<%BANNER%>

Role of Herbicides in Longleaf Pine Flatwoods Restoration: Pine Growth, Understory Vegetation Response and Fate of Appli...


PAGE 1

ROLE OF HERBICIDES IN LONGLEAF PINE FLATWOODS RESTORATION: PINE GROWTH, UNDERSTORY VEGETATION RESPONSE AND FATE OF APPLIED HERBICIDES By SANJAYA RANASINGHE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Sanjaya Ranasinghe

PAGE 3

To my beloved parents, sisters and Jenny.

PAGE 4

ACKNOWLEDGMENTS I would like to thank my supervisory committee members, Drs. Shibu Jose (Chair), Debbie Miller and Alan Long, for their advice, time and support. In particular, I thank Dr. Shibu Jose for providing me the opportunity to attend graduate school. I would like to thank Dr. Debbie Miller for her advice concerning vegetation analysis procedures. Her knowledge on longleaf ecosystems and its understory vegetation, was a valuable asset. Dr. Long with his vast knowledge on fire ecology and longleaf ecosystems provided invaluable improvements to my research. I would also like to thank Dr. Craig Ramsey for his assistance with the project from its beginning. His contributions, from field work to data analysis are much appreciated. The efforts of Sara Merritt, Cathy Hardin, Leah McCue, Lewayne White, Chris Atkins and Tim Baxley in data collection work are much appreciated. I would also like to thank my fellow graduate students Diomides Zamora, Robert Wanvestrout, Maheteme Gebremedhin and Andrew Ruth for their friendship and support throughout my study. I would also like to thank the Florida Division of Forestry for providing land, field assistance and funding for the project. A special thank you to Jenny whose constant encouragement, love and support provided me with the strength to strive harder. I would also like to thank my parents and sisters for always believing in me and supporting me all throughout my education. iv

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 LONGLEAF PINE ECOSYSTEMS AND THEIR RESTORATION IN FLATWOODS........................................................................................................1 Longleaf Pine Ecosystems and Their Decline..............................................................1 Role of Fire, Mechanical and Chemical Site Preparation as Restoration Techniques..............................................................................................................2 Longleaf Pine Flatwoods Restoration...........................................................................5 Fate of Applied Herbicides...........................................................................................8 2 LONGLEAF PINE SEEDLING AND UNDERSTORY VEGETATION RESPONSE TO HERBICIDAL VEGETATION CONTROL AND FATE OF APPLIED HERBICIDES IN FLATWOODS...........................................10 Introduction.................................................................................................................10 Pine Flatwoods....................................................................................................10 Vegetation Control and Pine Response...............................................................11 Fate of Applied Herbicides..................................................................................12 Materials and Methods...............................................................................................14 Study Area Description.......................................................................................14 Experimental Design and Treatments.................................................................14 Measurements......................................................................................................15 Pine survival and growth..............................................................................15 Vegetation control assessment.....................................................................15 Herbicide dissipation assessment.................................................................16 Statistical Analysis..............................................................................................17 Pine survival and growth..............................................................................17 Vegetation control........................................................................................17 Herbicide persistence analysis.....................................................................18 Results.........................................................................................................................18 v

PAGE 6

Pine Survival.......................................................................................................18 Pine RCD, Height and SVI..................................................................................18 Vegetation Control..............................................................................................19 Herbicide Dissipation..........................................................................................20 Discussion...................................................................................................................21 Conclusion..................................................................................................................24 3 UNDERSTORY SPECIES DYNAMICS FOLLOWING HERBICIDAL PINE RELEASE TREATMENTS ON A LONGLEAF FLATWOODS SITE...................33 Introduction.................................................................................................................33 Materials and Methods...............................................................................................36 Study Area Description.......................................................................................36 Experimental Design and Treatments.................................................................37 Measurements......................................................................................................38 Statistical Analysis..............................................................................................38 Results.........................................................................................................................39 Species Composition and Diversity.....................................................................39 Understory Shrub Cover and Density..................................................................41 Wiregrass Cover..................................................................................................42 Discussion...................................................................................................................42 Conclusions.................................................................................................................46 4 SUMMARY AND CONCLUSIONS.........................................................................55 APPENDIX SPECIES LIST..............................................................................................57 LIST OF REFERENCES...................................................................................................59 BIOGRAPHICAL SKETCH.............................................................................................64 vi

PAGE 7

LIST OF TABLES Table page 2-1 Soil description at four depths..................................................................................30 2-2 Mean foliar cover for major understory species.......................................................31 2-3 Stem numbers and stem heights...............................................................................32 3-1 Mean species evenness and richness of three vegetation surveys............................51 3-2 Percentage change in foliar cover............................................................................52 3-3 Mean stem numbers.................................................................................................53 3-4 Importance values (IV).............................................................................................54 vii

PAGE 8

LIST OF FIGURES Figure page 2-1 Monthly rainfall data for year 2002.........................................................................25 2-2 Longleaf pine seedling survival and growth............................................................26 2-3 Stem volume index of longleaf seedlings................................................................27 2-4 Mean herbicide concentrations in part per billion....................................................28 2-5 Change in herbicide concentration over time across soil profile.............................29 3-1 Canonical correspondence analysis (CCA) ordination............................................48 3-2 Mean Shannon diversity index.................................................................................50 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ROLE OF HERBICIDES IN LONGLEAF PINE FLATWOODS RESTORATION: PINE GROWTH, UNDERSTORY VEGETATION RESPONSE AND FATE OF APPLIED HERBICIDES By Sanjaya Ranasinghe December, 2003 Chair: Shibu Jose Major Department: School of Forest Resources and Conservation Longleaf pine ecosystems, once dominant in the southeastern Coastal Plain, have been drastically reduced in acreage by excessive logging, land use changes and fire suppression. Increased understory competition is a primary cause for inadequate regeneration of longleaf pine. In pine flatwoods understory competition is heightened due to the dense shrub understory. Effective vegetation control is achieved in flatwoods with the use of intensive mechanical and chemical site preparation methods. Although these methods are successful in increasing pine growth, they are also detrimental to native plant diversity. The renewed interests in restoration of longleaf ecosystems are often guided by multiple objectives. Protecting the rich biodiversity of these unique ecosystems is of great importance. Therefore, intensive site preparation methods may not be appropriate in vegetation control in areas concerned with restoration. As an alternative low intensive site preparation followed by over the top application of herbicides is suggested. ix

PAGE 10

In this study the application of three commercially available herbicides (hexazinone, sulfometuron and imazapyr) following site preparation by single drum chopping and burning was investigated. Pine growth and survival were measured after one growing season. Effect on understory vegetation was assessed three and nine months after herbicide application. In addition, fate of applied herbicides was determined 16, 90 and 240 days after treatment at 0-15, 15-30 and 45 cm soil depths. A bioassay was used to determine herbicide concentrations in soil. Herbicide applications significantly reduced longleaf seedling survival compared to control with the exception of hexazinone. However, imazapyr treatment significantly increased pine seedling growth compared to the control. Sulfometuron treatment displayed a stunting effect by reducing seedling growth compared to control. Imazapyr treatment resulted in significant control of overall shrub species, with other treatments exhibiting no effect. Species diversity increased with imazapyr application compared to other treatments. However, sulfometuron significantly reduced species diversity compared to control. Hexazinone and sulfometuron dissipated to very low concentrations 240 days after treatment. Imazapyr was found in very low concentrations 16 days after treatment and indicated persistence, as it did not dissipate rapidly with time. x

PAGE 11

CHAPTER 1 LONGLEAF PINE ECOSYSTEMS AND THEIR RESTORATION IN FLATWOODS Longleaf Pine Ecosystems and Their Decline Longleaf pine (Pinus palustris Mill) ecosystems once occupied an estimated 37 million ha in the south and southeastern United States (Frost 1993). These forests dominated the Coastal Plain areas ranging from Virginia to Texas through central Florida (Croker 1979, Landers, Van Lear and Boyer 1995), occupying a variety of sites ranging from xeric sandhills to wet poorly drained flatwoods (Carter et al. 1998, Brockway and Outcalt 2000, Boyer 1990). The extent of longleaf pine ecosystems has greatly declined since European settlement. At present, they occupy less than 2 million ha, about 5% of the original acreage (Kelly and Bechtold 1990, Busby et al. 1995). Excessive logging in the early 20 th century cleared vast areas of virgin longleaf forests with minimal effort in regeneration. As the demand for timber increased, land that was once under longleaf pine was converted to other faster growing pine species such as slash pine (Pinus elliottii Engelm.) and loblolly pine (Pinus taeda L.). Conversion of forestland in to agriculture reduced the longleaf acreage further more. Throughout its natural range, longleaf ecosystems are faced with many disturbances such as periodic fires and hurricanes, which play a very important ecological role. Through these disturbances and varying site conditions, longleaf ecosystems are comprised of a mosaic of community types that sustains a variety of flora and fauna (Brockway and Outcalt 2000, Croker 1979). The canopy gaps and the open midstory 1

PAGE 12

2 created by these disturbances allows more sunlight to reach the forest floor, thus inducing understory vegetation growth. Longleaf ecosystems have one of the richest species diversities outside the tropics (Peet and Allard 1993). Although the overstory is dominated by one species, the understory is host to a plethora of plant species. The diversity among the herbaceous plants is the main contributor to its high biodiversity. The composition of the understory is site specific, but is mainly dominated by grass species. In the western Gulf Coastal Plain, the understory is comprised mainly of bluestem grasses. In Florida and along the Atlantic Coast wiregrass (Aristida beyrichiana) is dominant, with Aristida stricta occurring from central South Carolina through North Carolina (Peet 1993). One of the significant causes for the reduction of longleaf regeneration was the interruption of natural fire cycles in the understory. Role of Fire, Mechanical and Chemical Site Preparation as Restoration Techniques Prior to European settlement, periodic fires were a frequent phenomenon in longleaf ecosystems and were ignited by lightning and native Americans (Croker 1979). Such frequent fires promoted longleaf dominance limiting the less fire adapted hardwood species to more mesic sites. Fire also created a forest structure with open midstory and a savanna like understory. The understory, mainly comprised of grasses, provided the necessary fuels and facilitated the spread of these fires. Understanding the role of fire and the autecology of longleaf pine is vital for the restoration of this ecosystem. Longleaf pine is a very intolerant pioneer species (Landers et al. 1995) and does not compete well for site resources with other more aggressive species (Brockway and Lewis 1997). With the removal of fire, the less fire adapted shrub species were allowed to spread into the understory. The encroaching hardwoods compete

PAGE 13

3 for site resources and light with the longleaf seedlings and hinder their growth and regeneration. Research shows that simulating the natural fire regime by applying frequent growing season burns increases the growth and survival of longleaf pine seedlings (Grelen 1978, 1983). Longleaf seedlings undergo an extended stem-less phase without height initiation. This phase, also known as the grass stage, varies in length depending on site resources and competition and may last as long as 10-15 years. Increased competition from the shrubs hinders height initiation of regenerating longleaf seedlings. The shrub species also outcompete the native grass and herbaceous species for site resources such as light. Studies show a significant increase in species diversity and richness on sites under a frequent fire regime (Brockway and Lewis 1997). In addition, the dangers of catastrophic fires are increased with the encroachment of shrub and other hardwood species. The shrubs and hardwoods occupy the once open midstory and provide ladder fuels that could cause crown fires and damage entire forests (Brockway and Lewis 1997). In addition to competition for light some hardwood scrub species display allelopathy against native pines including longleaf pine and native herbaceous species (Richardson 1985). Compared to other pine species, longleaf pine is not a prolific seeder. Longleaf pine seeds require over three years for their physiological development (Pederson et al. 1999). Thus good seed crops are infrequent and may arise once every 4-5 years. The seeds are large and heavy and do not disperse a great distance (Landers et al. 1995). The short dispersal of the seeds prevents longleaf pine from colonizing and establishing in areas far from the seed source. Longleaf seeds require a seedbed of exposed mineral soil free of surface litter. Fire exclusion results in accumulation of forest litter that hinders

PAGE 14

4 proper germination of longleaf pine seeds (Croker 1975). The number of regenerating seedlings is further lowered through predation by non-native feral hogs. One study done on the impact of feral hogs demonstrated that exclusion of hogs resulted in 500 seedlings in the fenced area compared to 8 seedlings in non fenced areas (Lipscomb 1989). Due to its rapid and continuing decline, longleaf forests are considered an endangered ecosystem. Therefore, there is heightened interest in developing techniques to restore longleaf pine ecosystems. Better understanding of the silvicultural requirements of longleaf pine has enabled foresters and landowners to successfully establish and manage longleaf forests. With the alarming increase of endangered species, there is also great amount of interest in maintaining the biodiversity of these ecosystems. The use of prescribed fire has greatly enhanced longleaf restoration efforts. The use of prescribed fire is suggested prior to seed fall to improve seedbed conditions. In areas sensitive to burning, mechanical scarification methods are suggested (Croker 1975). Fire is also used effectively to eliminate hardwood shrubs thus releasing regenerating pines from vegetation competition. Unfortunately, prescribed fire may not be applicable in some sites. The use of fire in forests near residential or commercial properties bears the risk of damaging human lives and property. Prescribed fire may only be used under favorable weather conditions and proper authorization. Forest fires also increase the potential for soil erosion, and prescribed fire should be used with caution on sites prone to erosion. With concern for catastrophic crown fires, sites with heavy fuel loads should not be burned until fuel loads are reduced by other methods. Due to such limitations of prescribed fire, alternative vegetation control methods are used.

PAGE 15

5 Understory shrub control could also be achieved through mechanical site preparation and vegetation removal. Intensive site preparation such as disking, harrowing and bedding has been successful in temporarily reducing understory shrub vegetation. On sites concerned with understory restoration such as wiregrass establishment, intensive site preparation should be limited. Research shows that intensive site preparation may adversely effect wiregrass populations (Clewell 1989). Excessive soil disturbance through mechanical site preparation may also increase the risk of soil erosion. Chemical treatments in the form of herbicides have been widely used to control understory vegetation. Herbicides are found in many forms, which differ in their mode of action, and target species. Therefore, herbicides can be used at different stages of stand development to selectively control undesired species. Longleaf Pine Flatwoods Restoration Pine flatwoods constituted a major forest type in the southeastern Coastal Plain and occupied nearly 50% of the Florida peninsula (Davis 1967). The soils within pine flatwoods are poorly drained saturated sandy soils with seasonal flooding (Stout and Marion 1993). Historically, this landscape was dominated by longleaf and slash pine depending on site conditions. Longleaf pine dominated the upland and moderate to poorly drained sites, frequented by fires. Slash pine was limited to the wetter sites where fires were not prevalent (Stout and Marion 1993). Flatwoods ecosystems are also characterized as having relatively dense understories, typically comprised of shrubs such as gallberry (Ilex glabra), saw palmetto (Serenoa repens), runner oak (Quercus pumila), fetterbush (Lyonia lucida), blueberry (Vaccinum myrsinites) and hairy wicky (Kalmia hirsuta) (Huck 1987).

