Specifically,

ignoring the interrelationship between soil nutrient availability and plant nutritional

quality could greatly skew our understanding of the ability of switchgrass intercropped

areas to support lactation-level deer diets (Lashley et al. 2015).

Several other studies have researched variability in deer forage nutritional quality.

In Louisiana, longleaf-flatwood systems had the poorest nutritional quality whereas

bottomland hardwood systems had the best nutritional quality; the authors found that crude protein and calcium content of forages best explained deer body growth, antler mass, and reproductive fertility (Horrell et al. 2015). In Mississippi, a soil geographic database was not sensitive enough to explain local variation in deer morphometrics and greater detail in soil nutrient spatial distribution was recommended (Strickland and

Demarais 2006). Jacobson (1984) regressed mean soil nutrient levels with mean mass and

antler measures to isolate soil nutrients phosphorus, potassium, calcium, magnesium,

organic matter, and power of hydrogen as correlative to deer morphometrics. Despite

these findings, little information is available on correlative nature of soil nutrient

availability and crude protein of deer forages.

Therefore, my purpose was to identify soil nutrients important to the nutritional

quality of deer forages to further understand drivers of nutritional variability across

intercropped and non-intercropped treatments. By identifying edaphic variables that contribute greatly to nutritional variance via a multivariate analysis, I was able to

illuminate any potential factors that might confound treatment effects. More importantly

though, I performed this analysis to obtain a richer understanding of ecological variability

and switchgrass intercropping effects on plant nutritional quality. This analysis helped

evaluate potential impact of switchgrass intercropping on plant nutrition with

consideration that switchgrass intercropping could be adopted across 15.8 million

hectares of planted pine forests and 12 million hectares of managed loblolly pine and

shortleaf pine (P. echinata Mill.) forest in the southeastern U.S. (Smith et al. 2009, Wear

and Greis 2012).

Methods

Study Area

As part of its landholdings, Weyerhaeuser Company managed a 25,000 hectare

block for pine timber production in Kemper County, Mississippi. Four recently clear-cut

experimental stands were selected from this area in 2009. During 2009-2010, a bulldozer

fitted with v-blade plow, subsoil ripper, and bedding plow cleared woody debris and

prepare raised beds for tree planting. In each stand loblolly pine seedlings were planted in raised soil beds (hereafter “pine beds”) 1.5 meters apart. Pine beds were spaced 6.1

meters apart leaving “interbeds” between pine rows for planting switchgrass. Wide

interbeds provided adequate light for switchgrass growth and space for machinery to operate during switchgrass establishment, maintenance, and harvest (Riffell et al. 2012).

A mixed tank banded application of sulfometuron methyl and imazapyr herbicides were

sprayed in pine beds during the first growing season to temporarily reduce competition.

These experimental stands, established as part of a study with Catchlight Energy,

LLC, a Weyerhaeuser Company and Chevron joint venture, were surrounded by

intensively managed pine forest (70%), mature pine-hardwood (17%), hardwood/riparian

areas (10%), and non-forested areas (3%) (Loman et al. 2013). The overall landscape had

an annual average minimum and maximum temperature of 10.1 ºC - 23.5 ºC and received

140.2 cm of precipitation annually (National Oceanic and Atmospheric Administration

2015).

Study Design

I used a complete randomized block design to assign one of three treatments to three 10 hectare treatment units, distanced ≥ 50 meters from each other, and within each 36

of the four experimental stands. Treatments included: traditionally managed pine control

(PINE), switchgrass intercropped (IC), and switchgrass monoculture (MONO).

Treatment descriptions are presented as follows. (1) (PINE): Treatment units selected for

PINE treatments underwent routine site preparation activities previously described. PINE treatment units were not harvested for timber or cellulosic feedstocks. (2) (IC): IC

treatment units were prepared similarly to PINE treatment units with several important

exceptions. In IC treatment units, most of the coarse woody debris (CWD) in interbeds

was displaced into pine beds and glyphosate was applied to interbeds to reduce

competition to switchgrass prior to disking and seeding. Alamo switchgrass was

broadcast seeded in interbeds from late May to early June, 2012. Alamo was selected

because of its propensity for efficient growth in the southeastern U.S. (Sanderson et al.

1996). Switchgrass was cut and baled during December, 2013 and again in December,

2014. Broadcast urea coated fertilizers were applied in interbeds to supply approximately

60 kg/ha of nitrogen in April each year. A banded herbicide application of triclopyr

(Garlon 4 Ultra®), metsulfuron methyl, and chlorsulfuron (Cimmaron Plus®) was

applied to interbeds in June each year to control woody encroachment and broadleaf

weeds. (3) (MONO): In MONO treatment units, practically all coarse woody debris was

displaced out of the treatment unit and glyphosate was administered before disking or

planting operations. Alamo switchgrass was broadcast seeded from late May to early

June, 2012. Switchgrass was cut and baled during December, 2013 and again in

December, 2014. Broadcast urea coated fertilizer was applied annually which supplied

approximately 60 kg/ha of nitrogen in April each year. A broadcast herbicide application

of triclopyr (Garlon 4 Ultra®), metsulfuron methyl, and chlorsulfuron (Cimmaron Plus®) was applied in June each year to control woody encroachment and broadleaf plants.

I used the diagonal distances from the upper (NW and NE) and lower corners

(SW and SE) of treatment units to position three interior midpoints 100 meters apart using geospatial software ArcGIS (Figure 2). Within ≤ 50 meters of midpoints, I actively

searched for three common non-leguminous plants from an empirically compiled list of

deer forages; they were Eupatorium serotinum Michx. (hereafter “thoroughwort”), Rubus

argutus Link (hereafter “blackberry”), and Solidago canadensis L. (hereafter

“goldenrod”) (Warren and Hurst 1981, Miller and Miller 1999, Jones et al. 2008, Jones et al. 2009, Gee et al. 1994, Wheat 2015) (Table 12). I collected physiologically mature and immature individuals of blackberry and goldenrod to control for plant maturation effects.

Mature blackberry was ≥ 1 meter tall and immature blackberry ≤ 1 meter tall. Mature

goldenrod was ≥ 0.5 meter tall and immature goldenrod ≤ 0.5 meter tall. All

thoroughwort collected was ≥ 0.5 meter tall. Mature and immature classifications were

not absolute criteria but comparative criteria constructed from field observations and

morphological data collected by Miller and Miller (1999). Overall plant collection sample

size was 15 plant samples from IC/PINE treatments and ≤ 9 plant samples from MONO

treatments.

I handpicked a minimum of 5 grams (dry weight) of determinately located leaf

tissue from mature and immature plant individuals to comprise species samples in June,

2014. I considered leaves to be the vegetative structure below plant buds and not bracts or involucres. I only incorporated leaf tissue from the uppermost 20.32 cm (8 inches) of

growing stems into samples to avoid inconsistent sampling effects on crude protein

estimates (Bailey 1967, Blair and Epps 1967). I avoided diseased, damaged, or otherwise

aberrant individuals, interbeds, and low-lying areas with standing water. I collected plant

tissues from pine beds and not interbeds to control for sunlight availability; the only

exception was MONO treatments which did not contain beds.

I stored leaf tissue samples in a commercial refrigerator for 4-5 days until drying

samples in a forced air oven for 72 hours at 60 ºC, and sent them to the Mississippi State

University Animal Nutrition Laboratory (MSUANL) for crude protein analysis (Haufler

& Servello 1996, Jones et al. 2008). At MSUANL, samples were ground in The Wiley

Mill® (Thomas Scientific™, Swedesboro NJ, USA) and analyzed for their nitrogen content using the Kjeldahl procedure. During the Kjeldahl procedure, dried samples were digested in a sulfuric acid and catalyst solution producing ammonia, which was then

captured and titrated to determine nitrogen concentration of samples (Haufler & Servello

1996). Nitrogen estimates were multiplied at MSUANL by 6.25 to obtain crude protein

estimates (Haufler & Servello 1996, Jones et al. 2008, Wheat 2015). The coefficient 6.25

was used because proteins on average are 16% nitrogen (100/16 = 6.25) (Haufler &

Servello 1996).

In addition to collecting forage samples of blackberry, goldenrod, and

thoroughwort, I compiled 90 soil cores per treatment unit (30 soil cores per midpoint)

with a 38.1 cm by 1.91 cm diameter (15 inch by 3/4 inch diameter) soil sampler at a

depth of 20.32 cm (8 inches) within pine beds during June, 2014. Soil cores were

extracted randomly on both sides of the midpoint and within the pine bed closest to midpoints (i.e. 15 cores per side) to control for sunlight availability (Figure 2). I brushed

away vegetative debris at soil core sites before sampling. I mixed all soil cores by hand,

placed them into labeled sample boxes, and transported them to the Mississippi State Soil

Laboratory for determination of available phosphorus, potassium, calcium, magnesium,

sulfur, zinc, organic matter, and power of hydrogen (pH).

Statistical Methods

Data exhibited multicollinearity and thus I used non-metric multidimensional scaling (NMDS) in program R to construct ordination models to best explain crude

protein of thoroughwort, blackberry, and goldenrod (R Core Team 2014). I square root

transformed data and submitted data to Wisconsin double standardization. I used Bray-

Curtis dissimilarity to define differences in edaphic variables. I assigned 0.95 confidence

level circumference ellipses around treatments to examine dissimilarities between treatments. I chose an a priori two dimensional representation for all dissimilarities.

Results

I collected 144 composite plant samples and 36 composite soil samples during

June, 2014. I did not collect immature thoroughwort because of insufficiently abundant

specimens across all treatments. Blackberry and mature goldenrod were not sufficiently

abundant in MONO treatments and thus were not collected.

Thoroughwort

Crude protein (cp) of thoroughwort was similar in ranking to power of hydrogen

(ph) in my NMDS model (Figure 3). I too found that sulfur (s), organic matter (om), and

sodium (na) were less dissimilar in ranking to thoroughwort crude protein than calcium

(ca), potassium (k), magnesium (mg), zinc (zn), and phosphorus (p) (Figure 3).

Intercropped areas were similar in ranking to non-intercropped areas and fell within

similar two dimensional space (Figure 4). Power of hydrogen ranged from 4.3 to 6.2 and crude protein ranged from 12.0 to 25.5.

Blackberry

Blackberry mature (matcp) and immature (immcp) crude protein was similar in

ranking to power of hydrogen, much like thoroughwort (Figure 5). Organic matter,

sodium, and calcium were more similar to crude protein than potassium, magnesium,

zinc, and phosphorus (Figure 5). Intercropped and non-intercropped areas were ranked

similarly amidst edaphic variables (Figure 6). Power of hydrogen ranged from 4.3 to 5.4, mature crude protein ranged from 8.7 to 16.3, and immature crude protein ranged from

9.2 to 14.3.