PAGE 16

6 The predominantly shrub understory amplifies the challenges faced by young longleaf seedlings due to the increased competition for site resources. Restoration techniques such as prescribed fire have been widely used to control understory shrub vegetation resulting in successful establishment of longleaf pine forests on many sites (Provencher et al 2001). However, vegetation control by prescribed fire is short-lived on flatwood sites due to the vigorous resprouting of the shrubs. Typical shrubs and herbaceous species of flatwoods sites have extensive rootstock systems that enable the vegetation to resprout rapidly following fire (Abrahamson 1984). Therefore, alternative restoration tools are required for successful establishment of longleaf pine on flatwood sites. Alternative vegetation control methods such as mechanical and chemical site preparation and pine release treatments have been widely studied on flatwood sites with slash pine (Kline et al 1994, Shiver et al 1991, Shiver et al 1990, Burger et al 1988). For example, Shiver et al (1990) reported significant growth increases in planted slash pine seedlings with site preparation methods such as chopping and bedding. Burger et al (1988) compared low intensity (chopping, burning) and high intensity (blading and harrowing) site preparation methods on flatwood sites and reported significant early growth gains with intensive site preparation. A majority of the flatwood studies used bedding as a standard site preparation method for better growth. Research done by Shiver et al (1991) on the use of picloram-triclopyr mixtures at site preparation rates reported 70 % control of gallberry, saw palmetto, wax myrtle and blueberry species. The study also reported 70% or greater control of blueberry species and staggerbush with imazapyr at site preparation rates. A similar study done by Kline et al (1994) using

PAGE 17

7 triclopyr and imazapyr mixtures at site preparation rates achieved good control over a majority of the shrub species. However, both studies found that even at high rates none of the herbicides yielded equal control on all flatwood species. The recent increase of interest in longleaf pine restoration is with multiple objectives such as biodiversity protection and enhancement of wildlife habitat. Therefore, the effect of restoration activities on native plant diversity is of concern. Mechanical site preparation such as bedding, disking and harrowing causes extensive soil disturbances. Although these methods are very effective in providing better growth gains among planted pine seedlings, they are detrimental to the native plant diversity. For instance, wiregrass a keystone species within longleaf pine ecosystems is very sensitive to ground disturbance. According to Clewell (1989), wiregrass is a poor seed producer and relies on vegetative propagation for its expansion. Wiregrass is also easily uprooted and once uprooted it is rarely successful in propagation. Therefore, intensive site preparation should not be used in areas concerned with understory restoration with wiregrass. Information regarding the effect of herbicides on the groundcover vegetation of natural flatwoods and sandhill communities is scarce (Litt et al 2001). An extensive literature review done by Litt et al (2001) on published research regarding herbicide effects on groundlayer vegetation found only 3 and 7 studies, respectively, on natural flatwoods and sandhill sites. A study done by Neary et al (1990) on a flatwood site in Florida found nearly 75% reduction in species richness following an intensive weed management regime with repeated application of sulfometuron, glyphosate and triclopyr treatments and mowing. Brockway et al (1998) studied the effects of hexazinone applied

PAGE 18

8 at 1.1 and 2.2 kg/ha with different application techniques on a sandhill site. They reported significant decrease in the cover of oak species and shrub control at the higher rate. Moreover, the higher rate application of hexazinone induced an 86% increase in biomass of wiregrass in the first growing season after treatment. With the broadcast application of hexazinone, a short-term reduction of forb species was also reported. The continuance of herbicide effects appears to be dependent on the rate and method of application. However, most treatment applications seem to cause only short-term changes in species diversity and composition. A long-term study of herbicide effect on understory plant diversity and richness was conducted by Boyd et al (1995) with one-time broadcast applications of imazapyr, glyphosate, and hexazinone. The reported results 7 years after treatment showed no significant herbicide effect on species diversity and richness. Fate of Applied Herbicides The application of herbicides creates a contamination risk of ground and surface water bodies. Extensive research has been conducted on the offsite movement of applied herbicides and the risk of contamination. Studies done by Michael and Neary (1993) with multiple herbicides on industrial forestry sites near a watershed reported levels of contamination less than the Health Advisory Levels (HAL). On some sites, herbicide concentrations exceeded HAL and drinking water standards when herbicide applications were directed onto surface water bodies, but did not persist for an extended period of time (Michael 2000, Neary et al. 1996). The groundwater level in flatwoods sites is prone to seasonal fluctuations and may rise as high as the soil surface level. On such sites, the application of herbicide may seem to involve high risk. In contrast, a study done on a

PAGE 19

9 flatwoods site by Neary et al. (1989), reported low levels of offsite movement and contamination of ground water using sulfometuron methyl. Phytotoxic effects of herbicide residue on non-target species are also of concern. Herbicide residue on agricultural fields has displayed phytotoxic effects on rotational crops (Brewster and Appleby 1985). Movement and persistence of forestry herbicides depend on weather, edaphic conditions, herbicide characteristics, method and time of application and site characteristics (Norris 1981). The majority of aforementioned research involved intensive mechanical and chemical site preparation treatments. Such methods may not be appropriate on most flatwoods sites with high water table and where multiple management objectives, including understory species richness and diversity, are of great concern. As a shrub control method in flatwoods, low-rate, over the top application of herbicide following less intensive site preparation could serve s an alternative to high intensity site preparation. In our study, three commonly used forestry herbicides, hexazinone, sulfometuron and imazapyr were used and their efficacy was examined. These herbicide treatments were applied following single drum chopping and prescribed burning. The three major research objectives addressed in this study were: Determine the extent of pine seedling response Quantify the fate of applied herbicides in the soil Determine the effect of applied herbicides on understory species diversity and composition

PAGE 20

CHAPTER 2 LONGLEAF PINE SEEDLING AND UNDERSTORY VEGETATION RESPONSE TO HERBICIDAL VEGETATION CONTROL AND FATE OF APPLIED HERBICIDES IN FLATWOODS Introduction Pine Flatwoods Pine flatwoods constituted a major forest type in the southeastern Coastal Plain and occupied nearly 50% of the Florida peninsula (Davis 1967). These landscapes were mostly dominated by longleaf pine (Pinus palustris Mill.) or slash pine (Pinus elliottii Engelm.) depending on site conditions. Longleaf pine dominated the upland and moderate to poorly drained sites, frequented by fires. Slash pine was limited to the wetter sites where fires were not prevalent (Stout and Marion 1993). Due to excessive logging and inadequate regeneration, the extent of longleaf pine ecosystems including flatwoods has greatly declined since European settlement. At present, longleaf pine forests occupy less than 2 million ha, about 5% of the original acreage (Kelly and Bechtold 1990, Busby et al. 1995). Due to the high demand for timber and the rise of plantation forestry, many longleaf dominated sites were converted to slash pine and loblolly pine (Pinus taeda L.) (Croker 1979). Longleaf pine was replaced by other pines because of the slow early growth characteristic of the species. It is a poor seed producer with infrequent seed crops (Boyer 1990, Pederson et al. 1999) and requires scarified seedbed with exposed mineral soil for adequate germination (Croker 1979). Once established, longleaf seedlings exhibit a slow growth phase with little to no height initiation for several years. This slow 10

PAGE 21

11 juvenile growth phase, also known as the grass stage is extended with increased competition for site resources (Haywood 2000, Jose et al. 2003). Such challenges are exacerbated on flatwoods sites due to the characteristic heavy understory of shrub species. These understories are typically comprised of shrubs such as gall berry (Ilex glabra L.), saw palmetto (Serenoa repens Bartr.), runner oak (Quercus pumila Walt.), fetterbush (Lyonia lucida Lam.), blueberry (Vaccinum myrsinites Chapman.) and hairy wicky (Kalmia hirsuta Walt.) (Huck 1987). Vegetation Control and Pine Response Prescribed fire has been widely used to control understory woody vegetation resulting in successful establishment of longleaf pine forests on many sites (Provencher et al. 2001). Vegetation control by prescribed fire is short-lived on flatwoods sites due to the vigorous resprouting of the vegetation. Typical shrubs and herbaceous species of flatwoods sites have extensive rootstock systems that enable the vegetation to resprout rapidly following fire (Abrahamson 1984). Alternative vegetation control methods such as mechanical and chemical site preparation and pine release treatments have been widely studied on flatwoods sites with slash pine (Kline et al. 1994, Shiver et al. 1990, 1991, Burger et al. 1988). For example, Shiver et al. (1990) reported significant growth increases in planted slash pine seedlings with site preparation methods such as chopping and bedding. Burger et al. (1988) compared low intensity (chopping, burning) and high intensity (burning, blading and harrowing) site preparation methods on flatwoods sites and reported significant early growth gains with intensive site preparation. A majority of the flatwoods studies used bedding as a standard site preparation method for better growth. Research done by Shiver et al. (1991) on the use of picloram-triclopyr mixtures at site preparation rates

PAGE 22

12 reported 70 % control of gallberry, saw palmetto, wax myrtle and blueberry species. The study also reported 70% or greater control of blueberry species and staggerbush with imazapyr at site preparation rates. A similar study done by Kline et al. (1994) using triclopyr and imazapyr mixtures at site preparation rates achieved good control over majority of the shrub species. However, both studies found that even at high rates none of the herbicides yielded equal control on all flatwoods species. The importance of site specific applications of herbicide treatments depending on the prevalent shrub species was emphasized by both studies. The aforementioned research on methods of vegetation control conducted on industrial forestry sites with slash pine may be applicable to the establishment of longleaf pine stands in flatwoods. However, the recent interests in restoring longleaf pine ecosystems are with multiple objectives in addition to increased seedling growth. In areas concerned with native biodiversity, sensitive plant populations and groundwater quality, intensive site preparation and high-rate herbicide applications may not be applicable. For example, wiregrass is reported as very sensitive to site disturbance. Therefore, on sites concerned with wiregrass regeneration intensive site preparation methods should be avoided (Clewell 1989). Fate of Applied Herbicides The offsite movement and persistence of applied herbicides and the risk of contamination has been widely studied. Movement and persistence of forestry herbicides depend on weather, edaphic conditions, herbicide characteristics, method and time of application and site characteristics (Norris, 1981). Studies done by Neary and Michael (1989) with sulfometuron methyl on a flatwoods site reported low levels of off-site movement and contamination of ground water. The study also reported that the high

PAGE 23

13 acidity (average pH = 4) of flatwoods sites further impedes the mobility of herbicides such as sulfometuron methyl. Michael and Neary (1993) reported on the off site movement of hexazinone, imazapyr, picloram and sulfometuron on industrial forestry sites in the south. Results from 23 studies showed levels of contamination less than the Health Advisory Levels (HAL) for all herbicides tested. On some sites, herbicide concentrations exceeded HAL and drinking water standards when herbicide applications were directed onto surface water bodies, but did not persist for an extended period. Phytotoxic effects of herbicide residue on non-target species are also of concern. Herbicide residue on agricultural fields has displayed phytotoxic effects on rotational crops (Brewster and Appleby 1985). Thus, persistent herbicides may impact seasonal variation in plant communities. Due to site specificity of herbicide behavior, care should be taken in estimating herbicide movement and persistence based on data from other sites (Michael 2000). Research on herbicidal vegetation control on flatwoods sites has focused primarily on pre-plant site preparation treatments. Information regarding the effect of over-the-top herbicide applications on longleaf pine seedlings and the understory vegetation of flatwoods sites is scarce. This study examined use of three commercially available herbicide treatments (hexazinone, sulfometuron and imazapyr) following low-intensive site preparation. The specific questions researched were: What is the response of longleaf pine seedling survival and growth to the applied treatments? What is the impact of herbicide application on major understory species foliar cover and density?

PAGE 24

14 What is the mobility and persistence of the applied herbicides in flatwoods soils? Materials and Methods Study Area Description This study was conducted on a flatwoods site at the Point Washington State Forest in Walton County, Florida (30 0 20.04 N, 86 0 4.22 W). Average annual high and low temperatures were 25.5 0 C and 12 0 C respectively. Annual precipitation was about 1640mm (2002) with most received in the late summer months (Figure 2-1). Soils were of low pH (pH <5) and sandy texture with low nutrient content (Table 2.1). The study area soils were mapped as sandy, siliceous, thermic aeric alaquods belonging to the Leon series, which is characterized by deep, poorly to very poorly drained soils. Soils of the flatwoods pinelands are formed on sandy quaternary formations derived from marine deposits (Stout and Marion 1993). Prior to study establishment the overstory was a planted slash pine stand with intermittent longleaf pine saplings. The average age of the stand was 26 years with a basal area of 1.85 m 2 and an average dbh of 19.1 cm. The understory was comprised mainly of species such as gallberry, saw palmetto, runner oak, dangleberry (Gaylussacia frondosa L.), hairy wicky, wiregrass (Aristida beyrichiana Trin.and Rupr.) and bluestem grasses. Dormant season fires were used on a three-year burn cycle to mitigate fuel build up. Experimental Design and Treatments Prior to site preparation, the overstory was harvested (August 2001). The harvest debris and the understory were roller chopped once and prescribed burned in October 2001. A randomized complete block design was used to examine the effects of herbicidal

PAGE 25

15 vegetation control methods on pine seedling growth and survival. The study incorporated six blocks with five treatment plots within each block. All treatment plots were 36.6m x 24.4 m, including a > 3m buffer strip between plots. In December 2001, one year old containerized longleaf pine seedlings were hand-planted at 3.1m x 1.8m spacing. Seedlings were planted in rows to facilitate the application of treatments. Each treatment plot included 100 seedlings amounting to 3000 seedlings for the entire study. In March 2002, four herbicide treatments [Sulfometuron (0.26 ai kg/ ha), Hexazinone (0.56 ai kg/ ha), Sulfometuron (0.26 ai kg/ ha) + Hexazinone (0.56 ai kg/ ha) mix, Imazapyr (0.21 ae kg/ ha)] were applied in a 1.2 m band over the top of seedlings via a knapsack sprayer. In each block, one treatment plot was kept herbicide free as a control. Measurements Pine survival and growth Pine survival was monitored six and 12 months (respectively, June 2002 and Dec 2002) after planting. Pine growth was measured after the first growing season (Dec 2002). Seedling height and root collar diameters (RCD) were measured as growth parameters on all planted seedlings. Seedling height was measured using a ruler, from soil surface to the top of the bud and RCD was measured using a digital caliper. Stem volume index (SVI) was calculated with the measured RCD and height data. Initial RCD of seedlings were also recorded prior to planting as baseline data. Within our study, post-planting burial was observed among the majority of the dead seedlings. In addition to growth measurements, the extent of post planting burial was estimated for each seedling. Vegetation control assessment A preliminary vegetation survey was conducted (June 2001) prior to overstory harvest and site preparation to assess the initial presence and percent cover of understory

PAGE 26

16 species. Following site preparation and herbicide application, two vegetation surveys were conducted. These surveys were done six and eleven months (June and Nov 2002) after pine planting. In each plot six randomly selected 1m 2 quadrats were sampled in the herbicide treated bands and were revisited for subsequent surveys. Percent cover was ocularly estimated for all species using the modified Daubenmire scale (Daubenmire 1959). Stem number and average stem heights were recorded for all shrub species. Herbicide dissipation assessment Soil was sampled over time in the hexazinone, sulfometuron and imazapyr plots in three blocks to monitor the persistence and movement of the herbicides. Soil samples were collected 16, 90 and 240 days after treatment. In each treatment plot, six random points within treated bands were sampled and the general location was revisited for subsequent sampling. Samples were collected at 0-15, 15-30 and 45-60cm depths. In each block, soil collected within a treatment plot was composited by depth to a single sample. Samples were kept frozen until the time of analysis. A bioassay was used to determine the herbicide concentration in soil samples (Ramsey et al 2004). Standard curves were established for each herbicide by subjecting brown top millet (Panicum ramosum L.) seeds to a known concentration of each herbicide. Three herbicide concentration ranges, 1-10 ppb at 1 ppb interval, 10-100 ppb at 10 ppb interval and 100-1000 ppb at 100 ppb interval were tested. The seeds were grown in a growth chamber with 16 hours of daylight (1800 mol m -2 ) at 301 0 C and 8 hours of darkness at 151 0 C. Growth chamber conditions were set as described by Hernandez-Sevillano et al. (2001). Plants were harvested after 13 days and dry plant weight was measured. The field soils

PAGE 27

17 was thawed at room temperature and brown top millet seeds were grown under the same growing conditions as the standards. Statistical Analysis Pine survival and growth Pine survival, RCD, height and SVI data after one growing season were analyzed using analysis of variance (ANOVA) within the framework of a randomized complete block design (RCBD) using SAS version 8.0 (SAS 2000). Significant treatment effects (=0.1) were separated using Duncans multiple range test. The initial RCD and the extent of post planting burial were used as covariates in the ANOVA model. Vegetation control The effect of treatments on stem counts and heights of major shrub species were analyzed using ANOVA for a randomized complete block design. Pre harvest uniformity of stem counts for each shrub species were tested prior to analysis. Only those shrub species with uniform stem counts prior to treatment application were considered for analysis. Percent control on shrub stem number and heights were calculated. Significant differences (=0.1) between treatments were detected by Duncans multiple range test. ANOVA was not used in the analysis of percent cover data, as it did not conform to the assumption of normality. Therefore, the Kruskal-Wallis test was used with PROC NPAR1WAY as a nonparametric alternative to ANOVA. Pre harvest percent cover uniformity of shrubs, grasses, forbs and major understory species were tested. Treatment effects on overall percent cover changes in shrubs, grasses and forbs were analyzed for both survey dates. Five major understory species that showed pre harvest uniform percent cover were used to test for treatment effects on percent cover of individual species. Percent control was calculated for each vegetation class and major species.

PAGE 28

18 Treatment plots were compared to the untreated control plots to establish significant differences (=0.1). Herbicide persistence analysis A power model was developed to predict herbicide concentration (Y) using the plant dry weight (X) as follows: Y= aX b The quality of fit of the model was tested for each herbicide by calculating R 2 value. Dry whole plant weight from seeds grown in the herbicide contaminated field soil was used in the power model to predict the herbicide concentrations in the field soil. The predicted concentrations were analyzed using ANOVA for differences in concentrations between sampling days and depths. Duncans multiple range test was used to detect significant differences between days and depths at a significance level of =0.1. Results Pine Survival Analysis of variance revealed a significant treatment effect on pine seedling survival after one growing season (P<0.0001). The control and hexazinone treatments had the highest survival (85.3% and 85% respectively) and were significantly greater than sulfometuron (78.3%), sulfometuron+hexazinone mix (67.8%) and imazapyr (65.6%) treatments (Fig 2-2a). Pine RCD, Height and SVI Analysis of first growing season RCD data revealed a significant treatment effect on RCD growth (p<0.0001). Initial RCD and extent of post-planting burial were significant as covariates (p <0.0001). Imazapyr treatment yielded the highest RCD (12.17mm) and was significantly larger than all other treatments. The RCD in

PAGE 29

19 hexazinone (11.75mm) and sulfometuron-hexazinone mix (11.25mm) treatments showed significant differences between the two treatments but did not differ from the control (11.65mm). Sulfometuron treatment resulted in the lowest RCD growth (10.80mm)and was significantly lower than all other treatments (Fig 2-2b). Analysis of height data showed significantly taller seedlings in the imazapyr (1.35cm) and sulfometuron+hexazinone mix (1.29cm) compared to other treatments. Hexazinone (1.18cm) did not significantly affect seedling height compared to the control (1.10cm). Sulfometuron (1.04cm) again resulted in lower growth and had the lowest height compared to all other treatments (Fig 2-2c). The SVI showed a similar trend as RCD and height. Imazapyr treatment resulted in the largest mean SVI of 2.46cm 3 and was significantly greater than all other treatments. Sulfometuron (1.59cm 3 ) yielded lowest SVI compared to all other treatments. Hexazinone and sulfometuron+hexazinone mix treatment did not significantly affect SVI compared to the control (Fig 2-3). Vegetation Control Imazapyr resulted in significant overall reduction of the percent cover of shrubs in both 3 MAT and 9 MAT surveys (66% and 59% control respectively). None of the other herbicide treatments was effective in reducing the overall percent cover of shrubs. Overall, percent cover of grasses and forbs were not reduced by any of the treatments. Analysis of major flatwoods species showed no treatment effect on wiregrass and saw palmetto percent cover. Imazapyr was successful in significantly reducing the percent cover of gallberry in the June 2002 survey (78% control) and showed sustained control through November 2002 survey (70% control). Gallberry percent cover was unaffected by the other treatments except in sulfometuron+hexazinone mix treatment

PAGE 30

20 where it increased in both surveys. None of the treatments reduced the percent cover of runner oak by June 2002 survey. However runner oak exhibited a delayed response to hexazinone (37% control), imazapyr (69% control) and sulfometuron+hexazinone mix (46% control) treatments as its percent cover was significantly reduced in the Nov 2002 survey. Bluestem species were not adversely affected by the applied treatments. The percent cover of bluestem species increased in the sulfometuron treatment plots by the Nov 2002 survey. Shrub stem counts and heights were analyzed for gallberry, runner oak and saw palmetto species (table 2-3). Gallberry and runner oak stem counts and heights were unaffected by the herbicide treatments except for imazapyr. Imazapyr significantly reduced (58%) the amount of resprouting stems of gallberry in June 2002 survey. An increased amount of control of gallberry was found (79%) in Nov 2002 and a significant reduction in the number of runner oak stems (65%). The effect on stem heights followed a similar trend with only imazapyr yielding reduced stem heights. In the June 2002 survey, imazapyr significantly reduced the stem heights of gallberry and runner oak species (64% and 40% respectively). The significant reduction in stem heights of gallberry and runner oak species by imazapyr was sustained in the Nov 2002 survey (37% and 29% respectively). None of the herbicide treatments significantly affected the stem counts or heights of saw palmetto species. Herbicide Dissipation Three predictive power models were developed for imazapyr, sulfometuron and hexazinone with pseudo R 2 values of 0.89, 0.65 and 0.60 respectively. The analysis of the predicted concentrations for the herbicide contaminated soils revealed no significant difference among soil depths and days (Fig 2-4). Herbicide concentrations of the entire

PAGE 31

21 sampled soil profile were analyzed over the three sampling dates. Hexazinone and sulfometuron showed significant decrease in herbicide concentrations between 16 DAT and 240 DAT. Imazapyr concentrations were not significantly different among the three dates (Fig 2-5). Discussion Successful establishment, survival and improved growth of pine seedlings on flatwoods sites rely on the efficacy of applied vegetation control methods. The majority of the studies reporting improved survival and growth were associated with intensive site preparation and high rate pre-plant herbicide applications which provide significant release from competition stress (Shiver et al. 1991, Kline et al. 1994). Within our study, overall survival of seedlings was moderate at best including the control. Post planting burial observed among the majority of the dead seedlings may have contributed to the overall seedling mortality. Metcalfe and Cantrell (1986) studied the effect of sulfometuron and sulfometuron+hexazinone treatments on first year survival of longleaf seedlings planted on a sandy soil in Florida. In their study with higher rates of sulfometuron and sulfometuron+hexazinone mix treatments, no significant effect on seedling survival was found compared to the control. Our results are contrary to the above findings as sulfometuron and sulfometuron+hexazinone mix treatments significantly lowered seedling survival compared to the control. Although imazapyr and sulfometuron+hexazinone mix treatments resulted in lowest survival, they both significantly increased seedling height with imazapyr also increasing RCD growth. Our results reveal an evident trade off in benefits of using herbicide applications such as imazapyr against control, providing forest managers the choice between increased survival of pine seedlings or increased growth depending on their silvicultural objectives.

PAGE 32

22 Application of sulfometuron displayed a stunting effect on seedling growth by yielding the lowest RCD and height. Our findings agree with previous research done by Gjerstad et al. (1983) where they reported detrimental effects such as stunting of pines by sulfometuron in sandy soils. Analysis of understory vegetation revealed significant reduction in foliar cover of shrubs by imazapyr treatment. However, imazapyr did not reduce the foliar cover of grasses and forb species. Runner oak displayed a delayed response to hexazinone, sulfometuron+hexazinone mix and imazapyr treatments. The benefits of runner oak percent cover reduction may be realized the following growing seasons. Although imazapyr had poor control over grasses and forbs, it resulted in increased growth of longleaf pine seedlings by controlling the shrub component alone. This result suggests an important competitive relationship between the flatwoods understory shrub species and planted longleaf seedlings. Our observations reiterate the importance of effective shrub control for improved growth of longleaf seedlings on flatwoods sites. Imazapyr was the only treatment to effectively reduce the resprouting of major flatwoods shrub species such as gallberry and runner oak by reducing the stem number and their heights. Interestingly, none of the treatments reduced the percent cover of wiregrass. Thus, in flatwoods areas concerned with restoration of longleaf-wiregrass ecosystems, the tested herbicide treatments at the specified rates may be applied for longleaf pine establishment without any detrimental effects on wiregrass. Bioassays are widely used to detect herbicide concentrations in soil. With the existing limitations of analytical methods due to inefficient residue extraction methods, bioassays provide a good alternative (Hernandez-Sevillano et al. 2001). They are

PAGE 33

23 effectively used to detect biologically active levels of herbicide residue at lower concentrations in soil (Anon 2000). Therefore, a bioassay is best suited for our study due to the relatively low herbicide rates applied. A study done by Hollaway et al. (1999) comparing different detection methods reported that bioassays were more accurate than high performance liquid chromatography (HPLC) when used to detect residues at lower concentrations. Many studies have also reported on the efficacy of bioassays in detecting herbicide residue concentrations as small as 1 ppb (Stork and Hannah 1996, Hollaway et al. 1999, Hernandez-Sevillano et al. 2001). The three predictive models developed through the bioassays had relatively high pseudo R 2 values indicating good predictive strength. However, bioassay results did not show significant differences in herbicide concentrations between soil depths for all herbicides. Although nonsignificant, the results reveal a trend in herbicide movement as concentrations increase in lower depths with time (fig 2.4). Lack of significance may be attributed to the large amount of variation encountered within the bioassay. For future studies, additional field sampling and increased replications within the bioassay are suggested to account for high variation. Herbicide dissipation analysis over the entire sampled soil profile reveals a significant decrease in herbicide concentrations of hexazinone and sulfometuron between 16 DAT and 240 DAT. Imazapyr was persistent in the soil at low concentrations for an extended period of time. Soil adsorption of imazapyr is improved with increased organic matter and low pH in soil (Dickens and Wehtje 1986). The accumulation of debris from the site preparation burn and low pH of the site may have contributed to the low predicted imazapyr concentrations by increasing

PAGE 34

24 its adsorption to soil. Herbicide concentrations less than 50 ppb were found for all three herbicides 240 DAT. Conclusion Within the first growing season after treatment application imazapyr treatment resulted in a significant decrease in seedling survival. However, it resulted in greater overall growth by significantly increasing RCD and height of seedlings. Imazapyr also resulted in significant reduction of shrub foliar cover, stem density and stem heights. The decrease in understory dominance of shrubs with imazapyr induced increased growth of longleaf seedlings. Application of hexazinone, sulfometuron and sulfometuron+hexazinone mix treatments failed to yield significant vegetation control and improve overall pine seedling growth and survival. Sulfometuron resulted in lowest RCD growth and its application is not recommended on sandy flatwoods soils. Applied herbicides did not significantly impact the understory grasses and forb species. More importantly, none of the herbicide treatments was detrimental or augmented wiregrass foliar cover. However, with the decline of shrub dominance the potential for improvement in wiregrass cover is increased. Hexazinone and sulfometuron do not seem to persist for extended periods. Imazapyr was present in very low concentration soon after application and exhibited a slower dissipation. However all three herbicides dissipated to very low concentrations less than 75ppb, 240 days after treatment. Overall, imazapyr provided the best-desired results with significant increase in pine growth and better control of shrub species with no significant effect on other understory species.

PAGE 35

25 050100150200250300350400JanFebMarAprMayJunJulAugSepOctNovDecmonthRainfall (mm) Figure 2-1. Monthly rainfall data for year 2002 at Point Washington State Forest, Walton County, Florida.