Goldenrod

Again, goldenrod mature and immature crude protein was least dissimilar from

power of hydrogen (Figure 7). Sulfur, organic matter, sodium, and calcium were less

dissimilar than potassium, zinc, magnesium, and phosphorus (Figure 7). Furthermore,

intercropped and non-intercropped areas overlapped in two dimensional ordination space

amidst edaphic variables. (Figure 8). Power of hydrogen ranged from 4.3 to 6.2, mature crude protein ranged from 12.6 to 18.0, and immature crude protein ranged from 9.9 to

19.3.

Summary

All three analyses yielded similar graphical responses demonstrating similarity between power of hydrogen and crude protein values. Crude protein varied for each

species tested despite controlling for sunlight, location, maturity, and plant parts included.

Discussion

From my analyses, it is clear that soil power of hydrogen is important to thoroughwort, blackberry, and goldenrod nutritional value. Variability in power of hydrogen impacts availability of nutrients for uptake in plants, a well-known phenomenon (Bianchini and Mallarino 2002, Olfs et al. 2013 and references therein). For

example, high power of hydrogen greatly reduced availability of zinc (Hafeez et al.

2013). A zinc deficiency can lead to chlorosis, growth suppression, and loss of nutritional

value (Hafeez et al. 2013). At my study site, crude protein variability observed by Wheat

(2015) and I can be explained by variability in power of hydrogen. However, there are

certainly more nuanced interactions between soil pH and nutrient expression that

influence crude protein of deer food plants not evidenced by this study. More important is

the evidence that supports edaphic quantity similarities between treatments. The lack of dissimilarity between intercropped and non-intercropped treatment suggests that there were no violations in assumptions of randomization for a complete randomized block design used at my study site.

Awareness of the complex difference in soil nutrient quantity and plant availability has provided insight into variability of plant nutrition. The understanding gained has guided me to prefer local site data over averaged data for estimation of nutritional value of deer food species and increased accuracy of deer nutritional carrying capacity estimates used to test for switchgrass intercropping effects on plant communities. Just as LTER research captured variability in decomposition rates that 42

improved global estimates, I too have increased estimation accuracy of plant community

nutrition in my switchgrass intercropped study area (Kratz et al. 2003). I recommend that

future analyses weight power of hydrogen so that additional information be gained

concerning other edaphic variables.

Management Implications

It is important to recognize that many variables influence nutritional outcomes.

Generally considered variables, such as nitrogen and phosphorus, should not take

precedence over less notable variables like pH. Additionally, simple supplementation

solutions are not always appropriate for manipulating soil nutrient availability due to

physical and chemical barriers. I recommend that managers become aware of both soil

nutrient availability and plant expression of nutrients because understanding availability, uptake, and inhibitory effects of edaphic variables are critical for meeting nutritional goals.

Figure 2 Treatment Unit Sampling Design for Soil and Deer Forage Collection

Figure 3 Thoroughwort NMDS Plot without Treatment

Figure 4 Thoroughwort NMDS Plot with Similar Treatment

Overlapping crude protein (cp) and power of hydrogen (ph) surrounded by sulfur (s), sodium (na), organic matter (om), calcium (ca), potassium (k), magnesium (mg), zinc (zn), phosphorus (p), and observation sites (gray circles). Treatment: SG = Switchgrass Monoculture, PC = Pine Control, IC = Switchgrass Intercropped. IC treatment falls behind PC.

Figure 5 Blackberry NMDS Plot without Treatment

Overlapping mature crude protein (matcp), immature crude protein (immcp), and power of hydrogen (ph) surrounded by sulfur (s), sodium (na), organic matter (om), calcium (ca), potassium (k), magnesium (mg), zinc (zn), phosphorus (p), and observation sites (gray circles).

Figure 6 Blackberry NMDS Plot with Similar Treatment

Figure 7 Goldenrod NMDS Plot without Treatment

Figure 8 Goldenrod NMDS Plot with Similar Treatment

Literature Cited

Bailey, J. A. 1967. Sampling deer browse for crude protein. The Journal of Wildlife Management 31:437-442.

Bianchini, A. A., and A. P. Mallarino. 2002. Soil-sampling alternatives and variable-rate liming for a soybean–corn rotation. Agronomy Journal 94:1355-1366.

Blair, R. M., and E. A. Epps JR. 1967. Distribution of protein and phosphorus in spring growth of rusty blackhaw. Journal of Wildlife Management 31:188-190.

Blair, R. M., R. Alcaniz, and A. Harrell. 1983. Shade intensity influences the nutrient quality and digestibility of southern deer browse leaves. Journal of Range Management 36:257-264.

Blaser, R., 1964. Symposium on forage utilization: effects of fertility levels and stage of maturity on forage nutritive value. Journal of Animal Science 23:246-253.

Buaphan, S., M. A. Froetschel, V. Pattarajinda, M. Duangjinda, Y. Opatpatanakit. 2011. Impact of forage variety and maturity at stage of harvest on in vitro rumen fermentation and microbial protein production. Proceedings of the 3rd international conference on sustainable animal agriculture for developing countries, Nakhon Ratchasima, Thailand, July, 2011. p. 284-287.

Cambardella, C. A., T. B. Moorman, J. M. Novak, T. B. Parkin, D. L. Karlen, R. F. Turco, A. E. Konopka. 1994. Field-scale variability of soil properties in central Iowa soils. Soil Science Society of America Journal 58:1501-1511.

Gee, K. L., M. D. Porter, S. Demarais, F. C. Bryant. 1994. White-tailed deer: their foods and management in the cross timbers. A Samuel Roberts Noble Foundation Publication, Ardmore, Oklahoma, USA 3-9.

Hafeez, B., Y. M. Khanif, and M. Saleem. 2013. Role of zinc in plant nutrition – a review. American Journal of Experimental Agriculture 3:374-391.

Haufler, J. B., and F. B. Servello. 1996. Techniques for Wildlife Nutritional Analyses. Research and Management Techniques for Wildlife and Habitats. Ed. T. A. Bookhout. Lawrence, Kansas USA: The Wildlife Society 307-323. Print.

Horrell, L. B., B. S. Cohen, K. V. Miller, M. J. Chamberlain. 2015. Geographical variation in nutritional quality of white-tailed deer forage plants in Louisiana. Journal of the Southeast Association of Fish and Wildlife Agencies 2:187-192.

Jacobson, H. A. 1984. Relationships between deer and soil nutrients in Mississippi. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 38:1–12.

Jones, P. D., S. Demarais, B. K. Strickland, S. L. Edwards. 2008. Soil region effects on white-tailed deer forage protein content. Southeastern Naturalist 7:595-606.

Jones, P. D., S. L. Edwards, S. D. Demarais, A. W. Ezell. 2009. Vegetation community responses to different establishment regimes in loblolly pine (Pinus taeda) plantations in southern MS, USA. Forest Ecology and Management 257:553-560.

Kempel, A., R. Brandl, and M. Schadler. 2009. Symbiotic soil microorganisms as players in aboveground plant-herbivore interactions- the role of rhizobia. Oikos 118:634- 640.

Knapp, A. K., P. A. Fay, J. M. Blair, S. L. Collins, M. D. Smith, J. D. Carlisle, C. W. Harper, B. T. Danner, M. S. Lett, J. K. McCarron. 2002. Rainfall variability, carbon cycling, and plant species diversity in a mesic grassland. Science 298:2202-2205.

Kratz, T. K., L. A. Deegan, M. E. Harmon, W. K. Lauenroth. 2003. Ecological variability in space and time: insights gained from the U. S. LTER program. BioScience 53:57-67.

Lashley, M. A., and C. A. Harper. 2012. The effects of extreme drought on native forage nutritional quality and white-tailed deer diet selection. Southeastern Naturalist 11:699-710.

Lashley, M. A., M. C. Chitwood, C. A. Harper, C. E. Moorman, C. S. DePerno. 2015. Poor soils and density-mediated body weight in deer: forage quality or quantity? Wildlife Biology 21:213-219.

Miller, J. H., and K. V. Miller. 1999. Forest plants of the southeast and their wildlife uses. Southern Weed Science Society, Craftmaster Printers, Auburn, Alabama, USA. Print.

National Oceanic and Atmospheric Administration. 2010. Southern Regional Climate Center, climate normals. Available from http://www.srcc.lsu.edu/climate_normals.html. Accessed 22 October 2015.

Oelberg, K. 1956. Factors affecting the nutritive value of range forage. Journal of Range Management 9:220-225.

Olfs, H. W., A. Borchert, and D. Trautz. 2013. Validation of an on-the-go soil pH sensing system to derive pH maps for variable lime application. Bornimer Agrartechnische Berichte, Heft 82, Potsdam-Bornim, Germany. 52

R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Riffell, S. K., J. Verschuyl, D. Miller, T. B. Wigley Jr. 2012. Potential biodiversity response to intercropping herbaceous biomass crops on forest lands. Journal of Forestry 110:42-47.

Smith, W. B., P. D. Miles, C. H. Perry, S. A. Pugh. 2009. Forest resources of the United States, 2007: a technical document supporting the forest service 2010 RPA assessment. General Technical Report-USDA Forest Service, WO-78.

Strickland, B. K., and S. Demarais. 2006. Effectiveness of the State Soil Geographic Database (STATSGO) to predict white-tailed deer morphometrics in Mississippi. Wildlife Society Bulletin 34:1264-1272.

Warren, R. C., and G. A. Hurst. 1981. Ratings of plants in pine plantations as white-tailed deer food. Mississippi Agricultural and Forestry Experiment Station Information Bulletin 18, Mississippi State University, Mississippi State, USA.

Wear, D. N., and J. G. Greis. 2012. The southern forest futures project: Summary report. United States Forest Service, General Technical Report SRS-168, Asheville, North Carolina, USA.

Wheat, B. 2015. Plant community response to intercropping switchgrass in pine plantations of Mississippi. Master’s Thesis. Mississippi State University. Mississippi State. Mississippi, USA.

White, L. M. 1983. Seasonal changes in yield, digestibility, and crude protein of vegetative and floral tillers of two grasses. Journal of Range Management 36:402- 405.