PAGE 36

26 acbac20406080100imzhexasulfosulfo+ hexaConSurvival (%)(a) b ccdba10.010.511.011.512.012.5imzhexasulfosulfo+ hexaconRCD (mm)(b) b cacba0.40.81.21.6imzhexasulfosulfo+ hexaContreatmentHeight (cm)(c) Figure 2-2. Longleaf pine seedling survival and growth. (a) mean survival, (b) mean root collar diameter (RCD), (c) mean seedling height by treatment. Means followed by the same letter are not significantly different (=0.1). (imz: imazapyr; hexa: hexazinone; sulfo: sulfometuron; sulfo+hexa: sulfometuron+hexazinone mix; con: control)

PAGE 37

27 caa ba bb1.01.31.51.82.02.32.52.83.03.33.5imzhexasulfosulfo+hexacontreatmentSVI (cm3) Figure 2-3. Stem volume index of longleaf seedlings after one growing season by treatments. Means associated with the same letters are not significantly different (=0.1). (imz: imazapyr; hexa: hexazinone; sulfo: sulfometuron; sulfo+hexa: sulfometuron+hexazinone mix; con: control)

PAGE 38

28 \000\000\000\000)TjETEMC/P <>BDCQ q166.5 636 14.28 -48 reW* n1 1 1 scnBT/T3_15 1 Tf1.92 0 0 -11.52 165.12 633.6 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_15 1 Tf1.92 0 0 -11.52 167.04 633.5999 Tm(\000)TjETEMC/P <>BDCq166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 172.8 633.5999 Tm(\000\000)TjETEMC/P <>BDCQ q166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 165.12 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_15 1 Tf1.92 0 0 -11.52 167.0399 622.08 Tm(\000)TjETEMC/P <>BDCq166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 172.8 622.08 Tm(\000\000)TjETEMC/P <>BDCQ q166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 165.12 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_15 1 Tf1.92 0 0 -11.52 167.0399 610.5599 Tm(\000)TjETEMC/P <>BDCq166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 172.7999 610.5599 Tm(\000\000)TjETEMC/P <>BDCQ q166.5 636 14.28 -48 reW* nBT/T3_15 1 Tf1.92 0 0 -11.52 165.1199 599.0399 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 0.78 w 0 j 0 J 166.5 636 14.28 -48 reS0 0 0 scn230.28 650.22 14.22 -62.22 refEMC/P <>BDCq230.28 650.22 14.22 -62.22 reW* n1 1 1 scnBT/T3_16 1 Tf1.92 0 0 -11.52 228.48 656.64 Tm(\000\000\000\000)TjETEMC/P <>BDCQ q230.28 650.22 14.22 -62.22 reW* n1 1 1 scnBT/T3_16 1 Tf1.92 0 0 -11.52 228.48 645.12 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_16 1 Tf1.92 0 0 -11.52 230.4 645.1199 Tm(\000)TjETEMC/P <>BDCq230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 236.16 645.1199 Tm(\000\000)TjETEMC/P <>BDCQ q230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 228.48 633.5999 Tm()TjETEMC/P <>BDCQ BT/T3_16 1 Tf1.92 0 0 -11.52 230.4 633.5999 Tm(\000)TjETEMC/P <>BDCq230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 236.1599 633.5999 Tm(\000\000)TjETEMC/P <>BDCQ q230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 228.4799 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_16 1 Tf1.92 0 0 -11.52 230.3999 622.08 Tm(\000)TjETEMC/P <>BDCq230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 236.1599 622.08 Tm(\000\000)TjETEMC/P <>BDCQ q230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 228.4799 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_16 1 Tf1.92 0 0 -11.52 230.3999 610.5599 Tm(\000)TjETEMC/P <>BDCq230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 236.1599 610.5599 Tm(\000\000)TjETEMC/P <>BDCQ q230.28 650.22 14.22 -62.22 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 228.4799 599.0399 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 230.28 650.22 14.22 -62.22 reS0 0 0 scn294.78 661.5 14.22 -73.5 refEMC/P <>BDCq294.78 661.5 14.22 -73.5 reW* n1 1 1 scnBT/T3_17 1 Tf1.92 0 0 -11.52 293.76 668.16 Tm(\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_17 1 Tf1.92 0 0 -11.52 295.6799 656.6399 Tm(\000 \000\000)TjETEMC/P <>BDCQ q294.78 661.5 14.22 -73.5 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 293.7599 645.1199 Tm()TjETEMC/P <>BDCQ BT/T3_17 1 Tf1.92 0 0 -11.52 295.6799 645.1199 Tm(\000 \000\000)TjETEMC/P <>BDCQ q294.78 661.5 14.22 -73.5 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 293.7599 633.5999 Tm()TjETEMC/P <>BDCQ BT/T3_17 1 Tf1.92 0 0 -11.52 295.6799 633.5999 Tm(\000 \000\000)TjETEMC/P <>BDCQ q294.78 661.5 14.22 -73.5 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 293.7599 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_17 1 Tf1.92 0 0 -11.52 295.6798 622.08 Tm(\000 \000\000)TjETEMC/P <>BDCQ q294.78 661.5 14.22 -73.5 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 293.7599 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_17 1 Tf1.92 0 0 -11.52 295.6798 610.5599 Tm(\000 \000\000)TjETEMC/P <>BDCQ q294.78 661.5 14.22 -73.5 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 293.7598 599.0399 Tm(\000\000\000\000 \000\000\000\000 \000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q244.5 633.72 15 -45.72 reW* n1 1 1 scnBT/T3_19 1 Tf1.92 0 0 -11.52 243.84 633.6 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_19 1 Tf1.92 0 0 -11.52 245.76 633.5999 Tm(\000)TjETEMC/P <>BDCq244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 251.52 633.5999 Tm(\000\000)TjETEMC/P <>BDCQ q244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 243.84 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_19 1 Tf1.92 0 0 -11.52 245.7599 622.08 Tm(\000)TjETEMC/P <>BDCq244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 251.52 622.08 Tm(\000\000)TjETEMC/P <>BDCQ q244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 243.84 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_19 1 Tf1.92 0 0 -11.52 245.7599 610.5599 Tm(\000)TjETEMC/P <>BDCq244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 251.5199 610.5599 Tm(\000\000)TjETEMC/P <>BDCQ q244.5 633.72 15 -45.72 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 243.8399 599.0399 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 244.5 633.72 15 -45.72 reS0 0 0 scn309 635.22 14.28 -47.22 refEMC/P <>BDCq309 635.22 14.28 -47.22 reW* n1 1 1 scnBT/T3_20 1 Tf1.92 0 0 -11.52 307.2 645.12 Tm(\000\000\000\000)TjETEMC/P <>BDCQ q309 635.22 14.28 -47.22 reW* n1 1 1 scnBT/T3_20 1 Tf1.92 0 0 -11.52 307.2 633.6 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_20 1 Tf1.92 0 0 -11.52 309.1199 633.5999 Tm(\000)TjETEMC/P <>BDCq309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 314.8799 633.5999 Tm(\000\000)TjETEMC/P <>BDCQ q309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 307.1999 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_20 1 Tf1.92 0 0 -11.52 309.1199 622.08 Tm(\000)TjETEMC/P <>BDCq309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 314.8799 622.08 Tm(\000\000)TjETEMC/P <>BDCQ q309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 307.1999 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_20 1 Tf1.92 0 0 -11.52 309.1199 610.5599 Tm(\000)TjETEMC/P <>BDCq309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 314.8799 610.5599 Tm(\000\000)TjETEMC/P <>BDCQ q309 635.22 14.28 -47.22 reW* nBT/T3_20 1 Tf1.92 0 0 -11.52 307.1999 599.0399 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 309 635.22 14.28 -47.22 reS0 0 0 scn195 591 14.28 -3 refEMC/P <>BDCq195 591 14.28 -3 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 193.92 599.04 Tm(\000\000\000\000 \000\000\000\000 \000\000\000\000 aaaaaaaaa0501001502002507.522.552.5Concentration (ppb) \000 16 DAT \000 \000 \000 hexazinone 90 DAT 240 DAT 90 DAT 240 DAT 90 DAT 240 DAT \000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_17 1 Tf0 Tc 1.92 0 0 -11.52 167.04 449.28 Tm(\000 \000\000)TjETEMC/P <>BDCQ q165.78 459.6 14.22 -24.78 reW* nBT/T3_17 1 Tf1.92 0 0 -11.52 165.12 437.76 Tm(\000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q230.28 460.32 14.22 -25.5 reW* n1 1 1 scnBT/T3_16 1 Tf1.92 0 0 -11.52 228.48 449.28 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_16 1 Tf1.92 0 0 -11.52 230.4 449.28 Tm(\000)TjETEMC/P <>BDCq230.28 460.32 14.22 -25.5 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 236.16 449.28 Tm(\000\000)TjETEMC/P <>BDCQ q230.28 460.32 14.22 -25.5 reW* nBT/T3_16 1 Tf1.92 0 0 -11.52 228.48 437.76 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 230.28 460.32 14.22 -25.5 reS0 0 0 scn294 440.82 14.28 -6 refEMC/P <>BDCq294 440.82 14.28 -6 reW* n1 1 1 scnBT/T3_23 1 Tf1.92 0 0 -11.52 293.76 449.28 Tm(\000\000\000\000 \000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q180 453.6 15 -18.78 reW* n1 1 1 scnBT/T3_24 1 Tf1.92 0 0 -11.52 178.56 449.28 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_24 1 Tf1.92 0 0 -11.52 180.48 449.28 Tm(\000)TjETEMC/P <>BDCq180 453.6 15 -18.78 reW* nBT/T3_24 1 Tf1.92 0 0 -11.52 186.24 449.28 Tm(\000\000)TjETEMC/P <>BDCQ q180 453.6 15 -18.78 reW* nBT/T3_24 1 Tf1.92 0 0 -11.52 178.56 437.76 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 180 453.6 15 -18.78 reS0 0 0 scn244.5 456.6 14.28 -21.78 refEMC/P <>BDCq244.5 456.6 14.28 -21.78 reW* n1 1 1 scnBT/T3_19 1 Tf1.92 0 0 -11.52 243.84 460.8 Tm(\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_19 1 Tf1.92 0 0 -11.52 245.76 449.28 Tm(\000 \000\000)TjETEMC/P <>BDCQ q244.5 456.6 14.28 -21.78 reW* nBT/T3_19 1 Tf1.92 0 0 -11.52 243.84 437.76 Tm(\000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q308.28 486.6 15 -51.78 reW* n1 1 1 scnBT/T3_25 1 Tf1.92 0 0 -11.52 307.2 483.84 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_25 1 Tf1.92 0 0 -11.52 309.1199 483.8399 Tm(\000)TjETEMC/P <>BDCq308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 314.8799 483.8399 Tm(\000\000)TjETEMC/P <>BDCQ q308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 307.1999 472.3199 Tm()TjETEMC/P <>BDCQ BT/T3_25 1 Tf1.92 0 0 -11.52 309.1199 472.3199 Tm(\000)TjETEMC/P <>BDCq308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 314.8799 472.3199 Tm(\000\000)TjETEMC/P <>BDCQ q308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 307.1999 460.7999 Tm()TjETEMC/P <>BDCQ BT/T3_25 1 Tf1.92 0 0 -11.52 309.1199 460.7999 Tm(\000)TjETEMC/P <>BDCq308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 314.8799 460.7999 Tm(\000\000)TjETEMC/P <>BDCQ q308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 307.1999 449.2799 Tm()TjETEMC/P <>BDCQ BT/T3_25 1 Tf1.92 0 0 -11.52 309.1198 449.2799 Tm(\000)TjETEMC/P <>BDCq308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 314.8799 449.2799 Tm(\000\000)TjETEMC/P <>BDCQ q308.28 486.6 15 -51.78 reW* nBT/T3_25 1 Tf1.92 0 0 -11.52 307.1999 437.7599 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 308.28 486.6 15 -51.78 reS0 0 0 scn195 455.82 14.28 -21 refEMC/P <>BDCq195 455.82 14.28 -21 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 193.92 460.8 Tm(\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 195.84 449.28 Tm(\000)TjETEMC/P <>BDCq195 455.82 14.28 -21 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 201.6 449.28 Tm(\000\000 \000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q258.78 450.6 14.22 -15.78 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 257.28 449.28 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 259.2 449.28 Tm(\000)TjETEMC/P <>BDCq258.78 450.6 14.22 -15.78 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 264.9599 449.28 Tm(\000\000)TjETEMC/P <>BDCQ q258.78 450.6 14.22 -15.78 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 257.2799 437.76 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 258.78 450.6 14.22 -15.78 reS0 0 0 scn323.28 482.82 14.22 -48 refEMC/P <>BDCq323.28 482.82 14.22 -48 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 322.56 483.84 Tm(\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 324.4799 472.32 Tm(\000 \000\000)TjETEMC/P <>BDCQ q323.28 482.82 14.22 -48 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 322.5599 460.8 Tm()TjETEMC/P <>BDCQ BT/T3_21 1 Tf1.92 0 0 -11.52 324.4799 460.8 Tm(\000 \000\000)TjETEMC/P <>BDCQ q323.28 482.82 14.22 -48 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 322.5599 449.28 Tm()TjETEMC/P <>BDCQ BT/T3_21 1 Tf1.92 0 0 -11.52 324.4799 449.28 Tm(\000 \000\000)TjETEMC/P <>BDCQ q323.28 482.82 14.22 -48 reW* nBT/T3_21 1 Tf1.92 0 0 -11.52 322.5599 437.76 Tm(\000\000\000\000 aaaaaaaaa01020304050607.522.552.5Concentration (ppb) \000 \000 16 DAT \000 \000 imazapyr \000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_28 1 Tf0 Tc 1.92 0 0 -11.52 168.96 299.52 Tm(\000 \000\000)TjETEMC/P <>BDCQ q167.28 300.12 14.22 -17.22 reW* nBT/T3_28 1 Tf1.92 0 0 -11.52 167.04 288 Tm(\000\000\000\000 \000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_23 1 Tf1.92 0 0 -11.52 232.32 299.52 Tm(\000 \000\000)TjETEMC/P <>BDCQ q231 302.4 14.28 -19.5 reW* nBT/T3_23 1 Tf1.92 0 0 -11.52 230.4 288 Tm(\000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* n1 1 1 scnBT/T3_26 1 Tf1.92 0 0 -11.52 293.76 357.12 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_26 1 Tf1.92 0 0 -11.52 295.6799 357.12 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4399 357.12 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7599 345.6 Tm()TjETEMC/P <>BDCQ BT/T3_26 1 Tf1.92 0 0 -11.52 295.6799 345.6 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4399 345.6 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7599 334.08 Tm()TjETEMC/P <>BDCQ BT/T3_26 1 Tf1.92 0 0 -11.52 295.6799 334.08 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4399 334.08 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7599 322.56 Tm()TjETEMC/P <>BDCQ BT/T3_26 1 Tf1.92 0 0 -11.52 295.6798 322.56 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4398 322.56 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7598 311.04 Tm()TjETEMC/P <>BDCQ BT/T3_26 1 Tf1.92 0 0 -11.52 295.6798 311.04 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4398 311.04 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7598 299.52 Tm()TjETEMC/P <>BDCQ BT/T3_26 1 Tf1.92 0 0 -11.52 295.6798 299.52 Tm(\000)TjETEMC/P <>BDCq295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 301.4398 299.52 Tm(\000\000)TjETEMC/P <>BDCQ q295.5 368.4 14.28 -85.5 reW* nBT/T3_26 1 Tf1.92 0 0 -11.52 293.7598 288 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 295.5 368.4 14.28 -85.5 reS0 0 0 scn181.5 297.12 14.28 -14.22 refEMC/P <>BDCq181.5 297.12 14.28 -14.22 reW* n1 1 1 scnBT/T3_27 1 Tf1.92 0 0 -11.52 180.48 299.52 Tm(\000\000\000\000 \000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q245.28 295.62 15 -12.72 reW* n1 1 1 scnBT/T3_22 1 Tf1.92 0 0 -11.52 243.84 288 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 245.28 295.62 15 -12.72 reS309.78 300.12 14.22 -17.22 refEMC/P <>BDCq309.78 300.12 14.22 -17.22 reW* n1 1 1 scnBT/T3_29 1 Tf1.92 0 0 -11.52 309.12 311.04 Tm(\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_29 1 Tf1.92 0 0 -11.52 311.0399 299.52 Tm(\000 \000\000)TjETEMC/P <>BDCQ q309.78 300.12 14.22 -17.22 reW* nBT/T3_29 1 Tf1.92 0 0 -11.52 309.1199 288 Tm(\000\000\000\000 \000\000\000\000)TjETEMC/P <>BDCQ q195.78 293.4 14.22 -10.5 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 193.92 288 Tm(\000\000\000\000)TjETEMC/Shape <>BDCQ 195.78 293.4 14.22 -10.5 reS260.28 291.12 14.22 -8.22 refEMC/P <>BDCq260.28 291.12 14.22 -8.22 reW* n1 1 1 scnBT/T3_21 1 Tf1.92 0 0 -11.52 259.2 299.52 Tm(\000\000\000\000 \000\000\000\000 \000\000\000\000)TjETEMC/Shape <>BDCQ 324 286.62 14.28 -3.72 reS1 j 1 J 174 300.12 m174 300.9 l171.78 300.9 m177 300.9 l237.78 302.4 m237.78 303.9 l235.5 303.9 m240.78 303.9 l302.28 368.4 m302.28 368.4 l300 368.4 m305.28 368.4 l188.28 297.12 m188.28 302.4 l186 302.4 m191.28 302.4 l252 295.62 m252 301.62 l249.78 301.62 m255 301.62 l316.5 300.12 m316.5 302.4 l314.28 302.4 m319.5 302.4 l202.5 293.4 m202.5 293.4 l200.28 293.4 m205.5 293.4 l267 291.12 m267 296.4 l264.78 296.4 m270 296.4 l330.78 286.62 m330.78 288.9 l328.5 288.9 m333.78 288.9 lS0.06 w 156.78 384.9 m156.78 282.9 l154.5 282.9 m156.78 282.9 l154.5 294.12 m156.78 294.12 l154.5 305.4 m156.78 305.4 l154.5 316.62 m156.78 316.62 l154.5 327.9 m156.78 327.9 l154.5 339.9 m156.78 339.9 l154.5 351.12 m156.78 351.12 l154.5 362.4 m156.78 362.4 l154.5 373.62 m156.78 373.62 l154.5 384.9 m156.78 384.9 l156.78 282.9 m348.78 282.9 l156.78 280.62 m156.78 282.9 l220.5 280.62 m220.5 282.9 l285 280.62 m285 282.9 l348.78 280.62 m348.78 282.9 lSEMC/P <>BDCBT/TT4 1 Tf8.28 0 0 8.28 171.78 307.6804 Tm(aaaaaaaaa01002003004005006007008009007.522.552.5soil depth (cm)concentration (ppb) \000 16 DAT \000 \000 \000 \000 sulfometuron Figure 2-4. Mean herbicide concentrations in part per billion (ug kg -1 ) by depth. Values associated with the same letters are not significantly different (=0.1). Significance established with duncans multiple range test.

PAGE 39

29 \000\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_8 1 Tf1.92 0 0 -11.52 176.64 622.08 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.72 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.64 610.5599 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 599.0399 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.6399 599.0399 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 587.5199 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.6399 587.5199 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 575.9999 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.6399 575.9999 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 564.4799 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.6399 564.4799 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 552.9599 Tm()TjETEMC/P <>BDCQ BT/T3_8 1 Tf1.92 0 0 -11.52 176.6399 552.9599 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q175.5 624.72 18 -86.22 reW* nBT/T3_8 1 Tf1.92 0 0 -11.52 174.7199 541.4399 Tm(\000\000\000\000\000 \000\000\000\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* n1 1 1 scnBT/T3_9 1 Tf1.92 0 0 -11.52 255.36 668.16 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_9 1 Tf1.92 0 0 -11.52 257.2799 668.1599 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8799 668.1599 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3599 656.6399 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2799 656.6399 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8799 656.6399 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3599 645.1199 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2799 645.1199 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8799 645.1199 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3599 633.5999 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2798 633.5999 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8799 633.5999 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3599 622.0799 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2798 622.08 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8798 622.08 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3598 610.56 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2798 610.5599 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8798 610.5599 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3598 599.0399 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2798 599.0399 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8798 599.0399 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3598 587.5199 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2797 587.5199 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8797 587.5199 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3597 575.9999 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2797 575.9999 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8797 575.9999 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3597 564.4799 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2797 564.4799 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8797 564.4799 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3597 552.9599 Tm()TjETEMC/P <>BDCQ BT/T3_9 1 Tf1.92 0 0 -11.52 257.2797 552.9599 Tm(\000\000)TjETEMC/P <>BDCq257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 266.8796 552.9599 Tm(\000\000)TjETEMC/P <>BDCQ q257.22 678 18.78 -139.5 reW* nBT/T3_9 1 Tf1.92 0 0 -11.52 255.3596 541.4399 Tm(\000\000\000\000\000)TjETEMC/Shape <>BDCQ 257.22 678 18.78 -139.5 reS0 0 0 scn339.72 545.22 18 -6.72 refEMC/P <>BDCq339.72 545.22 18 -6.72 reW* n1 1 1 scnBT/T3_10 1 Tf1.92 0 0 -11.52 337.92 552.96 Tm(\000\000\000\000\000)TjETEMC/P <>BDCQ q339.72 545.22 18 -6.72 reW* n1 1 1 scnBT/T3_10 1 Tf1.92 0 0 -11.52 337.92 541.44 Tm(\000\000\000\000\000)TjETEMC/Shape <>BDCQ 339.72 545.22 18 -6.72 reS193.5 597.72 18.72 -59.22 refEMC/P <>BDCq193.5 597.72 18.72 -59.22 reW* n1 1 1 scnBT/T3_11 1 Tf1.92 0 0 -11.52 192 599.04 Tm(\000\000\000\000\000)TjETEMC/P <>BDCQ q193.5 597.72 18.72 -59.22 reW* n1 1 1 scnBT/T3_11 1 Tf1.92 0 0 -11.52 192 587.52 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_11 1 Tf1.92 0 0 -11.52 193.92 587.5199 Tm(\000\000)TjETEMC/P <>BDCq193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 203.52 587.5199 Tm(\000\000)TjETEMC/P <>BDCQ q193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 192 575.9999 Tm()TjETEMC/P <>BDCQ BT/T3_11 1 Tf1.92 0 0 -11.52 193.92 575.9999 Tm(\000\000)TjETEMC/P <>BDCq193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 203.52 575.9999 Tm(\000\000)TjETEMC/P <>BDCQ q193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 192 564.4799 Tm()TjETEMC/P <>BDCQ BT/T3_11 1 Tf1.92 0 0 -11.52 193.9199 564.4799 Tm(\000\000)TjETEMC/P <>BDCq193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 203.5199 564.4799 Tm(\000\000)TjETEMC/P <>BDCQ q193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 191.9999 552.9599 Tm()TjETEMC/P <>BDCQ BT/T3_11 1 Tf1.92 0 0 -11.52 193.9199 552.9599 Tm(\000\000)TjETEMC/P <>BDCq193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 203.5199 552.9599 Tm(\000\000)TjETEMC/P <>BDCQ q193.5 597.72 18.72 -59.22 reW* nBT/T3_11 1 Tf1.92 0 0 -11.52 191.9999 541.4399 Tm(\000\000\000\000\000)TjETEMC/Shape <>BDCQ 193.5 597.72 18.72 -59.22 reS0 0 0 scn276 606 18 -67.5 refEMC/P <>BDCq276 606 18 -67.5 reW* n1 1 1 scnBT/T3_12 1 Tf1.92 0 0 -11.52 274.56 610.56 Tm(\000\000\000\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* n1 1 1 scnBT/T3_12 1 Tf1.92 0 0 -11.52 274.56 599.04 Tm()TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_12 1 Tf1.92 0 0 -11.52 276.4799 599.0399 Tm(\000\000)TjETEMC/P <>BDCq276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 286.08 599.0399 Tm(\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 274.56 587.5199 Tm()TjETEMC/P <>BDCQ BT/T3_12 1 Tf1.92 0 0 -11.52 276.4799 587.5199 Tm(\000\000)TjETEMC/P <>BDCq276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 286.0799 587.5199 Tm(\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 274.5599 575.9999 Tm()TjETEMC/P <>BDCQ BT/T3_12 1 Tf1.92 0 0 -11.52 276.4799 575.9999 Tm(\000\000)TjETEMC/P <>BDCq276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 286.0799 575.9999 Tm(\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 274.5599 564.4799 Tm()TjETEMC/P <>BDCQ BT/T3_12 1 Tf1.92 0 0 -11.52 276.4799 564.4799 Tm(\000\000)TjETEMC/P <>BDCq276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 286.0799 564.4799 Tm(\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 274.5599 552.9599 Tm()TjETEMC/P <>BDCQ BT/T3_12 1 Tf1.92 0 0 -11.52 276.4798 552.9599 Tm(\000\000)TjETEMC/P <>BDCq276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 286.0798 552.9599 Tm(\000\000)TjETEMC/P <>BDCQ q276 606 18 -67.5 reW* nBT/T3_12 1 Tf1.92 0 0 -11.52 274.5598 541.4399 Tm(\000\000\000\000\000)TjETEMC/Shape <>BDCQ 276 606 18 -67.5 reS0 0 0 scn357.72 548.22 18.78 -9.72 refEMC/P <>BDCq357.72 548.22 18.78 -9.72 reW* n1 1 1 scnBT/T3_13 1 Tf1.92 0 0 -11.52 357.12 552.96 Tm(\000\000\000\000\000)TjETEMC/P <>BDCQ q357.72 548.22 18.78 -9.72 reW* n1 1 1 scnBT/T3_13 1 Tf1.92 0 0 -11.52 357.12 541.44 Tm(\000\000\000\000\000)TjETEMC/Shape <>BDCQ 357.72 548.22 18.78 -9.72 reS212.22 545.22 18 -6.72 refEMC/P <>BDCq212.22 545.22 18 -6.72 reW* n1 1 1 scnBT/T3_14 1 Tf1.92 0 0 -11.52 211.2 552.96 Tm(\000\000\000\000\000 \000\000\000\000\000 \000\000\000\000\000 )TjETEMC/P <>BDCQ 1 1 1 scnBT/T3_14 1 Tf1.92 0 0 -11.52 295.6799 564.4799 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q294 570 18.72 -31.5 reW* nBT/T3_14 1 Tf1.92 0 0 -11.52 293.76 552.9599 Tm()TjETEMC/P <>BDCQ BT/T3_14 1 Tf1.92 0 0 -11.52 295.6799 552.9599 Tm(\000\000 \000\000)TjETEMC/P <>BDCQ q294 570 18.72 -31.5 reW* nBT/T3_14 1 Tf1.92 0 0 -11.52 293.7599 541.4399 Tm(\000\000\000\000\000 \000\000\000\000\000 \000\000\000\000\000 aaaaaa babb050100150200250300hexazinonesulfometuronimazapyrherbicideconcentration (ppb) \000)TjETEMC/P <>BDCQ q356.22 683.22 4.5 -4.5 reW* n1 1 1 scnBT/T3_9 1 Tf0 Tc 1.92 0 0 -11.52 355.2 679.68 Tm(\000)TjETEMC/Shape <>BDCQ 0.72 w 356.22 683.22 5.28 -5.22 reSEMC/P <>BDCBT/TT4 1 Tf-0.0127 Tc 0.003 Tw 8.28 0 0 8.28 363.72 678.7804 Tm(16 DAT \000)TjETEMC/P <>BDCQ q356.22 669 4.5 -4.5 reW* n1 1 1 scnBT/T3_15 1 Tf0 Tc 0 Tw 1.92 0 0 -11.52 355.2 668.16 Tm(\000)TjETEMC/Shape <>BDCQ 356.22 669 5.28 -5.28 reSEMC/Shape <>BDC356.22 654.72 4.5 -4.5 refEMC/P <>BDCq356.22 654.72 4.5 -4.5 reW* n1 1 1 scnBT/T3_14 1 Tf0 Tc 0 Tw 1.92 0 0 -11.52 355.2 656.64 Tm(\000)TjETEMC/Shape <>BDCQ 356.22 654.72 5.28 -5.22 reSEMC/P <>BDC0.004 0 0 scnBT/TT4 1 Tf8.28 0 0 8.28 363.72 664.5604 Tm(90 DAT 240 DAT Figure 2-5. Change in herbicide concentration over time across soil profile (60 cm). Concentration values followed by different letter are significantly different from other dates for the respective herbicide. Significance declared at =0.1 using Duncans multiple range test.

PAGE 40

30 Table 2-1. Soil description at four depths of Point Washington State Forest, Walton County, Florida. Soil property Soil depths (cm) 0-15 15-30 30-45 45-60 Texture sandy sandy sandy sandy Soil pH 4.40 4.60 4.80 4.60 CEC (meq/100g) 3.80 2.20 2.50 2.50 Organic matter (%) 1.38 0.62 0.58 0.37 Phosphorus (kg ha -1 ) 16.67 4.44 4.44 4.44 Potassium (kg ha -1 ) 42.23 14.44 10.00 8.89 Magnesium (kg ha -1 ) 60.02 12.22 10.00 8.89 Calcium (kg ha -1 ) 330.11 50.01 44.46 33.33

PAGE 41

31 Table 2-2. Mean foliar cover for major understory species and the total foliar cover of vegetation types by treatment. Control Sulfometuron Hexazinone Sulfo+Hexa 1 Imazapyr Species Wiregrass June 2002 4.50 0.9 6.02 1.1 6.5 1.4 5.00 1.0 5.81 1.7 Nov 2002 11.5 2.5 13.972.7 13.65 2.6 12.86 2.1 12.40 1.4 Bluestem spp June 2002 0.96 0.2 1.05 0.3 1.08 0.2 1.00 0.2 0.50 Nov 2002 4.75 1.8 2.04 0.5 a 3.80 0.6 5.16 1.3 3.83 1.2 Gallberry June 2002 9.15 1.7 10.32 1.2 10.911.5 13.34 1.9 a 2.10 0.4 a Nov 2002 14.20 2.1 15.21 2.2 17.13 1.9 19.12 2.4 a 4.31 0.9 a Runner oak June 2002 19.09 4.6 21.60 4.9 12.55 3.3 10.89 1.5 6.12 1.6 Nov 2002 29.21 4.2 32.20 5.4 18.50 3.2 a 15.96 2.0 a 9.25 2.0 a Saw palmetto June 2002 11.22 4.1 8.04 2.3 8.70 2.3 7.94 4.1 11.30 3.6 Nov 2002 10.83 3.9 11.50 2.4 9.62 2.4 13.64 4.5 10.75 2.1 Veg-type All shrub June 2002 50.7913.6 44.5811.6 34.284.9 36.936.8 23.264.2 a Nov 2002 68.6510.1 67.9112.4 57.26.1 60.426.7 34.412.6 a All grasses June 2002 11.462.2 8.081.7 8.991.0 7.931.7 7.51.4 Nov 2002 26.002.7 18.633.9 24.925.8 20.653.6 22.483.4 All forbs June 2002 4.601.5 3.020.6 4.221.2 3.101.3 5.522.3 Nov 2002 8.364.7 7.101.0 6.221.2 6.502.3 12.324.1 Note: Values followed by letter a are significantly different from the control along rows at =0.1. 1 Sulfometuron+Hexazinone mix treatment