PLANT BIOMASS PRODUCTION AND WHITE-TAILED DEER NUTRITIONAL

CARRYING CAPACITY RESPONSE TO INTERCROPPING SWITCHGRASS

IN LOBLOLLY PINE PLANTATIONS

Switchgrass intercropping is an agroforestry practice that facilitates energy

feedstock and wood commodity production simultaneously by cultivating switchgrass

(Panicum virgatum L.) between planted rows of loblolly pine (Pinus taeda L.) (Riffell et

al. 2012). Switchgrass intercropping has potential to impact terrestrial biodiversity if applied across 15.8 million hectares of planted pine forests and 12 million hectares of managed loblolly and shortleaf pine (Pinus echinata Mill.) forest in the southeastern U.S.

(Smith et al. 2009, Wear and Greis 2012). Switchgrass is an attractive biofuel feedstock as yields are superior to other feedstocks; switchgrass sequesters and stores carbon in perennial roots and requires little energy, water, and chemical inputs for cultivation

(McLaughlin et al. 2002). Additionally, policy requirements mandate production of large amounts of cellulosic feedstocks, which can be partially met by incorporating switchgrass production into an already expansive managed landscape (McLaughlin et al. 2002,

Fletcher et al. 2011, Boundy et al. 2011, Perlack et al. 2011, Albaugh et al. 2012, Riffell

et al. 2012).

Intensification of management required for switchgrass intercropping in pine

plantations may negatively affect terrestrial biodiversity and the interrelationships among

those species (Flynn et al. 2009, Fletcher et al. 2011, Riffell at al. 2012). Influence of

management strategies like thinning, herbicide, burning, and site preparation in pine

plantations are well documented, but the long term synergistic effects of these strategies

combined with switchgrass intercropping is currently hypothetical (Edwards et al. 2004,

Jones et al. 2009, Mixon et al. 2009, Iglay et al. 2010a, Iglay et al. 2010b, Iglay et al.

2012, Jones et al. 2012, Lashley et al. 2015a, Campbell et al. in press). We do understand

how switchgrass establishment results in reductions of available snags for cavity nesters

(Loman et al. 2013), instigates initial diversification of plant communities (Iglay et al.

2012), and alters rodent community dynamics to favor invasive house mouse (Mus

musculus L.) over white-footed mouse (Peromyscus leucopus Rafinesque) in some

landscapes (Homyack et al. 2014). Even still, our understanding of switchgrass

intercropping and its effect on wildlife species is limited to avian, plant, and small

mammal communities.

To further understand effects of switchgrass intercropping on biotic communities

I examined the propensity for switchgrass intercropped pine plantations to support overall

plant growth (hereafter “plant biomass”) and white-tailed deer (Odocoileus virginianus

Zimmerman) (hereafter “deer”) forage. The economic and cultural importance of deer,

widespread distribution of deer populations, and the high population density of deer

justified examination alone (Waller and Alverson 1997). More importantly, deer are

keystone herbivores and impacts to deer habitat are certain to have consequences for

other terrestrial species found in managed pine landscapes (Waller and Alverson 1997,

Rooney and Waller 2003, Brockerhoff et al. 2008, Miller et al. 2009, Pawson et al. 2013).

Therefore, I researched effects of switchgrass intercropping on plant biomass production,

deer forage production, and deer nutritional carrying capacity (NCC) in 2014-2015. I quantified total plant biomass (kg/ha), total plant biomass in planted pine rows, total plant biomass in planted switchgrass rows, total deer forage biomass, total deer forage biomass

in planted pine rows, total deer forage biomass in planted switchgrass rows, deer NCC, and tested for differences between these metrics in switchgrass intercropped and non- switchgrass intercropped pine treatments in Kemper County, Mississippi, USA.

Methods

Study Area

As part of its landholdings, Weyerhaeuser Company managed a 25,000 hectare

block for pine timber production in Kemper County, Mississippi. Four recently clear-cut

experimental stands were selected from this area in 2009. During 2009-2010, a bulldozer

fitted with v-blade plow, subsoil ripper, and bedding plow cleared away woody debris

and prepared raised beds for tree planting. In each stand loblolly pine (Pinus taeda L.)

seedlings were planted in raised soil beds (hereafter “pine beds”) 1.5 meters apart. Pine beds were spaced 6.1 meters apart leaving “interbeds” between pine rows for planting switchgrass. Wide interbeds provided adequate light for switchgrass growth and space for

machinery to operate during switchgrass establishment, maintenance, and harvest (Riffell

et al. 2012). A mixed tank banded application of sulfometuron methyl and imazapyr

herbicides were sprayed in pine beds during the first growing season to temporarily

reduce competition.

These experimental stands, established as part of a study with Catchlight Energy,

LLC, a Weyerhaeuser Company and Chevron joint venture, were surrounded by

intensively managed pine forest (70%), mature pine-hardwood (17%), hardwood/riparian 56

areas (10%), and non-forested areas (3%) (Loman et al. 2013). The overall landscape had an annual average minimum and maximum temperature of 10.1 ºC - 23.5 ºC and received

140.2 cm of precipitation annually (National Oceanic and Atmospheric Administration

2015).

Study Design

I used a complete randomized block design to assign one of two treatments to two

10 hectare treatment units, distanced ≥ 50 meters from each other, and within each of

four experimental stands. Treatments included: traditionally managed pine control

(PINE) and switchgrass intercropped (IC). Treatment descriptions are as follows. (1)

(PINE): Treatment units selected for PINE treatments underwent routine site preparation

activities previously described. PINE treatment units were not harvested for timber or

cellulosic feedstocks. (2) (IC): IC treatment units were prepared similarly to PINE

treatment units with several important exceptions. In IC treatment units most of the

coarse woody debris (CWD) in interbeds was pushed into pine beds with a bulldozer, and

glyphosate was applied to interbeds to reduce competition prior to disking and seeding.

Heavy equipment broadcast seeded Alamo switchgrass in interbeds from late May to early June, 2012. Alamo was selected because of its propensity for efficient growth in the southeastern U.S. (Sanderson et al. 1996). Switchgrass was cut and baled during

December, 2013 and again in December, 2014. Broadcast urea coated fertilizers were applied in interbeds supplying approximately 60 kg/ha of nitrogen in April each year. A banded herbicide treatment of triclopyr (Garlon 4 Ultra®), metsulfuron methyl, and

chlorsulfuron (Cimmaron Plus®) was applied to interbeds in June each year to control

woody encroachment and broadleaf weeds. 57

I used the diagonal distances from the upper (NW and NE) and lower corners

(SW and SE) of treatment units to position three interior midpoints 100 meters apart using geospatial software ArcGIS (Figure 9). I used a random point spatial analyst function in ArcGIS with ≤ 50 meters programmed separation distance around each of the three interior midpoints and five meter separation distance between individual points to establish random sampling points (Roberts-Pichette and Gillespie 1999, Wheat 2015). I

generated thirty random points per treatment unit for a larger study on plant community

diversity. I used each geo-located random point to place a 1m2 sampling frame in the center of both the pine bed and interbed, equaling 60 frames per treatment unit (n=60).

The closest pine bed and interbed to the random point was sampled. I positioned frames

in pine beds or interbeds opposite from my direction of entry to standardize choice and

remove bias. I numbered each of the 10 random points around each midpoint (30 points

per treatment unit) from 0 to 9 and collected plant biomass from interbeds at even

numbered random points and biomass from pine beds at odd numbered random points,

omitting points 0 and 9. I collected plant biomass from 24 (n = 24: 12 interbeds and 12

pine beds) of the 60 plant diversity frames in each treatment unit. I clipped all plant

biomass detected by hand, separated these into stems and leaves, placed them in paper

bags, stored them in a commercial refrigerator for less than 5 days, dried them in a forced

air oven for 72 hours at 60 ºC, and weighed them to obtain a dry weight (kg/ha) biomass estimate (Haufler & Servello 1996, Jones et al. 2008, Wheat 2015).

I compiled a list of deer forages known to occupy the study area from existing

literature and via a survey of 5 wildlife biologists from Mississippi, Louisiana, and

Alabama; I ranked each forage species as high-use forages, moderate-use forages, and

low or no-use forages (Warren and Hurst 1981, Miller and Miller 1999, Jones et al. 2008,

Jones et al. 2009, Gee et al. 1994, Wheat 2015). I combined leaves of plant biomass

known to be high or moderate-use deer forages into ≥ 2 gram samples (i.e. consisting of

multiple individual plants representative of each treatment plot) and sent them to the

Mississippi State University Animal Nutrition Laboratory (MSUANL) for crude protein

analysis (Haufler & Servello 1996, Wheat 2015). The MSUANL ground forage samples

through a 1.0 mm screen in The Wiley Mill® (Thomas Scientific™, Swedesboro NJ,

USA) and analyzed each sample for nitrogen content using the Kjeldahl procedure.

During the Kjeldahl procedure, dried samples were digested in a sulfuric acid and catalyst

solution producing ammonia, which was then captured and titrated to determine nitrogen

concentration of samples (Haufler & Servello 1996, Jones et al. 2008, Sáez-Plaza et al.

2013). I multiplied nitrogen estimates by 6.25 to obtain crude protein estimates of forage

nutritional quality (Haufler & Servello 1996, Jones et al. 2008, Wheat 2015). I used

coefficient 6.25 because proteins on average are 16% nitrogen (i.e. 100/16 = 6.25)

(Haufler & Servello 1996).

I also used crude protein estimates of high and moderate-use deer forages to

calculate deer NCC. Deer NCC estimates were based on a comparative forage quality and

quantity index calculated via a constraints algorithm (Hobbs and Swift 1985). Therefore,

I made no assumptions that the index obtained absolute estimates of deer NCC but only

relative estimates useful for comparing treatment effects (Hobbs and Swift 1985, Stewart

et al. 2000, Lashley et al. 2011). Adult deer require 4-12% crude protein diets for general

metabolic maintenance whereas a lactating female nursing one fawn requires a 14%

crude protein diet during late summer (see Edwards et al. 2004, Jones et al. 2009).

Therefore, I determined if a target diet of 6% and 14% crude protein could be obtained from each treatment unit during July, 2014 and July, 2015 (Jones et al. 2009, Iglay et al.

2010b, Wheat 2015). I organized forages by treatment unit and descending crude protein value (Hobbs and Swift 1985). I then calculated mean diet quality and maximum forage

amount needed to acquire targeted crude protein for each treatment unit (Hobbs and Swift

1985). I assumed that daily dry matter intake for deer equaled 1.36 kg and used this

number to divide maximum forage needed to yield deer NCC in deer days per hectare

(Edwards et al. 2004, Jones et al. 2009 and references therein).