PAGE 42

Table 2-3. Stem numbers and stem heights for three understory shrub species by treatment. Shrub stem numbers (stems/ m 2 ) Species Control Sulfometuron Hexazinone Sulfo+Hexa 1 Imazapyr 3 MAT Gallberry 18.37 2.6 20.60 2.7 16.93 1.5 16.60 1.8 7.60 0.9 b Runner oak 31.40 5.3 31.29 5.9 23.14 3.9 20.32 2.5 17.62 3.6 Saw palmetto 2.20 0.5 2.45 0.4 2.00 0.4 2.55 0.6 2.90 0.6 9 MAT Gallberry 34.50 4.9 39.80 5.3 38.33 3.7 43.00 5.9 7.03 1.6 b Runner oak 52.07 7.8 59.20 12.1 37.57 6.4 36.57 4.7 18.04 3.3 b Saw palmetto 2.75 0.5 3.44 0.8 2.00 0.6 3.28 1.3 3.33 0.8 Shrub heights (cm) Species Control Sulfometuron Hexazinone Sulfo+Hexa 1 Imazapyr 3 MAT Gallberry 18.89 1.1 18.64 0.7 20.57 1.0 19.52 1.1 6.72 0.4 b Runner oak 22.92 1.9 a b 22.66 1.6 a 19.40 1.4 b 18.00 0.8 b 13.66 0.2c Saw palmetto 36.00 5.5 30.45 5.0 51.20 6.3 25.11 4.8 36.10 4.0 9 MAT Gallberry 25.24 1.5 26.71 1.4 28.16 1.5 27.66 1.6 15.85 1.6 b Runner oak 29.92 2.1a 26.37 2.2a b 26.00 1.7a b 23.82 1.3b 21.16 1.8 c Saw palmetto 24.81 4.2 26.11 5.9 35.75 9.0 30.85 6.1 30.25 5.4 32 Note: values followed by different letters are significantly different along rows (=0.1) 1 Sulfometuron+Hexazinone mix treatment

PAGE 43

CHAPTER 3 UNDERSTORY SPECIES DYNAMICS FOLLOWING HERBICIDAL PINE RELEASE TREATMENTS ON A LONGLEAF FLATWOODS SITE Introduction Prior to European settlement in the United States, the southeastern coastal plain was dominated by large tracts of longleaf pine forests, covering an estimated 37 million ha (Frost. 1993). These forests ranged from southeastern Virginia to eastern Texas through central Florida (Croker 1979, Landers et al. 1995), occupying a variety of sites ranging from xeric sandhills to wet poorly drained flatwoods (Boyer 1990, Carter et al.1998). At present, longleaf acreage has been reduced to less than 5% of its former range with an estimate of less than 2 million ha remaining (Kelly and Bechtold 1990, Busby et al. 1995). The longleaf pine range within the southeastern Coastal Plain is prone to many natural disturbances such as tropical storms and forest fire (Delcourt et al 1993). Coupled with these disturbances and varying site conditions, longleaf ecosystems included a mosaic of community types that sustained a high diversity of flora and fauna (Croker 1979, Brockway and Outcalt 2000). Longleaf pine ecosystems are well adapted to such natural calamities and were dependent on them to maintain their structure and functions. The frequent fires that swept through these forests prevented the encroachment of less fire-adapted hardwood species and maintained an open midstory. These natural disturbances coupled with competition intolerance of longleaf pine created 33

PAGE 44

34 a widely spaced overstory. Such structural characteristics of these forests facilitated increased light transmitance facilitating understory plant growth. Longleaf ecosystems have one of the richest species diversities outside the tropics (Peet and Allard 1993). Although the overstory is mostly monospecific, the understory is host to a plethora of plant species. The composition of the understory is site specific and mainly dominated by grass species (Outcalt 2000). In the western gulf Coastal Plain, the understory is comprises mainly of bluestem grasses. In Florida and along the Atlantic Coast wiregrass (Aristida beyrichiana Trin and Rupr) is dominant with Aristida stricta Michx., occupying from central South Carolina through North Carolina (Peet 1993). Some of the common forb species include Asclepias spp., Aster spp., Liatris spp., Rhexia spp., Carphephorus spp. and Eupatorium spp. among others. (Platt et al. 1988, Huck 1987). On the more mesic flatwoods sites the understory is comprised of a variety of shrub species such as gallberry (Ilex glabra L.), runner oak (Quercus pumila Walt.), fetterbush (Lyonia lucida Lam.), saw palmetto (Seronoa repens Bartr.), blueberry (Vaccinum myrsinites Chapman.) and hairy wicky (Kalmia hirsuta Walt.) (Huck. 1987). The forests of the southeastern Coastal Plain also include a large number of rare plants (Collins et al. 2001). With the decline of these forests and changes in land use and development activities, some of the rare and native plant populations were fragmented and eliminated reducing genetic diversity (Collins et al. 2001). Due to the rapid loss of these forests and the associated highly diverse understory, its restoration has gained importance. In restoration efforts focus on controlling the understory shrub dominance has gained importance. Shrub dominance increases resource competition stress of longleaf pine

PAGE 45

35 seedlings and suppresses the ground-layer understory vegetation. On flatwoods sites, the competition for site resources and shrub dominance is elevated due to the characteristic dense understory. Prescribed fire has been widely used to control understory vegetation resulting in successful establishment of longleaf pine forests on many sites (Provencher et al. 2001). Vegetation control by prescribed fire is short-lived on flatwoods sites due to the vigorous resprouting of many species. Typical shrub and herbaceous species of flatwoods sites have extensive rootstock systems that enable them to resprout rapidly following fire (Abrahamson 1984). The uses of mechanical and chemical site preparation and release treatments are suggested as alternatives to reduce the understory dominance of shrub species. Although intensive mechanical site preparation provides good shrub control (Burger et al. 1988), it also causes extensive ground disturbance that may be detrimental to desired native species. For instance, wiregrass, one of the keystone understory species of longleaf pine ecosystems, is reported to be very sensitive to ground disturbance (Clewell 1989). Chemical vegetation control methods are widely used as site preparation and pine release treatments. Herbicide applications at high rates (site preparation) were studied by Kline et al. (1994) and Shiver et al. (1991) on industrial flatwoods sites, with significant reduction of the understory shrub component. Information regarding the effect of herbicides on the ground-layer vegetation of natural flatwoods and sandhill communities is scarce (Litt et al. 2001). An extensive literature review done by Litt et al. (2001) regarding herbicide effects on groundlayer vegetation found only 3 and 7 studies respectively, on natural flatwoods and sandhill sites. A study done by Neary et al. (1990) on a flatwoods site in Florida found nearly 75% reduction in species richness following

PAGE 46

36 an intensive weed management regime with repeated applications of sulfometuron, glyphosate and triclopyr mowing. Brockway et al. (1998) studied the effects of hexazinone applied at 1.1 and 2.2 kg ha -1 with different application techniques on a sandhill site. They reported significant decreases in the cover of oak species and increases in shrub control at the higher rate. Moreover, the higher rate application of hexazinone induced an 86% increase in biomass of wiregrass in the first growing season after treatment. With the broadcast application of hexazinone, a short-term reduction of forbs species was also reported. A long-term study of herbicide effect on understory plant diversity and richness was conducted by Boyd et al (1995) on sandhill and piedmont site with one-time broadcast applications of imazapyr, glyphosate, and hexazinone. The reported results seven years after treatment showed no significant herbicide effect on species diversity and richness. Our study was conducted to examine the effects of three common forestry herbicides (imazapyr, hexazinone and sulfometuron) on the understory species of a flatwoods site. The specific questions addressed were: The effect of herbicides on the overall vegetation composition and species diversity; The effect of herbicides on cover and density of major understory shrub species; The effect of herbicides on the percent cover of wiregrass. Materials and Methods Study Area Description This study was conducted on a flatwoods site in Point Washington State Forest, Walton County, Florida (30 0 20.04 N, 86 0 4.22 W). Average annual highest and

PAGE 47

37 lowest temperatures were 25.5 0 C and 12 0 C, respectively. Annual precipitation was about 1640mm (2002) with most received in the late summer months (Figure 2.1). Soils are of low pH (pH <5) and sandy texture with low nutrient content (Table 2.1). The study area soils were mapped as sandy, siliceous, thermic aeric alaquods belonging to the Leon series, which is characterized by deep, poorly to very poorly drained soils. Soils of the flatwoods pinelands are formed on sandy quaternary formations derived from marine deposits (Stout and Marion 1993). Prior to study establishment the overstory was a stand dominated by slash pine with the occasional longleaf pine saplings. The average age of the stand was 26 years with a basal area of 1.85m 2 and an average dbh of 19.1cm. The understory was comprised mainly of species such as gallberry, saw palmetto, runner oak, dangleberry, hairy wicky, wiregrass and bluestem grasses. Dormant season fires were used on a three-year burn cycle to mitigate fuel build up. Experimental Design and Treatments Prior to site preparation, the overstory was harvested (August 2001). The harvest debris and the understory were roller chopped once and prescribed burned in October 2001. A randomized complete-block design (RCB) with five replicates was used to examine the effect of four herbicidal vegetation control methods on the understory vegetation of the site. The study included five treatment plots within each block. All treatment plots were 26.6m x 24.4 m, and included a > 3m buffer strip between plots. In December 2001, one year old containerized longleaf pine seedlings were hand-planted at 3.1m x 1.8m spacing. Seedlings were planted in rows to facilitate the application of treatments. In March 2002, four herbicide treatments [Sulfometuron (0.26 ai kg ha -1 ), Hexazinone (0.56 ai kg ha -1 ), Sulfometuron (0.26 ai kg ha -1 ) + Hexazinone (0.56 ai kg

PAGE 48

38 ha -1 ) mix, Imazapyr (0.21 ae kg ha -1 )] were applied in a 1.2 m band over the top of seedlings using a knapsack sprayer. In each block, one treatment plot was kept herbicide free as a control plot. Measurements A preliminary vegetation survey was conducted (June 2001) to assess the initial species composition and diversity, prior to harvesting and site preparation. The experimental blocks and treatment plots were demarcated and within all plots, three randomly selected 1m 2 sampling quadrats were sampled. After study establishment and herbicide application, two more vegetation surveys were conducted. These surveys were done three and nine (June and Nov 2002) months after herbicide treatment (MAT) and sampling was done on the herbicide applied strips along pine seedlings rows. In each treatment plot six randomly selected 1M 2 quadrats were sampled and the same location was revisited for subsequent survey. In every survey, all plants found within a quadrat were assigned to vegetation class of shrubs, grass, forbs, vines and ferns. Percent cover was ocularly estimated for all species using the modified Daubenmire scale (Daubenmire 1959). Stem number and average stem heights were recorded for all shrub species. Statistical Analysis Uniformity of percent cover of understory species and stem counts of major shrub species was tested within the pre-harvest survey data. Only species that displayed uniformity in percent cover and stem counts were used in comparison with post-treatment survey data. Percent cover data did not conform to the assumption of normality even after data transformation. Therefore, comparison of percent cover between treatments and the control plots for each survey date was performed with Kruskal-Wallis test at a significance level of =0.05. Treatment effects on stem counts were analyzed using

PAGE 49

39 analysis of variance (ANOVA) techniques within the framework of a randomized complete block design (RCBD) using SAS version 8.0 (SAS 2000). Significant treatment effects were separated using Duncans multiple range test. Importance value (IV) was calculated for shrubs, grasses and forbs. Shrub species IV was calculated as an average of relative frequency, relative cover and relative density. Average of relative frequency and relative cover was used as IV for grass and forb species. IV values were log transformed before analysis. Differences of IV values between treatments were tested with Duncans multiple range test at a significance level of =0.05. Species diversity was calculated using Shannon-Wiener diversity indices, which incorporate species richness and evenness of the species. Species diversity was calculated for all three vegetation surveys using percent cover data and were analyzed using Multivariate Statistical Package (MVSP) version 3.1 (Kovach 1999). Canonical Correspondence Analysis (CCA) was used to examine the overall distribution of species among the study plots. CCA is a direct gradient analysis which constraint the species distribution along specified environmental variables (Palmer 1993). The applied treatments were used as nominal environmental variables and time since treatment was used as a quantitative environmental variable in CCA analysis. CCA analysis was performed for each survey separately. Monte Carlo permutation test was used to detect significant (=0.05) environmental variables. Results Species Composition and Diversity CCA analysis with pre-harvest survey and control plots revealed a change in species composition with time without the effect of herbicides (Figure 3-1 a). Monte

PAGE 50

40 Carlo permutation tests showed time (p= 0.015) as a significant factor. The compositional change may be attributed to the seasonal variation within the plant community and the effects of site preparation. Analysis of 3 MAT vegetation survey data showed no significant changes in the species composition and distribution between the treatments (Fig 3-1 b). Monte Carlo permutation tests also revealed that 3 MAT none of the treatments were significant. However, 9 MAT data showed a distinct separation between herbicide treatment and the control (Fig 3-2 c). The first axis was highly correlated with absence of herbicides and the second axis although nonsignificant was correlated with imazapyr treatment. Monte Carlo permutation revealed absence of herbicide (P= 0.04) to be the most significant factor for the separation of sample plots within the ordination. The permutation tests did not reveal any significant herbicide treatment effect. A total of 81 species were recorded within the duration of the study. A larger number of species was found within post-treatment surveys compared to the survey done prior to harvest. The applied herbicides had no significant effect on species richness 3 and 9 MAT. Within post-treatment surveys, overall species evenness increased across all treatments (Table 3-1). However, in both surveys none of the herbicides significantly changed species evenness among the treatments. Although non-significant, imazapyr treatment resulted in the highest evenness in both surveys with sulfometuron resulting in the lowest evenness values. Species diversity increased 3 MAT across all the treatments compared to pre-harvest vegetation (Figure 3-2). However, none of the herbicide treatments had a

PAGE 51

41 significant effect on the understory species diversity 3 MAT (Figure 3-2). Further increase in diversity was observed across all treatments 9 MAT. The largest increase occurred with imazapyr treatment and was significantly greater than all other treatments, including control. Sulfometuron markedly reduced species diversity 9 MAT and was significantly lower than imazapyr, hexazinone treatments and control (Figure 3-2). Overall, shrub foliar cover was not significantly affected in both 3 and 9 MAT surveys in all treatments except imazapyr (Table 3-2). Imazapyr resulted in a reduction of 65% of shrub foliar cover 3 MAT and 58% reduction 9 MAT compared to control (Table 3-2). Overall grass and forb foliar cover were not significantly affected by any of the herbicides (Table 3-2). Understory Shrub Cover and Density Prior to harvest and site preparation, the understory was dominated by gallberry, runner oak, and saw palmetto. These species were abundantly found with percent frequency of 76, 56 and 42, respectively. These species were used to examine herbicide effects. Among the major understory shrub species, saw palmetto cover was not significantly affected by any herbicides in both post-treatment surveys (Table 3-2). Runner oak cover was not affected by the herbicides 3 MAT (Table 3-2). However, it showed a delayed response to imazapyr, hexazinone and sulfometuron+hexazinone mix treatments with a foliar cover reduction of 67%, 36% and 45%, respectively, compared to the control, 9 MAT (Table 3-2). The greatest reduction of cover occurred in gallberry in the imazapyr treatment with 77% reduction 3 MAT and 70% reduction 9 MAT. Gallberry cover was not affected by any other treatments (Table 3-2). Shrub density differences between treatments were analyzed for gallberry, runner-oak and saw palmetto, which showed uniform density across the study site prior to

PAGE 52

42 treatment application. Saw palmetto density was not affected by any of the applied treatments. The density of runner oak was not affected by any herbicides 3 MAT but showed delayed response to imazapyr with a 65% reduction of stem density 9 MAT (Table 3-3). Gallberry density was unaffected by the applied treatments except imazapyr. Within the imazapyr treatment gallberry density reduced by 58% and 79% 3 and 9 MAT, respectively. Analysis of mean importance values revealed no significant change due to applied herbicides except imazapyr (Table3-4). Imazapyr treatment significantly reduced the importance value of gallberry 3 MAT (> 50%) and was further reduced (>60%) 9 MAT. Wiregrass Cover Prior to harvest, wiregrass was the dominant grass species with 63 percent frequency. Interestingly none of the herbicide treatments negatively affected the foliar cover of wiregrass. Discussion Ecological disturbance such as periodic fires, hurricanes and plant mortality due to pathogens or insects influence the species diversity within an ecosystem by changing dominance, interspecific competition and initiating succession (Perry 1994). Herbicide application through its selective activity may also influence dominance of species, interspecific competition and species composition. Thus, with herbicide application changes in species diversity, richness and composition can be expected. CCA ordination on pre-harvest vegetation and the control plots revealed a change in species composition with time. The difference in compositional change may be a response to site disturbance and seasonal variation within the plant community. For instance, an increase in number of forb species such as Aster tortifolius, Liatris tenuifolia

PAGE 53

43 and Sabatia brevifolia were found in the latter survey conducted in late fall. The ordination analysis for 9 MAT vegetation data showed a significant separation of treatments due to the absence of herbicide. The separation between herbicide treatments and the control reveals a herbicide effect on overall species composition. In our study, post harvest vegetation surveys included a higher number of species compared to the pre harvest vegetation. The increase in species richness occurred in all treatment plots and may be attributed to the disturbances brought forth by site preparation treatments and prescribed burning. Similarly, Brockway et al. (1997) found increases in species richness following disturbances such as prescribed fire. Species richness did not change between the herbicide treatments indicating no apparent effect of treatments on species richness. This is contrary to published research, which shows a decline in species richness within the first growing season after herbicide application. A study done by Wilkins et al. (1993) with hexazinone on a flatwoods site reported significant reduction in species richness with increasing application rates within the first year. A similar study by Brockway et al. (1998) on a sandhill site with a broadcast application of hexazinone at 1.1 ai kg/ha reduced the species richness by 28%. It should be noted that both the above studies used site preparation rates of hexazinone and used broadcast application methods. The relatively lower rate application of herbicides and the banded application method used in our study did not significantly affect species richness. The increase in species diversity across the study site is a characteristic response to site disturbances. Usually a decrease in species diversity within the first growing season after herbicide treatments is common. Brockway et al. (1998) reported an overall reduction in species diversity on a sandhill site treated with high rates of hexazinone.

PAGE 54

44 With the exception of imazapyr and sulfometuron, the herbicide treatments used in our study did not significantly impact understory species diversity at the applied rates. Imazapyr was the only treatment to significantly increase species diversity 9 MAT. In addition, an increase in species evenness was also observed. Species evenness measures the similarity of abundance between species found within a given area. Higher evenness values signify a lack of dominance by a fewer species and reveals similar proportions of abundance. The significant reduction of shrub foliar cover and density by imazapyr may have reduced the dominance of shrub species, thereby increasing resource availability to other species. The absence of shrub dominance may have facilitated similar distribution of abundance among other species increasing species evenness and in turn improving species diversity. In contrast to imazapyr, sulfometuron resulted in a significant decrease in species diversity. The decrease in diversity is attributable to the lower species richness and evenness found in the sulfometuron treatment. Although imazapyr showed significant control over the dominant shrub species, it did not cause a reduction in species richness within the treated area. This reveals the selectivity of imazapyr on its target species and its inability to severely damage or eliminate them from the understory at the applied rate. The herbicide application method used in our study may also have contributed to the lack of severity of herbicide effect on species diversity and richness. In contrast to broadcast application, the band application method provides a herbicide free buffer between treatment bands. The buffer strips provide a suitable seed source for re-colonization within the treated plots. Studies done with broadcast and spot application treatments have reported greater decline in species

PAGE 55

45 richness and diversity with the broadcast application of herbicides compared to the spot application method (Brockway et al. 1998). One of the important concerns with longleaf pine restoration techniques is their effect on wiregrass regeneration and growth. Clewell (1989) reported on the detrimental effects of soil disturbance on wiregrass. In his research 85% reduction in wiregrass density was found with soil disturbance in flatwoods. In addition, the negative impact of site preparation methods such as chopping was also reported. In contrast to the above findings, the site preparation chopping performed in this study did not have any negative effect on wiregrass foliar cover. However, the damaging effects of soil disturbances is lessened along an increasing moisture gradient (Clewell 1989). Thus, the recovery of wiregrass following site disturbances may be site specific. A study done on a sandhill site reported an increase in wiregrass cover in a linear relationship with increasing application rates of hexazinone within the first growing season (Wilkins et al. 1993). They also reported a significant reduction in the dominant oak species that may have led to the increase in wiregrass cover. A similar study done by Brockway et al. (1998) reported an 86% increase in wiregrass cover following a 93% reduction in dominant shrub cover. The above research indicates a positive response of wiregrass to reduction of shrub dominance within the first growing season. In our study, the herbicide treatments did not significantly effect the cover of wiregrass. However, imazapyr resulted in a significant reduction in gallberry cover and runner oak showed delayed significant reduction to hexazinone, imazapyr and sulfometuron+hexazinone mix treatments. Thus, the reduced shrub competition may yield greater wiregrass cover in the

PAGE 56

46 following years. Wiregrass is also reported to experience a lag time and requires at least two growing seasons to respond to reduced competition (Brockway et al. 1998). Overall, imazapyr yielded the most desirable results by reducing the shrub dominance while improving the understory biodiversity. Imazapyr was also successful in yielding improved growth of planted pine seedlings. Hexazinone yielded a delayed control of runner oak with no significant impact on the understory biodiversity. The benefits of hexazinone with its delayed control of runner oak may be realized in the coming years. However after one growing season hexazinone did not yield greater benefits compared to control. Sulfometuron yielded poor results by reducing the biodiversity with no significant control over shrub dominance. The application of sulfometuron was also found detrimental to planted longleaf pine seedlings on flatwoods sites as described in chapter two. The sulfometuron+hexazinone mix treatments did not result in any improved benefits. Conclusions Within the first growing season, imazapyr treatment significantly increased the species diversity while increasing evenness. In contrast, sulfometuron decreased the species diversity and reduced species evenness. None of the other herbicides significantly impacted the understory plant diversity. Imazapyr treatment resulted in an overall reduction of shrub foliar cover with a significant reduction of gallberry and runner oak. However, shrub species such as saw palmetto and huckleberry species displayed resistance to imazapyr at the applied rate. Imazapyr caused a higher mortality rate of pine seedlings compared to control, but resulted in a significant increase in growth. Hexazinone was not effective in the control of major understory shrub species with the exception of runner oak. Runner oak displayed a delayed response to hexazinone by

PAGE 57

47 significant reduction of foliar cover 9 MAT. None of the other herbicide treatments was effective in reducing shrub dominance. Therefore, on flatwoods sites with an understory dominated by gallberry and runner oak, application of imazapyr seems to yield the best control of shrub dominance one-year after treatment. The use of sulfometuron at the above-applied rates is not recommended on flatwoods sites as it decreased understory species diversity while hindering longleaf pine seedling growth.

PAGE 58

48 Figure 3-1.Canonical correspondence analysis (CCA) ordination for understory vegetation. (a) CCA ordination with pre harvest and post treatment control plots. (b) CCA ordination of vegetation 3 months after treatment (MAT). (c) CCA ordination of vegetation 9 MAT. The tested nominal variables are plotted as centroids.

PAGE 59

49 CCA case scores pre-treatment June 01 control June 02 control Nov 02Axis 2Axis 1 -0.41-0.83-1.24-1.66-2.070.410.831.241.662.07 -0.41-0.83-1.24-1.66-2.070.410.831.241.662.07 time Vector scalin g : 1.73 Novem b er 2002 June 2002 June 2001 (a) CCA case scores sulfo+hexa hexa sulfo cont imazAxis 2Axis 1 -0.9-1.8-2.6-3.5-4.40.91.82.63.54.4 -0.9-1.8-2.6-3.5-4.40.91.82.63.54.4 hexa sulfo sulfo+hexa imaz cont Vector scalin g : 4.43 (b) CCA case scores sulfo+hexa hexa sulfo cont imazAxis 2Axis 1 -0.7-1.4-2.1-2.8-3.50.71.42.12.83.5 -0.7-1.4-2.1-2.8-3.50.71.42.12.83.5 hexa sulfo sulfo+hexa imaz cont Vector scalin g : 3.33 control (c)

PAGE 60

50 abaa bc1.522.5pre-harvest3 MAT9 MATDateShannon's diversity index control sulfometuron hexazinone sulfo+hexa imazapyr Figure 3-2. Mean Shannon diversity index for each treatment by date. Shannons index values for each treatment followed by a different letter is significantly different from other treatments within each date. Significance established through Duncans multiple range test at =0.1. (sulfo+hexa: sulfometuron+hexazinone mix treatment)

PAGE 61

51 Table 3-1. Mean species evenness and richness of three vegetation surveys Treatment Evenness Richness Pre harvest survey 0.70 10.40 3 MAT 2 survey Control 0.76 14.80 Sulfometuron 0.74 11.60 Hexazinone 0.78 13.00 Sulfo+Hexa 1 0.77 12.40 Imazapyr 0.81 13.80 9 MAT 2 survey Control 0.83 13.00 Sulfometuron 0.77 11.20 Hexazinone 0.83 12.60 Sulfo+Hexa 1 0.81 12.80 Imazapyr 0.87 14.80 1 Sulfometuron+Hexazinone mix treatment 2 Months after treatment

PAGE 62

52 Table 3-2. Percentage change in foliar cover (% cover) of major understory species and vegetation classes by treatment Species/ Sulfometuron Hexazinone Sulfo+Hexa Imazapyr Vegetation class Wiregrass June 2002 33.86 + 44.44 + 11.11 + 29.29 + November 2002 21.53 + 18.69 + 11.82 + 7.82 + Bluestem spp June 2002 9.27 + 12.14 + 3.51 + 48.24 November 2002 56.93 a 19.83 8.77 + 19.29 Gallberry June 2002 12.80 + 19.27 + 45.79 + 77.04 a November 2002 7.16 + 20.65 + 34.68 + 69.61 a Runner oak June 2002 13.17 + 34.22 42.93 67.91 November 2002 10.26 + 36.66 a 45.34 a 68.33 a Saw palmetto June 2002 28.29 22.45 29.19 0.71 + November 2002 6.18 + 11.12 25.97 + 0.73 + All shrub June 2002 12.22 32.50 27.28 54.20 a November 2002 1.07 16.67 11.98 49.87 a All grasses June 2002 29.49 21.55 30.80 34.55 November 2002 28.34 4.15 20.57 13.53 All forbs June 2002 34.34 8.26 32.60 20.00 + November 2002 15.07 25.59 22.24 47.36 + 1 sulfometuron+hexazinone mix treatment + and indicate increase and decrease of foliar cover in relation to control a significantly different from control within rows. Significance established with KruskalWallis test at = 0.1

PAGE 63

53 Table 3-3. Mean stem numbers for three major understory shrub species by treatment Shrub stem numbers (stems/m 2 ) Species Pre harvest Gallberry 22.422.1 Runner oak 32.163.8 Saw palmetto 2.220.2 Control Sulfometuron Hexazinone Sulfo+Hexa Imazapyr 3 MAT 1 Gallberry 18.372.6 20.602.7 16.931.5 16.601.8 7.600.9 a Runner oak 31.405.3 31.295.9 23.143.9 20.322.5 17.623.6 Saw palmetto 2.200.5 2.450.4 2.000.4 2.550.64 2.900.6 9 MAT 1 Gallberry 34.504.9 39.805.3 38.333.7 43.005.9 7.031.6 a Runner oak 52.077.8 59.2012.1 37.576.4 36.574.7 18.043.3 a Saw palmetto 2.750.5 3.440.8 2.000.6 3.281.3 3.330.8 1 Months after treatment a significant differences between treatment within rows. Significance established by Duncan's multiple range test at = 0.1

PAGE 64

Table 3-4. Importance values (IV) for major understory species in three vegetation surveys by treatment. 54 Species 1 Treatment Blue Arbe Gadu* Ilgl* Kahi* Limi* Pani Qupu* Sere* Pre harvest MAT MAT 0.79 0.60 0.99 1.05 1.02 1.11 0.70 1.05 0.954 3 Control 0.70 0.54 1.65 1.22 1.65 1.68 0.88 1.32 1.32 Sulfometuron 0.85 0.48 1.51 1.26 1.68 0.23 0.71 1.12 1.23 Hexazinone 0.92 0.62 1.50 1.21 0.40 1.54 0.17 1.13 0.78 Sulfo+Hexa 2 0.73 0.52 0.92 1.27 0.39 0.89 0.65 1.06 0.95 Imazapyr N/A 3 0.45 N/A 3 0.45 a N/A 3 1.06 0.29 0.63 1.20 9 Control 0.60 0.47 1.59 1.12 2.38 1.22 1.04 1.41 1.11 Sulfometuron 0.53 0.58 1.61 1.24 N/A 3 N/A 3 0.46 1.24 1.45 Hexazinone 0.60 0.65 1.03 1.35 N/A 3 1.39 0.28 1.09 0.73 Sulfo+Hexa 2 0.77 0.63 0.57 1.42 0.37 1.13 0.10 1.07 1.05 Imazapyr 0.71 0.56 1.26 0.36 a 1.07 1.06 0.34 0.51 1.18 1 Blue:Bluestem spp, Arbe:Wiregrass, Gadu:Dwarf huckleberry, Ilgl:Gallberry, Kahi:Hairy wicky, Limi:Gopher apple, Pani:Panicum spp, Qupu:Runner oak, Sere:Serenoa repens 2 Sulfometuron+Hexazinone mix treatment 3 N/A: Not present in treatment plots at the time of survey IV values calculated as an average of relative frequency, cover and density a Significantly different along rows. Significance established with Duncan's multiple range test at = 0.1