Additionally, I compiled biomass and nutritional quality data collected during this

study in 2011-2013 and calculated deer NCC for each treatment unit (Wheat 2015). I

averaged crude protein data from 2011-2013 treatments with data from years 2014 and

2015 to determine crude protein values of missing data. I used all data from 2011-2015 in

statistical analyses to better understand deer NCC across years.

Statistical Methods

I used repeated measures mixed model analysis of variance in program R to examine differences in biomass of plant classes, biomass of deer forages, and deer NCC between IC and PINE treatments, across years, and a treatment x year interaction (R Core

Team 2014). I used treatment (IC and PINE) as a fixed effect, stand and treatment unit (1| stand: treatment unit) as random effects, and year (2014 and 2015) as a repeated measure.

Plant classes included forbs, graminoids (grasses, sedges, and rushes), herbaceous vines, switchgrass, and woody plants (lianas, subshrubs, shrubs, and trees). I used Kenward-

Roger denominator degrees of freedom correction to avoid inflated type I error

(Gutzwiller and Riffell 2007, Wheat 2015). I calculated least square means for each 60

mixed model and performed post-hoc pairwise comparisons of the adjusted means to tease out significant interactions. I used Šidák correction of post-hoc analyses to avoid type-I error (Šidák 1967). I considered results significant at alpha = 0.10.

Results

I collected 201.9 kg (445.1 lbs.) of plant biomass from 336 biomass sampling

frames comprised of 164 identified species (Table 12). Total IC biomass in 2014 equated

to 5452 kg/ha whereas total PINE biomass in 2014 was 4374 kg/ha. In 2015, total IC biomass was 29608 kg/ha and PINE biomass was 25032 kg/ha.

Total Biomass

There were no differences in total biomass collected across treatment (F1, 5 = 0.36,

P = 0.57), year (F1, 5 = 2.09, P = 0.21), or treatment x year (F1, 5 = 0.01, P = 0.91) (Table

6). However, forb biomass differed across treatment (F1, 6 = 6.99, P = 0.04) and year (F1,

4 = 9.30, P = 0.04) (Table 6) and graminoid biomass differed across treatment (F1, 6 =

13.40, P = 0.01), year (F1, 4 = 10.66, P = 0.03), and treatment x year (F1, 4 = 56.99, P =

0.01) (Table 6). Forb biomass was greater in PINE treatments than IC treatments and

graminoid biomass was higher in IC treatments than PINE treatments. Total deer forage

biomass was not significantly different across treatments (Table 7). Overall, total biomass

produced in each treatment was similar but total forb and graminoid biomass was

significantly different across treatments and through time.

Interbed Biomass

Total interbed biomass did not differ by treatment (F1, 6 = 1.56, P = 0.26) but there

was a significant interaction across years (F1, 4 = 1.56, P = 0.04) (Table 8). A treatment

(F1, 6 = 14.21, P = 0.01) and year (F1, 5 = 11.70, P = 0.02) effect was detected for interbed

forb biomass (Table 8). Interbed forb biomass of IC treatments was smaller than interbed

forb biomass of PINE treatments (Table 9). Graminoid biomass in interbeds was

significantly different across treatments (F1, 6 = 14.63, P = 0.01), year (F1, 4 = 16.66, P =

0.02), and treatment x year (F1, 4 = 36.33, P = 0.01) (Table 8). Herbaceous vine biomass

in interbeds was significantly different across treatments (F1, 6 = 7.64, P =0.03) as was

woody biomass (F1, 6 = 12.97, P =0.01) (Table 8). I also detected year (F1, 5 = 4.63, P =

0.09) and treatment x year (F1, 5 = 4.69, P = 0.09) interactions for interbed woody

biomass (Table 8). High-use deer forages in interbeds differed by treatment (F1, 6 = 24.22,

P = 0.01) and low-use deer forages in interbeds differed by year (F1, 5 = 4.20, P =0.10)

(Table 8). Post hoc corrected analyses supported mixed model significance in all areas when averaged over years or treatments (Table 9). In general, interbeds of IC and PINE treatments were different in biomass composition but not in total biomass production.

Pine Bed Biomass

Pine beds of IC and PINE treatments were similar. I only detected a significant difference in graminoid biomass across years (F1, 5 = 15.07, P = 0.01) (Table 10).

Biomass composition of pine beds did not vary as much as biomass composition of

interbeds, but like interbeds, overall biomass productivity did not differ between

treatments. The only exception was that interbeds of IC treatments had more woody

biomass, primarily highbush blackberry (Rubus argutus L.), than did interbeds of PINE

treatments. Regardless, the compositional difference was not significant (Table 11).

White-Tailed Deer Nutritional Carrying Capacity

Across all years (2011-2015), I detected no treatment differences in deer NCC at a

6% crude protein diet (Table 7) but did detect a significant difference across years (F4, 20

= 13.52, P = <0.001) (Table 6). The treatment-averaged post hoc differences existed between year 2011-2014 (t = -6.69, P = < 0.001), 2012-2014 (t = -5.70, P = < 0.001),

2013-2014 (t = -5.21, P = < 0.001), and 2014-2015 (t = 5.64, P = < 0.001). In 2014, only

3 replicates were sampled due to logistical constraints, thus I suspect that this averaged

significant result across years is non-significant due to the change in sampling intensity.

A significant interaction across years (F4, 20 = 8.24, P = < 0.001) was detected

when testing for differences among 14% crude protein diet statistics (Table 6). This result was supported by treatment averaged post hoc analyses with significant interactions only between years 2011-2012 (t = 4.15, P = 0.01), 2011-2013 (t = 4.63, P = 0.01), 2011-2014

(t = 3.77, P = 0.01), and 2011-2015 (t = 5.41, P = < 0.001). Exactly like 2014, only 3

replicates were sampled in 2011 due to logistical constraints, thus it is likely that the significant result across years is again non-significant because of unequal sampling intensity. Neither IC nor PINE treatments supported a target nutritional diet of 14% crude

protein in all treatment units from 2012 to 2015. Overall, deer NCC in both treatments

was higher in years immediately following establishment disturbance (i.e. 2011, 2012)

and lower during later years (2013, 2015) but no treatment differences were observed

(Table 7). Deer NCC was similar in both IC and PINE treatments and decreased across

years (2011-2015).

Summary

IC and PINE treatment total biomass (interbeds + pine beds) did not differ, however, forbs and graminoids made up different ratios of the total biomass for each

treatment. Despite inherent differences between IC interbeds and PINE interbeds and

much like total biomass (interbeds + pine beds) insignificance, there were no functional

differences in maintenance level (6% crude protein) or lactation level (14% crude

protein) deer diet between treatments.

Discussion

Total Biomass

No differences in biomass of IC and PINE treatments were detected in this study,

continuing the trend observed in 2013 by Wheat (2015). My data and Wheat’s (2015)

data show that total vegetative productivity was similar between IC and PINE treatments.

However, our data also suggest that the total vegetative biomass composition of forbs and

graminoids were different (Wheat 2015). Unsurprisingly, I suspect that this difference

was the result of switchgrass and other graminoids in IC interbeds occupying the space of

woody and herbaceous plant biomass in PINE interbeds but remaining approximately

equal in mass. This result was expected because mechanical disturbance and herbicide

applications can promote graminoids and suppress woody plants (Jones et al. 2009).

Regardless of changing biomass composition and switchgrass harvest disturbance in

intercropped pine stands, it is likely that overall plant biomass productivity will remain

approximately equal in switchgrass intercropped pine plantations and non-intercropped

pine plantations.

Interbed Biomass

Distinct inconsistencies in treated versus untreated areas of IC treatment gives

insight into effects of switchgrass intercropping on deer. Wheat (2015) observed no

differences in total interbed biomass of intercropped and traditionally managed pine

stands but observed a reduction in forb and deer forage biomass in interbeds during

switchgrass establishment. Wheat (2015) found that in 2012, reestablishment of

switchgrass in intercropped areas resulted in treatment level biomass differences.

However, in 2013 these differences were no longer evident (Wheat 2015). My data show a similar trend in 2014 and 2015 with reductions in forbs and high-use deer forages in

interbeds with no treatment level differences. Previous work shows that herbicide

applications usually cause effects within years and not among years, ceasing to exert any

influence 1-3 years later (Jones et al. 2009). However, because intercropped interbeds

were treated each June with herbicides that reduced forb biomass and favored switchgrass

dominance, this effect could be perpetual (Wigley et al. 2002, Miller et al. 2010). Site

preparation and land management strategies employed to intercrop switchgrass, such as

herbicide applications, are known to reduce forb and high-quality deer forage biomass

(Jones et al. 2009). Jones et al. (2009) demonstrated that chemical weed control reduced forb biomass and constrained nutritional environments to abundant lower quality forages and few higher quality forages in southeastern U.S. pine plantations. The loss of forbs and high-use deer forages in interbeds is important to recognize because higher quality

forages are the limiting nutritional factor for deer populations (Jones et al. 2009, Lashley

et al. 2015b). Therefore, should the trend of diminished high-use forages in interbeds

persist, intercropping switchgrass in pine plantations may limit nutritional environments important for deer maintenance and lactation in the southeastern United States.

Pine Bed Biomass

Unlike interbeds, pine beds in IC treatment were similar to pine beds in PINE treatment, both in total biomass and biomass composition from 2012-2015 (Wheat 2015).

Pine beds of switchgrass intercropped pine stands may be critical areas for deer nutrition

due to the loss of high-use forages in interbeds. However, pine beds may lose the ability

to support deer populations as shade and competition increases; especially considering

that forbs and legumes can comprise >33% of available forage one to two years after tree

planting but can decrease to 7% when pine stands reach six to eight years old (Campbell et al. in press).

White-Tailed Deer Nutritional Carrying Capacity

My data suggest that there were no detrimental treatment effects on deer NCC.

Therefore, it is likely that deer populations dependent on switchgrass intercropped pine stands will not be affected by switchgrass intercropping. Nevertheless, it is possible that successional reductions in sunlight combined with interbed intercropping disturbance may diminish deer NCC in switchgrass intercropped pine plantations. Lashley et al.

(2011) observed a similar effect in hardwood communities when reductions in light

availability diminished deer diets in hardwood stands. Loss of forbs and the most

nutritious deer forages in interbeds could be detrimental to deer nutritional environments

going forward. This reduction in available nutrition could be more important than initially perceived when consideration is given to influence of light on plant nutritional

quality (Blair et al. 1983), loss of deer accessibility to forages in pine beds, diminished

quality of bedding site characteristics due to thermal losses, fawn predation risk in an

intercropped row structure (Chitwood et al. 2015), foraging conditions in shaded dense

litter layers (Edwards et al. 2004), and importance of high-quality deer forages for

lactation-level deer diets (Lashley et al. 2015b).

By comparison, Jones et al. (2009) found that chemical or chemical plus

mechanical treatments when combined with herbaceous plant control best increased deer

forage production through time. However, the treatment was not perpetual and the results

hinged upon availability of open canopy conditions (Jones et al. 2009). Jones et al. (2009)

also found high-use deer forages to be the primary driver of nutritional carrying capacity

treatment differences. Blair and Enghardt (1976) echoed importance of open canopy

conditions for deer forages. In other research, herbicide treatments directed at woody

plants increased herbaceous plants but over time the change diminished (Miller et al.

1995). Overall, timing, duration, and targeted component of plant communities impact

vegetative outcomes in intensively managed pine plantations. Switchgrass intercropping

management is no different, and thus with changes to management procedures, deer

nutritional environments can perhaps be maintained or enhanced.

Summary

Switchgrass intercropping did not adversely impact total plant biomass, total deer

forages, or total deer NCC, and it is likely that this trend could continue. However, forb and graminoid compositional changes resulting from switchgrass intercropping practices in interbeds did reduce high-use deer forages. It is uncertain what effect the loss of high-

use deer forages in interbeds will have at larger spatial scales and time spans, but it is 67

plausible that with fluctuating conditions, such as drought, population size, seasonality,

and canopy closure, negative effects could ensue. I recommend that long term effects of

switchgrass intercropping be investigated to further understand potential impacts while

simultaneously exploring management alternatives to maintain availability of high-use

deer forages.

Management Implications

Switchgrass establishment can be particularly difficult, resulting in unnecessary

disturbance if perceived to be incomplete (Schmer et al. 2006, Garland 2008, Albaugh et

al. 2012, Wheat 2015). Recurrent herbicide applications promote switchgrass, but they

also reduce availability of important plant species needed for deer diets. Managers should

seek alternatives to intense switchgrass establishment and allow for patchy mosaics of

switchgrass. Using alternative herbicide applications that do not eliminate high-use deer

forages may be a simple and effective way to mitigate the loss of valuable deer forage

plants.

Table 6 Total vegetative biomass mixed model ANOVA results.

Plant Class Effect MS F-statistic P-value Total Biomass Treatment 2257917 0.36 0.57 Year 12984503 2.09 0.21 Treatment*Year 85074 0.01 0.91 Forbs Treatment 89532 6.99 0.04* Year 119143 9.30 0.04* Treatment*Year 10409 0.81 0.42 Graminoids Treatment 173766 13.40 0.01* Year 138223 10.66 0.03* Treatment*Year 738928 56.99 0.01* Herbaceous Vines Treatment 2881 0.08 0.80 Year 6494 0.17 0.70 Treatment*Year 58489 1.52 0.27 Woody Treatment 1266272 0.21 0.66 Year 12411147 2.07 0.21 Treatment*Year 1047805 0.18 0.69 Deer Forages High-Use Treatment 333046 2.68 0.15 Year 82082 0.66 0.46 Treatment*Year 591 0.01 0.95 Moderate-Use Treatment 441695 0.26 0.63 Year 1895 0.01 0.98 Treatment*Year 566561 0.33 0.59 Low-Use Treatment 4786155 0.74 0.43 Year 15227148 2.36 0.18 Treatment*Year 215829 0.03 0.86 Nutritional CC 6% CP Diet Treatment 205932 0.60 0.47 Year 4636564 13.52 <0.001* Treatment*Year 574244 1.68 0.19 14% CP Diet Treatment 9418 0.47 0.52 Year 165857 8.24 <0.001* Treatment*Year 26133 1.30 0.30 Mixed model analysis of variance results showing significant plant class, white-tailed deer forage use, and white-tailed deer nutritional carrying capacity interactions across switchgrass intercropped and traditionally managed pine controls in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Mean square estimates (MS) are in kg/hectare. P-values signified by * were considered significant at α = 0.10.

Table 7 Total vegetative biomass collected by treatment and year.

Plant Class Year IC PINE P-value Total Biomass 2014 5452 (329) 4374 (590) 1.00 2015 7402 (2199) 6258 (1067) 0.99 Forbs 2014 314 (106) 740 (199) 0.15 2015 147 (36) 466 (118) 0.35 Graminoids 2014 1229 (431) 581 (104) 0.41 2015 1937 (319) 276 (33) 0.01* Herbaceous Vines 2014 37 (15.6) 198 (63) 0.94 2015 212 (179) 108 (43) 0.99 Switchgrass 2014 1049 (504) 0 (0) NA 2015 923 (404) 0 (0) NA Woody 2014 2824 (461) 2855 (617) 1.00 2015 4183 (1898) 5408 (1132) 0.99 Deer Forages High-Use 2014 1378 (381) 1753 (331) 0.77 2015 1048 (256) 1672 (253) 0.64 Moderate-Use 2014 1515 (431) 1564 (485) 1.00 2015 1911 (1062) 1119 (287) 0.96 Low-Use 2014 2528 (275) 980 (222) 0.99 2015 4428 (2156) 3429 (923) 1.00 Nutritional CC 6% CP Diet 2011 869 (391) 1032 (100) 1.00 (deer days/ ha) 2012 1058 (245) 1529 (197) 1.00 2013 1449 (197) 1450 (243) 1.00 2014 2718 (626) 3710 (654) 0.96 2015 1557 (578) 1070 (249) 1.00 14% CP Diet 2011 507 (278) 395 (135) 1.00 (deer days/ ha) 2012 221 (124) 5 (3) 1.00 2013 143 (112) 7 (7) 1.00 2014 96 (72) 189 (34) 1.00 2015 28 (25) 0 (0) 1.00 Total vegetative biomass means and standard errors for significant plant class, white- tailed deer forage use, and white-tailed deer nutritional carrying capacity post hoc pairwise interactions across switchgrass intercropped (IC) and traditionally managed pine controls (PINE) in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Plant class and deer forages means and standard errors are in kg/hectare; nutritional carrying capacity means and standard errors are in deer days/hectare. P-values signified by * were considered significant at α = 0.10. 70

Table 8 Mixed model ANOVA results for interbeds.

Plant Class Effect MS F-statistic P-value Interbed Biomass Treatment 3755810 1.56 0.26 Year 19725181 8.20 0.04* Treatment*Year 10132476 4.21 0.10 Forbs Treatment 396614 14.21 0.01* Year 326448 11.70 0.02* Treatment*Year 24701 0.89 0.39 Graminoids Treatment 897105 14.63 0.01* Year 1021597 16.66 0.02* Treatment*Year 2228673 36.33 0.01* Herbaceous Vines Treatment 20244 7.64 0.03* Year 6895 2.60 0.17 Treatment*Year 6495 2.45 0.18 Woody Treatment 43349215 12.97 0.01* Year 15462383 4.63 0.10* Treatment*Year 15677104 4.70 0.09* Deer Forages High-Use Treatment 5067170 24.22 0.01* Year 43829 0.21 0.67 Treatment*Year 246 0.01 0.97 Moderate-Use Treatment 411990 1.74 0.24 Year 83695 0.35 0.58 Treatment*Year 166325 0.70 0.44 Low-Use Treatment 99762 0.02 0.88 Year 17985246 4.20 0.10* Treatment*Year 8976053 2.10 0.21 Mixed model analysis of variance results for interbed areas showing significant plant class, white-tailed deer forage use, and white-tailed deer nutritional carrying capacity interactions across switchgrass intercropped and traditionally managed pine controls in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Mean square estimates (MS) are in kg/hectare. P-values signified by * were considered significant at α = 0.10.

Table 9 Interbed vegetative biomass collected by treatment and year.

Plant Class Year IC PINE P-value Interbed Biomass 2014 4481 (368) 4822 (1151) 1.00 2015 4975 (726) 8851 (1987) 0.30 Forbs 2014 244 (43) 920 (210) 0.04* 2015 3.82 (1.99) 502 (167) 0.11 Graminoids 2014 2091 (803) 608 (154) 0.27 2015 3572 (660) 323 (86) 0.01* Herbaceous Vines 2014 1.41 (0.71) 163 (83) 0.09* 2015 0 (0) 77 (26) 0.64 Switchgrass 2014 2097 (1008) 0 (0) NA 2015 1373 (493) 0 (0) NA Woody 2014 46 (14) 3692 (767) 0.57 2015 27 (13) 7949 (2196) 0.01* Deer Forages High-Use 2014 141 (63) 2120 (600) 0.02* 2015 25 (18) 2059 (418) 0.01* Moderate-Use 2014 1188 (314) 1903 (445) 0.69 2015 1172 (312) 1480 (328) 0.99 Low-Use 2014 3094 (309) 1277 (407) 0.93 2015 3771 (715) 5267 (1896) 0.94 Interbed vegetative biomass means and standard errors for significant plant class, white- tailed deer forage use, and white-tailed deer nutritional carrying capacity post hoc pairwise interactions across switchgrass intercropped (IC) and traditionally managed pine controls (PINE) in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Plant class and deer forages means and standard errors are in kg/hectare. P-values signified by * were considered significant at α = 0.10.

Table 10 Mixed model ANOVA results for pine beds.

Plant Class Effect MS F-statistic P-value Pine Bed Biomass Treatment 60032268 2.75 0.16 Year 7910584 0.36 0.57 Treatment*Year 10744798 0.49 0.51 Forbs Treatment 15333 0.93 0.37 Year 16662 1.01 0.37 Treatment*Year 1067.9 0.06 0.81 Graminoids Treatment 23499 2.12 0.20 Year 166673 15.07 0.01* Treatment*Year 25112 2.27 0.19 Herbaceous Vines Treatment 13804 1.77 0.24 Year 400.7 0.05 0.83 Treatment*Year 20728 2.65 0.17 Woody Treatment 78643221 3.73 0.11 Year 15822427 0.75 0.42 Treatment*Year 6040357 0.29 0.61 Deer Forages High-Use Treatment 594021 2.22 0.19 Year 120358 0.45 0.54 Treatment*Year 5186 0.02 0.90 Moderate-Use Treatment 5035107 0.90 0.39 Year 93302 0.02 0.90 Treatment*Year 1300489 0.23 0.65 Low-Use Treatment 18218437 1.04 0.35 Year 12999418 0.74 0.43 Treatment*Year 3928250 0.22 0.66 Mixed model analysis of variance results for pine bed areas showing significant plant class, white-tailed deer forage use, and white-tailed deer nutritional carrying capacity interactions across switchgrass intercropped and traditionally managed pine controls in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Mean square estimates (MS) are in kg/hectare. P-values signified by * were considered significant at α = 0.10.

Table 11 Pine bed vegetative biomass collected by treatment and year.

Plant Class Year IC PINE P-value Pine Bed Biomass 2014 6424 (908) 3925 (599) 0.99 2015 9829 (4071) 3665 (1000) 0.44 Forbs 2014 383 (176) 560 (193) 0.94 2015 278 (69) 429 (143) 0.97 Graminoids 2014 367 (59) 554 (57) 0.41 2015 224 (57) 229 (61) 1.00 Herbaceous Vines 2014 73 (31) 233 (44) 0.43 2015 128 (67) 139 (69) 1.00 Switchgrass 2014 0 (0) 0 (0) NA 2015 0.23 (0.23) 0 (0) NA Woody 2014 5602 (931) 2018 (592) 0.95 2015 9199 (4011) 2868 (927) 0.39 Deer Forages High-Use 2014 2615 (701) 1387 (215) 0.78 2015 2072 (507) 1284 (129) 0.78 Moderate-Use 2014 1841 (713) 1224 (560) 1.00 2015 2649 (2046) 758 (268) 0.87 Low-Use 2014 1961 (859) 683 (39) 1.00 2015 5085 (3684) 1591 (805) 0.84 Pine bed vegetative biomass means and standard errors for significant plant class, white- tailed deer forage use, and white-tailed deer nutritional carrying capacity post hoc pairwise interactions across switchgrass intercropped (IC) and traditionally managed pine controls (PINE) in Kemper County, Mississippi, USA. Plant class “Graminoids” includes grasses, sedges, and rushes. Plant class “Woody” includes lianas, subshrubs, shrubs, and trees. Plant class and deer forages means and standard errors are in kg/hectare. P-values signified by * were considered significant at α = 0.10. No significant interactions were detected.

Figure 9 Ten Hectare Treatment Unit Sampling Design for Biomass Sampling

Example of 10 hectare treatment unit containing 50 m buffer area, midpoints, and 3 ≤ 50 m circular separation areas with (n=60) diversity and (n=24) biomass sampling frames at random locations. Interbeds and pine beds are oriented west to east.

Literature Cited

Blair, R. M., and H. G. Enghardt. 1976. Deer forage and overstory dynamics in a loblolly pine plantation. Journal of Range Management 29:104-108.

Blair, R. M., R. Alcaniz, and A. Harrell. 1983. Shade intensity influences the nutrient quality and digestibility of southern deer browse leaves. Journal of Range Management 36:257-264.

Boundy, B., S. W. Diegel, L. Wright, S. C. Davis. 2011. Biomass energy data book edition 4. ORNL/TM-2011/446. Oak Ridge National Laboratory: Oak Ridge, Tennessee, USA.

Brockerhoff, E. G., H. Jactel, J. A. Parrotta, C. P. Quine, J. Sayer. 2008. Plantation forests and biodiversity: oxymoron or opportunity? Biodiversity and Conservation 17:925-951.

Campbell, T. N., P. D. Jones, S. Demarais, and A. W. Ezell. 2011. Forage biomass production and deer carrying capacity in intensively established pine plantations at crown closure. Master’s Thesis. Mississippi State University. Mississippi State. Mississippi, USA.

Campbell, T. N., P. D. Jones, S. Demarais, and A. W. Ezell. In press a. Vegetation communities in intensively established loblolly pine plantations at crown closure. Journal of Forestry.

Chitwood, M. C., M. A. Lashley, J. C. Kilgo, K. H. Pollock, C. E. Moorman, C. S. DePerno. 2015. Do biological and bedsite characteristics influence survival of neonatal white-tailed deer? Public Library of Science One. DOI: 10.1371.

Dollar, J. G., S. Riffell, H. L. Adams, L. W. Burger Jr. 2014. Evaluating butterflies as surrogates for birds and plants in semi-natural grassland buffers. Journal of Insect Conservation 18:171-178.

Edwards, S. L., S. Demarais, B. Watkins, B. K. Strickland. 2004. White-tailed deer forage production in managed and unmanaged pine stands and summer food plots in Mississippi. Wildlife Society Bulletin 32:739-745.

Garland, C.D. 2008. Growing and harvesting switchgrass for ethanol production in Tennessee. Publication no. SP701-A. University of Tennessee Extension, Knoxville, Tennessee, USA.

Gee, K. L., M. D. Porter, S. Demarais, and F. C. Bryant. 1994. White-tailed deer: their foods and management in the cross timbers. A Samuel Roberts Noble Foundation Publication, Ardmore, Oklahoma, USA 3-9.

Gutzwiller, K. J., and S. K. Riffell. 2007. Using statistical models to study temporal dynamic of animal-landscape relations. Pages 93-118 in J. A. Bisonette and I. Storch, editors. Temporal dimensions of landscape ecology: wildlife responses to variable resources. Springer-Verlag, New York, New York, USA.

Hobbs, N. T., and D. M. Swift. 1985. Estimates of habitat carrying capacity incorporating explicit nutritional constraints. Journal of Wildlife Management 49:814-822.

Iglay, R. B., B. D. Leopold, D.A. Miller, and L. W. Burger Jr. 2010a. Effect of plant community composition on plant response to fire and herbicide treatments. Forest Ecology and Management 260:543-548.

Iglay, R. B., P. D. Jones, D. A. Miller, S. Demarais, B. D. Leopold, and L. W. Burger Jr. 2010b. Deer carrying capacity in mid-rotation pine plantations of Mississippi. Journal of Wildlife Management 74:1003-1012.

Jones, P. D., S. Demarais, B. K. Strickland, S. L. Edwards. 2008. Soil region effects on white-tailed deer forage protein content. Southeastern Naturalist 7:595-606.

Lashley, M. A., C. A. Harper, G. E. Bates, P. D. Keyser. 2011 Forage availability for white-tailed deer following silvicultural treatments in hardwood forests. The Journal of Wildlife Management 75:1467-1476.

Lashley, M. A., M. C. Chitwood, R. Kays, C. A. Harper, C. S. DePerno, C. E. Moorman. 2015a. Prescribed fire affects female white-tailed deer habitat use during summer lactation. Forest Ecology and Management 348:220-225.

Lashley, M. A., M. C. Chitwood, C. A. Harper, C. E. Moorman, C. S. DePerno. 2015b. Poor soils and density-mediated body weight in deer: forage quality or quantity? Wildlife Biology 21:213-219.

Miller, D. A., T. B. Wigley, K. V. Miller. 2009. Managed forests and conservation of terrestrial biodiversity in the southern United States. Journal of Forestry 107:197- 203.

Miller, J. H., B. R. Zutter, S. M. Zedaker, M. B. Edwards, R. A. Newbold. 1995. Early plant succession in loblolly pine plantations as affected by vegetation management. Southern Journal of Applied Forestry 19:109-126.

Miller, J. H., and K. V. Miller. 1999. Forest plants of the southeast and their wildlife uses. Southern Weed Science Society, Craftmaster Printers, Auburn, Alabama, USA. Print.

Mixon, M. R., S. Demarais, P. D. Jones, B. J. Rude. 2009. Deer forage response to herbicide and fire in mid‐rotation pine plantations. The Journal of Wildlife Management 73:663-668.

National Oceanic and Atmospheric Administration. 2010. Southern Regional Climate Center, climate normals. Available from http://www.srcc.lsu.edu/climate_normals.html. Accessed 22 October 2015.

Pawson, S. M., A. Brin, E. G. Brockerhoff, D. Lamb, T.W. Payn, A. Paquette, J. A. Parrotta. 2013. Plantation forests, climate change and biodiversity. Biodiversity and Conservation 22:1203-1227.

R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Riffell, S. K., J. Verschuyl, D. Miller, T. B. Wigley Jr. 2012. Potential biodiversity response to intercropping herbaceous biomass crops on forest lands. Journal of Forestry 110:42-47.

Roberts-Pichette, P., and L. J. Gillespie. 1999. Terrestrial vegetation biodiversity monitoring protocols. EMAN Occasional Paper Series Report No. 9, Ontario, Canada.

Rooney, T. P., and D. M. Waller. 2003. Direct and indirect effects of white-tailed deer in forest ecosystems. Forest Ecology and Management 181:165-176.

Sáez-Plaza, P., T. Michalowski, M. J. Navas, A. G. Asuero, S. Wybraniec. 2013. An overview of the Kjeldahl method of nitrogen determination. Part I. Early history, chemistry of the procedure, and titrimetric finish. Critical Reviews in Analytical Chemistry 43:178-223.

Schmer, M. R., K. P. Vogel, R. B. Mitchell, L. E. Moser, K. M. Eskridge, R. K. Perrin. 2006. Establishment stand thresholds for switchgrass grown as a bioenergy crop. Crop Science 46:157-161.

Šidák, Z. 1967. Rectangular confidence regions for the means of multivariate normal distributions. Journal of the American Statistical Association 62:626-633.

Stewart, K. M., T. E. Fulbright, and D. L. Drawe. 2000. White-tailed deer use of clearings relative to forage availability. The Journal of Wildlife Management 64:733-741.

Waller, D.M., and W.S. Alverson. 1997. The white-tailed deer: a keystone herbivore. Wildlife Society Bulletin 25:217–226.

Wear, D. N., and J. G. Greis. 2012. The southern forest futures project: Summary report. United States Forest Service, General Technical Report SRS-168, Asheville, North Carolina, USA.

Wheat, B. 2015. Plant community response to intercropping switchgrass in pine plantations of Mississippi. Master’s Thesis. Mississippi State University. Mississippi State. Mississippi, USA.

APPENDIX A

PLANT SPECIES DETECTED DURING STUDY IN KEMPER COUNTY, MS

Table 12 Plant species detected (2014-2015) in Kemper County MS, USA.

Family Scientific Name Common Use Detecte Class Name d Euphorbiaceae Acalypha gracilens A. Gray slender H BOTH F threeseed mercury Aceraceae Acer rubrum L. red maple H BOTH W Hippocastanaceae Aesculus pavia L. red buckeye L BOTH W Fabaceae Albizia julibrissin Durazz. silktree U BOTH W Liliaceae Allium canadense L. meadow garlic N DIV F Liliaceae Allium sativum L. cultivated garlic N DIV F Asteraceae Ambrosia artemisiifolia L. annual ragweed H BOTH F Poaceae Andropogon virginicus L. broomsedge L BOTH G bluestem Fabaceae Apios americana Medik. groundnut U BOTH W Apocynaceae Apocynum cannabinum L. Indianhemp H BOTH F Araliaceae Aralia spinosa L. devil's N DIV W walkingstick Caryophyllaceae Arenaria lanuginosa (Michx.) spreading U BOTH F Rohrb. sandwort Asclepiadaceae Asclepias variegata L. redring N DIV F milkweed Aspleniaceae Asplenium platyneuron (L.) Britton, ebony N DIV F Sterns & Poggenb. spleenwort Asteraceae Baccharis halimifolia L. eastern L BOTH W baccharis Rhamnaceae Berchemia scandens (Hill) K. Koch Alabama N DIV W supplejack Asteraceae Bidens aristosa (Michx.) Britton bearded H BOTH F beggarticks Bignoniaceae Bignonia capreolata L. crossvine M BOTH W Urticaceae Boehmeria cylindrica (L.) Sw. smallspike false N DIV F nettle Poaceae Briza minor L. little N DIV G quakinggrass Poaceae Bromus secalinus L. rye brome N DIV G Verbenaceae Callicarpa americana L. American H BOTH W beautyberry Convolvulaceae Calystegia catesbeiana Pursh Catesby's false M BOTH H bindweed Bignoniaceae Campsis radicans (L.) Seem. ex trumpet creeper M BOTH W Bureau Cyperaceae Carex annectens (E.P. Bicknell) yellowfruit L BOTH G E.P. Bicknell sedge Cyperaceae Carex complanata Torr. & Hook. hirsute sedge L BOTH G Cyperaceae Carex glaucodea Tuck. ex Olney blue sedge L BOTH G Cyperaceae Carex grayi Carey Gray's sedge L BOTH G

Table 12 (Continued)

Cyperaceae Carex jamesii Schwein. James' sedge L BOTH G Cyperaceae Carex longii Mack. Long's sedge L BOTH G Cyperaceae Carex lurida Wahlenb. shallow sedge L BOTH G Juglandaceae Carya glabra (Mill.) Sweet pignut hickory L BOTH W Juglandaceae Carya ovalis (Wangenh.) Sarg. red hickory L BOTH W Juglandaceae Carya tomentosa (Lam.) Nutt. mockernut L BOTH W hickory Ulmaceae Celtis laevigata Willd. sugarberry N DIV W Rubiaceae Cephalanthus occidentalis L. common N DIV W buttonbush Caryophyllaceae Cerastium glomeratum Thuill. sticky N DIV F chickweed Fabaceae Chamaecrista fasciculata (Michx.) partridge pea N DIV F Greene Poaceae Chasmanthium sessiliflorum (Poir.) longleaf L BOTH G Yates woodoats Oleaceae Chionanthus virginicus L. white fringetree N DIV W Apiaceae Cicuta maculata L. spotted water N DIV F hemlock Asteraceae Cirsium horridulum Michx. yellow thistle U BOTH F Ranunculaceae Clematis virginiana L. devil's darning U BOTH W needles Fabaceae Clitoria mariana L. Atlantic N DIV F pigeonwings Commelinaceae Commelina virginica L. Virginia N DIV F dayflower Asteraceae Conoclinium coelestinum (L.) DC. blue mistflower H BOTH F Asteraceae Conyza canadensis (L.) Cronquist Canadian H BOTH F horseweed Asteraceae Coreopsis major Walter greater tickseed H BOTH F Asteraceae Coreopsis tripteris L. tall tickseed N DIV F Rosaceae Crataegus marshallii Eggl. parsley N DIV W hawthorn Rosaceae Crataegus uniflora Münchh. dwarf hawthorn N DIV W Asteraceae Crepis pulchra L. smallflower U BOTH F hawksbeard Fabaceae Crotalaria sagittalis L. arrowhead L BOTH F rattlebox Euphorbiaceae Croton capitatus Michx. hogwort M BOTH F Cyperaceae Cyperus echinatus (L.) Alph. Wood globe flatsedge L BOTH G Cyperaceae Cyperus eragrostis Lam. tall flatsedge L BOTH G Cyperaceae Cyperus pseudovegetus Steud. marsh flatsedge L BOTH G Poaceae Danthonia sericea Nutt. downy N DIV G danthonia Fabaceae Desmodium canescens (L.) DC. hoary ticktrefoil M BOTH F

Table 12 (Continued)

Fabaceae Desmodium laevigatum (Nutt.) DC. smooth N DIV F ticktrefoil Fabaceae Desmodium rotundifolium DC. prostrate L BOTH F ticktrefoil Fabaceae Desmodium strictum (Pursh) DC. pine barren H BOTH F ticktrefoil Poaceae Dichanthelium aciculare (Desv. ex needleleaf M BOTH G Poir.) Gould & C.A. Clark rosette grass Poaceae Dichanthelium acuminatum (Sw.) tapered rosette M BOTH G Gould & C.A. Clark grass Poaceae Dichanthelium commutatum (Schult. variable M BOTH G ) Gould panicgrass Poaceae Dichanthelium dichotomum (L.) cypress M BOTH G Gould panicgrass Poaceae Dichanthelium laxiflorum (Lam.) openflower M BOTH G Gould rosette grass Poaceae Dichanthelium scoparium (Lam.) velvet panicum M BOTH G Gould Poaceae Digitaria ciliaris (Retz.) Koeler southern H BOTH G crabgrass Poaceae Digitaria sanguinalis (L.) Scop. hairy crabgrass H BOTH G Fabaceae Dioclea multiflora (Torr. & A. Gray) Boykin's H BOTH H C. Mohr clusterpea Rubiaceae Diodia virginiana L. Virginia H BOTH F buttonweed Dioscoreaceae Dioscorea villosa L. wild yam H BOTH H Ebenaceae Diospyros virginiana L. common L BOTH W persimmon Poaceae Echinochloa crus-galli (L.) P. barnyardgrass N DIV G Beauv. Cyperaceae Eleocharis obtusa (Willd.) Schult. blunt spikerush L BOTH G Asteraceae Elephantopus tomentosus L. devil's N DIV F grandmother Poaceae Eragrostis spectabilis (Pursh) Steud. purple L BOTH G lovegrass Asteraceae Erechtites hieraciifolius (L.) Raf. ex American M BOTH F DC. burnweed Asteraceae Erigeron annuus (L.) Pers. eastern daisy N DIV F fleabane Apiaceae Eryngium prostratum Nutt. ex DC. creeping N DIV F eryngo Asteraceae Eupatorium altissimum L. tall U BOTH F thoroughwort Asteraceae Eupatorium capillifolium (Lam.) dogfennel L BOTH F Small Asteraceae Eupatorium perfoliatum L. common H BOTH F boneset Asteraceae Eupatorium rotundifolium L. roundleaf N DIV F thoroughwort

Table 12 (Continued)

Asteraceae Eupatorium serotinum Michx. lateflowering H BOTH F thoroughwort Euphorbiaceae Euphorbia corollata L. flowering H BOTH F spurge Asteraceae Euthamia graminifolia (L.) Nutt. flat-top N DIV F var. graminifolia goldentop Cyperaceae Fimbristylis autumnalis (L.) Roem. slender fimbry N DIV G & Schult. Oleaceae Fraxinus americana L. white ash N DIV W Oleaceae Fraxinus pennsylvanica Marshall green ash N DIV W Rubiaceae Galium circaezans Michx. licorice H BOTH F bedstraw Rubiaceae Galium concinnum Torr. & A. Gray shining N DIV F bedstraw Rubiaceae Galium uniflorum Michx. oneflower N DIV F bedstraw Asteraceae Gamochaeta purpurea (L.) Cabrera spoonleaf L BOTH F purple everlasting Loganiaceae Gelsemium sempervirens (L.) W.T. evening N DIV W Aiton trumpetflower Geraniaceae Geranium carolinianum L. Carolina L BOTH F geranium Lamiaceae Hedeoma hispida Pursh rough false N DIV F pennyroyal Asteraceae Helenium flexuosum Raf. purplehead U BOTH F sneezeweed Asteraceae Helianthus divaricatus L. woodland M BOTH F sunflower Malvaceae Hibiscus lasiocarpos Cav. rosemallow M BOTH F Rubiaceae Houstonia purpurea L. Venus' pride N DIV F Clusiaceae Hypericum hypericoides (L.) Crantz St. Andrew's H BOTH F cross Clusiaceae Hypericum punctatum Lam. spotted St. L BOTH F Johnswort Clusiaceae Hypericum sphaerocarpum Michx. roundseed St. N DIV F Johnswort Aquifoliaceae Ilex decidua Walter possumhaw N DIV W Poaceae Imperata cylindrica (L.) P. Beauv. cogongrass N BOTH G Convolvulaceae Ipomoea pandurata (L.) G. Mey. man of the H BOTH H earth Juncaceae Juncus coriaceus Mack. leathery rush L BOTH G Juncaceae Juncus diffusissimus Buckley slimpod rush L BOTH G Juncaceae Juncus effusus L. common rush L BOTH G Juncaceae Juncus elliottii Chapm. Elliott's rush L BOTH G Juncaceae Juncus gymnocarpus Coville Pennsylvania L BOTH G rush

Table 12 (Continued)

Juncaceae Juncus marginatus Rostk. grassleaf rush L BOTH G Juncaceae Juncus polycephalos Michx. manyhead rush L BOTH G Juncaceae Juncus scirpoides Lam. needlepod rush L BOTH G Juncaceae Juncus tenuis Willd. poverty rush L BOTH G Cupressaceae Juniperus virginiana L. eastern N DIV W redcedar Acanthaceae Justicia americana (L.) Vahl American N DIV F water-willow Acanthaceae Justicia ovata (Walter) Lindau looseflower N DIV F water-willow Asteraceae Krigia caespitosa (Raf.) K.L. weedy N DIV F Chambers dwarfdandelion Asteraceae Krigia dandelion (L.) Nutt. potato N DIV F dwarfdandelion Fabaceae Kummerowia striata (Thunb.) Japanese clover N DIV F Schindl. Asteraceae Lactuca canadensis L. Canada lettuce H BOTH F Asteraceae Lactuca floridana (L.) Gaertn. woodland H BOTH F lettuce Lamiaceae Lamium purpureum L. purple N DIV F deadnettle Fabaceae Lespedeza procumbens Michx. trailing H BOTH F lespedeza Fabaceae Lespedeza repens (L.) W.P.C. creeping H BOTH F Barton lespedeza Fabaceae Lespedeza virginica (L.) Britton slender M BOTH F lespedeza Oleaceae Ligustrum sinense Lour. Chinese privet H BOTH W Hamamelidaceae Liquidambar styraciflua L. sweetgum L BOTH W Campanulaceae Lobelia siphilitica L. great blue N DIV F lobelia Poaceae Lolium perenne L. Italian ryegrass N DIV G subsp. multiflorum (Lam.) Husnot Caprifoliaceae Lonicera japonica Thunb. Japanese M BOTH W honeysuckle Caprifoliaceae Lonicera sempervirens L. trumpet N DIV W honeysuckle Onagraceae Ludwigia alternifolia L. seedbox H BOTH F Onagraceae Ludwigia palustris (L.) Elliott marsh seedbox L BOTH F Primulaceae Lysimachia ciliata L. fringed N DIV F loosestrife Liliaceae Maianthemum racemosum (L.) Link false lily of the N DIV F valley Asclepiadaceae Matelea gonocarpos (Walter) angularfruit N DIV H Shinners milkvine Scrophulariaceae Mecardonia acuminata (Walter) axilflower M BOTH F Small

Table 12 (Continued)

Fabaceae Medicago sativa L. alfalfa N DIV F Poaceae Melica mutica Walter twoflower N DIV G melicgrass Menispermaceae Menispermum canadense L. common U BOTH W moonseed Asteraceae Mikania scandens (L.) Willd. climbing M BOTH H hempvine Lamiaceae Monarda fistulosa L. wild bergamot N DIV F Moraceae Morus rubra L. red mulberry M BOTH W Vitaceae Nekemias arborea (L.) J. Wen & peppervine L BOTH W Boggan Cornaceae Nyssa sylvatica Marshall blackgum H BOTH W Oxalidaceae Oxalis dillenii Jacq. slender yellow M BOTH F woodsorrel Oxalidaceae Oxalis stricta L. common yellow L BOTH F oxalis Oxalidaceae Oxalis violacea L. violet L BOTH F woodsorrel Poaceae Panicum anceps Michx. beaked H BOTH G panicgrass Poaceae Panicum dichotomiflorum Michx. fall panicgrass M BOTH G Poaceae Panicum repens L. torpedo grass N DIV G Poaceae Panicum virgatum L. switchgrass L BOTH S Vitaceae Parthenocissus quinquefolia (L.) Virginia L BOTH W Planch. creeper Poaceae Paspalum dilatatum Poir. dallisgrass M BOTH G Poaceae Paspalum urvillei Steud. Vasey's grass L BOTH G Passifloraceae Passiflora incarnata L. purple L BOTH H passionflower Passifloraceae Passiflora lutea L. yellow N DIV H passionflower Scrophulariaceae Penstemon pallidus Small pale N DIV F beardtongue Polemoniaceae Phlox carolina L. thickleaf phlox N DIV F Polemoniaceae Phlox pilosa L. downy phlox N DIV F Solanaceae Physalis heterophylla Nees clammy M BOTH F groundcherry Solanaceae Physalis virginiana Mill. Virginia M BOTH F groundcherry Phytolaccaceae Phytolacca americana L. American M BOTH F pokeweed Pinaceae Pinus taeda L. loblolly pine N BOTH W Plantaginaceae Plantago virginica L. Virginia N DIV F plantain Polygalaceae Polygala sanguinea L. purple milkwort U BOTH F

Table 12 (Continued)

Polygonaceae Polygonum hydropiperoides Michx. swamp M BOTH F smartweed Buddlejaceae Polypremum procumbens L. juniper leaf M BOTH F Rosaceae Potentilla simplex Michx. common M BOTH F cinquefoil Lamiaceae Prunella vulgaris L. common M BOTH F selfheal Rosaceae Prunus americana Marshall American plum N DIV W Rosaceae Prunus serotina Ehrh. black cherry N DIV W Asteraceae Pseudognaphalium obtusifolium (L.) rabbit-tobacco L BOTH F Hilliard & B.L. Burtt Dennstaedtiaceae Pteridium aquilinum (L.) Kuhn western L BOTH F brackenfern Apiaceae Ptilimnium capillaceum (Michx.) herbwilliam H BOTH F Raf. Lamiaceae Pycnanthemum incanum (L.) hoary M BOTH F Michx. mountainmint Asteraceae Pyrrhopappus carolinianus (Walter) Carolina desert- N DIV F DC. chicory Fagaceae Quercus alba L. white oak N DIV W Fagaceae Quercus falcata Michx. southern red N DIV W oak Fagaceae Quercus laurifolia Michx. laurel oak N DIV W Fagaceae Quercus nigra L. water oak L BOTH W Fagaceae Quercus stellata Wangenh. post oak N DIV W Fagaceae Quercus velutina Lam. black oak L BOTH W Melastomataceae Rhexia mariana L. Maryland M BOTH F meadowbeauty Anacardiaceae Rhus copallinum L. winged sumac L BOTH W Anacardiaceae Rhus glabra L. smooth sumac M BOTH W Cyperaceae Rhynchospora globularis (Chapm.) globe L BOTH G Small beaksedge Cyperaceae Rhynchospora macrostachya Torr. tall horned L BOTH G ex A. Gray beaksedge Rosaceae Rosa carolina L. Carolina rose N DIV W Rosaceae Rosa multiflora Thunb. multiflora rose N DIV W Rosaceae Rubus argutus Link sawtooth H BOTH W blackberry Rosaceae Rubus flagellaris Willd. northern H BOTH W dewberry Rosaceae Rubus trivialis Michx. southern N DIV W dewberry Asteraceae Rudbeckia hirta L. blackeyed H BOTH F Susan Acanthaceae Ruellia strepens L. limestone wild M BOTH F petunia

Table 12 (Continued)

Gentianaceae Sabatia angularis (L.) Pursh rosepink N DIV F Gentianaceae Sabatia macrophylla Hook. largeleaf rose N DIV F gentian Poaceae Saccharum alopecuroides (L.) Nutt. silver L BOTH G plumegrass Lamiaceae Salvia lyrata L. lyreleaf sage M BOTH F Caprifoliaceae Sambucus nigra L. American black M BOTH W subsp. canadensis (L.) R. Bolli elderberry Apiaceae Sanicula canadensis L. Canadian M BOTH F blacksnakeroot Cyperaceae Scirpus cyperinus (L.) Kunth woolgrass L BOTH G Cyperaceae Scleria ciliata Michx. fringed nutrush L BOTH G Cyperaceae Scleria triglomerata Michx. whip nutrush L BOTH G Lamiaceae Scutellaria integrifolia L. helmet flower M BOTH F Asteraceae Silphium asteriscus L. starry N DIV F rosinweed Iridaceae Sisyrinchium angustifolium Mill. narrowleaf U BOTH F blue-eyed grass Smilacaceae Smilax bona-nox L. saw greenbrier H BOTH W Smilacaceae Smilax glauca Walter cat greenbrier H BOTH W Smilacaceae Smilax rotundifolia L. roundleaf H BOTH W greenbrier Solanaceae Solanum americanum Mill. American black H BOTH F nightshade Solanaceae Solanum carolinense L. Carolina M BOTH F horsenettle Asteraceae Solidago canadensis L. Canada M BOTH F goldenrod Asteraceae Solidago gigantea Aiton giant goldenrod N DIV F Asteraceae Solidago odora Aiton anisescented H BOTH F goldenrod Loganiaceae Spigelia marilandica (L.) L. woodland N DIV F pinkroot Fabaceae Strophostyles leiosperma (Torr. & slickseed N DIV F A. Gray) Piper fuzzybean Fabaceae Stylosanthes biflora (L.) Britton, sidebeak N DIV F Sterns & Poggenb. pencilflower Asteraceae Symphyotrichum dumosum (L.) G.L. rice button aster H BOTH F Nesom Asteraceae Symphyotrichum pilosum (Willd.) hairy white H BOTH F G.L. Nesom oldfield aster Asteraceae Symphyotrichum undulatum (L.) wavyleaf aster U BOTH F G.L. Nesom Fabaceae Tephrosia spicata (Walter) Torr. & spiked N DIV F A. Gray hoarypea Anacardiaceae Toxicodendron radicans (L.) Kuntze eastern poison H BOTH W ivy

Table 12 (Continued)

Apocynaceae Trachelospermum difforme (Walter) climbing M BOTH W A. Gray dogbane Commelinaceae Tradescantia ohiensis Raf. bluejacket H BOTH F Commelinaceae Tradescantia virginiana L. Virginia H BOTH F spiderwort Campanulaceae Triodanis biflora (Ruiz & Pav.) small Venus' N DIV F Greene looking-glass Campanulaceae Triodanis perfoliata (L.) Nieuwl. clasping Venus' N DIV F looking-glass Ulmaceae Ulmus alata Michx. winged elm H BOTH W Ericaceae Vaccinium arboreum Marshall farkleberry M BOTH W Ericaceae Vaccinium elliottii Chapm. Elliott's H BOTH W blueberry Valerianaceae Valerianella radiata (L.) Dufr. beaked N DIV F cornsalad Verbenaceae Verbena brasiliensis Vell. Brazilian M BOTH F vervain Asteraceae Verbesina alternifolia (L.) Britton wingstem M BOTH F ex Kearney Asteraceae Vernonia gigantea (Walter) Trel. giant ironweed M BOTH F Fabaceae Vicia sativa L. garden vetch M BOTH F Fabaceae Vicia tetrasperma (L.) Schreb. lentil vetch N DIV F Violaceae Viola sagittata Aiton arrowleaf violet M BOTH F Violaceae Viola sororia Willd. common blue M BOTH F violet Violaceae Viola triloba Schwein. three-lobe L BOTH F violet Vitaceae Vitis aestivalis Michx. summer grape M BOTH W Vitaceae Vitis rotundifolia Michx. muscadine H BOTH W Forage “Use” (white-tailed deer): H = High-Use, M = Moderate-Use, L = Low or no- use, N = Not Applicable, U = Unknown (Warren and Hurst 1984, Miller and Miller 1995, Jones et al. 2008, Jones et al. 2009, Gee et al. 1994, Wheat 2015, and Independent Biologist Survey). Sampling period when species were “Detected”: DIV = Diversity Sampling Only, BOTH = Diversity and Biomass Sampling. Plant “Class” or category: F = Forb, G = Graminoid (i.e. grass, sedge, rush), H = Herbaceous Vine, S = Switchgrass (Panicum virgatum L.), W = Woody (lianas, subshrubs, shrubs, trees.) Plant classes are determined by USDA Plant Database Descriptions