PAGE 65

CHAPTER 4 SUMMARY AND CONCLUSIONS Successful restoration of longleaf ecosystems and their associated understory may depend on the control of competing vegetation. On flatwoods sites, the need for vegetation control is magnified due to the dense shrub understory that smothers young longleaf seedlings and the herbaceous vegetation. Although prescribed fire is an effective restoration tool in many longleaf sites, in flatwoods its shrub control is short lived due to the vigorous sprouting of the shrub vegetation (Abrahamson 1984). Intensive mechanical site preparation, while yielding enhanced growth of seedlings (Burger et al. 1988), may cause extensive damage to the understory vegetation (Clewell 1989). In our study, we examined the effects of over the top application of herbicides following site preparation with chopping and burning. The use of herbicides such as imazapyr exhibited an evident trade off between non-application and application of herbicides. Herbicide treatments such as imazapyr, which resulted in significantly higher growth within the first growing season, also resulted in significantly lower survival. Where as in control, higher survival was observed but seedling growth was significantly lower compared to imazapyr treatment. On the other hand hexazinone had no effect on survival and did not improve longleaf pine seedling growth. The use of sulfometuron on flatwoods is not recommended as it significantly lowered survival and growth of planted longleaf pine seedlings. The sulfometuron+hexazinone mix treatment did not yield any additional benefit. Imazapyr was the only treatment that yielded significant control of the 55

PAGE 66

56 overall shrub foliar cover with no effect on the herbaceous component of the understory. The significant increase in seedling growth by controlling the shrub component alone reiterates the importance of shrub vegetation control on flatwoods sites. Hexazinone yielded delayed control of runner oak which was one of the dominant shrub species and may yield desirable results in the coming years. The understory species composition exhibited a change without the influence of herbicide treatments. These changes could be attributed to the seasonal changes in the vegetation and the site preparation activities. However, the herbicide treatments as a whole showed a distinct influence on the overall species composition. The understory species diversity was increased across all treatments compared to pre harvest vegetation. The herbicide treatments did not have a significant effect on the species diversity with the exception of imazapyr and sulfometuron. Imazapyr increased species diversity 9 MAT with nonsignificant increases in species evenness and richness. In contrast, sulfometuron decreased species diversity while resulting in lowered evenness and richness. Contrary to published research, the site preparation conducted in our study did not significantly affect the foliar cover of wiregrass. The herbicide treatments also had no significant affect on wiregrass. In conclusion, imazapyr yielded the best-desired effects with increased longleaf pine seedling growth, significant shrub control and improved understory plant diversity. In contrast, sulfometuron resulted in lower survival, growth and species diversity. Therefore, application of sulfometuron for shrub control on flatwoods sites is not recommended.

PAGE 67

APPENDIX SPECIES LIST Scientific name Code Common name Shrubs Asimina incana Asin Wooly paw paw Cyrilla racemiflora Cyra Ti ti Gaylussacia dumosa Gadu Drawf huckleberry Gaylussacia frondosa Gafr Dangleberry Ilex cariacea Ilca Large gallberry Ilex glabra Ilgl Gallberry Ilex vomitoria Ilvo Yaupon Kalmia hirsuta Kahi Hairy wicky Licania michauxii Limi Gopher apple Lyonia lucida Lylu Fetterbush Magnolia virginiana Mavi sweet bay Myrica cerifera Myce Wax myrtle Photinia pyrifolia Phpy Red choke berry Quercus pumila Qupu Runner oak Seronoa repens Sere Saw palmetto Stillingia sylvatica Stsy Queens delight Vaccinium spp Vacc Blueberry spp Grasses Bluestem spp Blue Bluestem grasses Arisitida beyrichiana Arbe Wiregrass Calamovilfa curtissii Cacu Curtis sandgrass Ctenium aromaticum Ctar Toothache grass Cyperus Cype Sedge spp Eragrostis spectabilis Ersp Purple lovegrass Panicum Pani Panicun spp Panicum erectifolium Paer Erect leaf witchgrass Panicum laxiflorum Pala Velvet Witchgrass Scleria Scle Nutrush spp Xyris caroliniana Xyca Yellow eyed grass Forbs Asclepias viridula Asvi Southern milkweed Aster adnatus Asad Scaleleaf aster Aster eryngiifolius Aser Thistleleaf aster Aster reticulatus Asre White top aster Aster tortifolius Asto Dixie aster 57

PAGE 68

58 Species list continued Carphephorous pseudoliatris Caps Bristleleaf chaffhead Carphephorus odoratissimus Caod Deer tounge Chrysopsis Chry Silkgrass spp Conyza canadensis Coca Canadian horseweed Coreopsis linifolia Coli Texas tickseed Desmodium rotundifolium Dero Tricklyfoil Drosera capillaris Drca Pink sundew Elephantopus tomentosus Elto Devils granmother Eupatorium capillifolium Euca Dog fennel Eupatorium compositifolium Euco Yankee weed Eupatorium mohrii Eumo Mohrs thoroughwort Eupatorium pilosum Eupi Rough Boneset Euthamia graminifolia Eugr Flat top goldenrod Gelsemium sempervirens Gese Yellow jessamine Gratiola hispida Grhi Rough Hedgehyssop Hypericum hypericoides Hyhy St. Andrews cross Hypoxis sessilis Hyse Glossyseed yellow stargrass Hypoxis spp Hypo Stargrass spp Lachnanthes caroliana Laca Carolina redroot Lechea Lech Pineweed spp Lechea pulchella Lepu Leggetts pineweed Liatris gracilis Ligr Slender gayfeather Liatris tenuifolia Lite Shortleaf gayfeather Mimosa quadrivalvis Miqu Sensitive brier Oenothera fruticosa Oefr Evening primrose Opuntia humifusa Ophu Prickly pear Pityopsis graminifolia Pigr Silkgrass Pterocaulon pycnostachyum Ptpy Blackroot Rhexia alifanus Rhal Meadow beauty Rhexia petiolata Rhpe Fringed meadow beauty Sabatia brevifolia Sabr Shortleaf Rosegentian Seymeria cassioides Seca Yaupon Blacksenna Smilax laurifolia Smla Laurel green brier Smilax pumila Smpu Green brier Solidago odora Sood goldenrod Stylisma patens Stpa Coastal plain dawn flower Tragia urens Trur Wavyleaf noseburn Verbena brasiliensis Vebr Brazilian vervain Viola septemloba Vise Blue violet Vitis rotundifolia Viro Muscadine

PAGE 69

LIST OF REFERENCES Anon. 2000. Guidance document on residue analytical methods, European Commission, Directorate General Health and Consumer Protection. SANCO/825/00 rev 6, 20 June 2000. 16p. Abrahamson, W. G. 1984. Post-fire recovery of Florida Lake wales ridge vegetation. Am. J. Bot. 71 (1): 9-21. Boyd, R. S., J. D. Freeman, J. H. Miller and M. B. Edwards. 1995. Forest herbicide influences on floristic diversity seven years after broadcast pine release treatments in central Georgia, USA. New Forests. 10: 17-37. Boyer, W. D. 1990. Pinus Palustris, Longleaf pine. In: Burns, R. M. and B. H. Honkala. Silvics of North America, Vol. 1. Conifers. USDA Forest Serv. Washington, DC, pp 405-412. Brewster, B. D. and A. P. Appleby. 1983. Response of wheat (Triticum aestivum) and rotation crops to chlorosulfuron. Weed Sci. 31: 801-805. Brockway, D. G., and C. E. Lewis. 1997. Long-term effect s of dormant-season prescribed fire on plant community diversity, structure and productivity in a longleaf pine wiregrass ecosystem. Forest. Ecol. Manage. 96: 167-183. Brockway, D. G., and K. W. Outcalt. 2000. Restoring longleaf pine wiregrass ecosystems: Hexazinone application enhances effects of prescribed fire. Forest. Ecol. Manage. 137: 121-138. Brockway, D. G., K. W. Outcalt, and R. N. Wilkins. 1998. Restoring longleaf pine wiregrass ecosystems: Plant cover, diversity and biomass following low rate hexazinone application on Florida sandhills. Forest. Ecol. Manage. 103: 159-175. Burger, J. A., and W. L. Pritchett. 1988. Site preparation effects on soil moisture and available nutrients in a pine plantation in the Florida flatwoods. Forest. Sci. 34: 77-87. Busby, R. L., C. E. Thomas, R. E. Lohrey, K. H. N. Le, and M. B. Edwards. 1995. Rapidly restoring large diameter longleaf pine ecosystems at a low cost: A preliminary appraisal. Proc. of the Eighth Bienn. South. Silvic. Res. Conf., Auburn, Alabama, 1994. Gen. Tech. Rep. South. Res. Stn. US Dept. Agric. Forest. Serv. SRS-1, 192-196. 59

PAGE 70

60 Carter, R. E., M. D. Mackenzie, D. H. Gjerstad and T. A. Waldrop. 1998. Ecological land classification of longleaf pine ecosystems in the Southern Loam Hills of South Alabama. Proc. 9 th Bienn. South. Silvic. Res. Conf. Gen. Tech. Rep. US Dept. Agric. Forest. Serv. SRS 20, 81-85. Clewell, A. F. 1989. Natural history of wiregrass (Aristida stricta). Nat. Areas J. 9, 223-233. Collins, B., P. S. White and D. W. Imm. 2001. Introduction to ecology and management of rare plants of the southeast. Nat. Areas J. 21(1): 4-11. Croker, T. C., Jr., 1975. Seedbed preparation aids natural regeneration of longleaf pine. USDA. For. Ser. Res. Pap, South. Forest Exp. Stn. No. SO-112. Croker, T. C., Jr., 1979. Longleaf pine. The longleaf pine story. J. Forest History. 23(1): 32-43. Daubenmire, R. F. 1959. Canopy coverage method of vegetation analysis. Northwest Sci. 33: 43-64. Davis, J. H. 1967. General map of natural vegetation of Florida. Inst. Food and Agric. Sci. Circ. S-178, University of Florida, Gainesville. Delcourt, P. A., H. R. Delcourt, D. F. Morse and P. A. Morse. History, evolution and organization of vegetation and human culture. In: Martin, W. H., S. G. Boyce and A. C. Echternacht. Ed. Biodiversity of the Southeastern United States, Lowland Terrestrial Communities. John Wiley and Sons, Inc., New York pp 47-80 Dickens, R., and G. Wehtje. 1986. Mobility and soil solution characteristics of imazapyr and sulfometuron methyl. Proc. South. Weed Sci. Soc. Conf. 39: 368. Frost, C. C. 1993. Four centuries of changing landscape patterns in the longleaf pine ecosystem. In: Hermann, S. M. (ed), Proc. Tall Timbers Fire Ecol. Conf. Vol. 18. Tall Timbers Res. Stn., Tallahassee, FL, pp 17-43. Gjerstad, D. H., L. R. Nelson. 1983. New herbaceous weed control in pine outplantings. Proc. South. Weed Sci. Soc. Conf. 36: 245. Grelen, H. E. 1978. May burns stimulate growth of longleaf pine seedlings. US Forest Serv. Res. Note, South. Forest Exp. Stn. No. SO-234. Grelen, H. E. 1983. May burning favors survival and early height growth of longleaf pine seedlings. South. J. Appl. For. 7 (1):, 16-20. Haywood, J. D. 2000. Mulch and hexazinone herbicide shorten the time longleaf pine seedlings are in the grass stage and increases height growth. New Forests. 19: 279-290.

PAGE 71

61 Hernandez-Sevillano, E., M. Villarroya, J. L. Alonso-Prados and J. M. Garcia-Baudin. 2001. Bioassay to detect MON-37500 and triasulfuron residues in soils. Weed Technology. 15: 447-452. Hollaway, K. L., R. S. Kookana, D. J. McQuinn, M. R. Moerkerk, D. M. Noy and M. A. Smal. 1999. Comparison of sulfonylurea herbicide residue detection in soil by bioassay, enzyme-linked immunosorbent assay and hplc. Weed Res. Oxford. 39(5): 383-397. Huck, R. B. 1987. Plant communities along an edaphic continuum in a central Florida watershed. Florida scientist. 50: 111-128. Jose, S., S. Merritt and C. L. Ramsey. 2003. Growth, nutrition, photosynthesis and transpiration responses of longleaf pine seedlings (Pinus palustris Mill.) to light, water and nitrogen. Forest Ecology and Management 180:335-344. Kelly, J. F., and W. A. Bechtold. 1990. The longleaf pine resource. In: Farrar, R. M. (Ed). Proc. Symp. Manage. Longleaf pine. USDA Forest Serv. South. Forest Exp. Stn. Gen. Tech. Rep. SO-75, New Orleans LA, pp. 11-22. Kline, W. N., J. L. Troth and B. D. Shiver. 1994. Maximizing pine yields in the southern flatwoods using Garlon herbicide. Down to Earth. 49: 14-22. Kovach, W. L., 1999. MVSP-A multivariate statistical package for windows, ver. 3.1. Kovach Computing Services, Pentraeth, Wales, UK. Landers, J. L., D. H. Van Lear, and W. D. Boyer. 1995. The Longleaf Pine Forests of the Southeast: Requiem or Renaissance? J. For. 93: 39-44. Lipscomb, D. J. 1989. Impacts of feral hogs on longleaf pine regeneration. South. J. Appl. For. 13(4): 177-181. Litt, A. R., B. J. Herring and L. Provencher. 2001. Herbicide effects on ground layer vegetation in southern pine lands (USA): A review. Nat. Areas J. 21 (2): pp 177-188 Metcalfe, C. S., and R. L. Cantrell. 1986. Secondry screening for herbaceous weed control in longleaf pine plantations-Oust regestration data. Auburn University Silvic. Herbicide Cooperative. Res. note 86-1. Michael, J. L. 2000. Pesticides used in forestry and their impacts on water quality. Proc. South. Weed Sci. Soc. 53: 81-87. Michael, J. L., E. C. Webber, Jr., D. R. Bayne, J. B. Fischer, H. L. Gibbs and W. C. Seesock. 1999. Canadian J. Forest Res. 29: 1170-1181 Michael, J. L., and D. G. Neary. 1993. Herbicide dissipation studies in southern forest ecosystems. Environ. Toxicology and Chemistry. 12: 405-410.

PAGE 72

62 Neary, D. G., and J. L. Michael. 1989. Effect of sulfometuron methyl on groundwater and stream quality in coastal plain forest watersheds. Water Resour. Bull. 25 (3): 617-623. Neary, D. G., J. L. Michael, J. A. Griffith. 1996. HerbicidesProtecting long term sustainability and water quality in forest ecosystems. Second Int. Conf. Forest vegetation manage. 26: 1-2, 241-264. Norris, L. A. 1981. The movement, persistence, and fate of the Phenoxy herbicides and TCDD in the forest. Residue Rev. 80: 65-135. Outcalt, K. W., 2000. The longleaf pine ecosystem of the south. Native Plants J. 1 (1): 43-51. Palmer, M. W. 1993. Putting things in even better order: The advantages of canonical correspondence analysis. Ecology. 74: 2215-2230. Pederson, N., J. S. Kush, R. S. Meldahl, W. D. Boyer, J. D. Haywood. 1999. Longleaf pine cone crops and climate: a possible link. Gen. Tech. Rep. South. Res. Stn, USDA Forest Serv. No.SRS-30, 255-258. Peet, R. K., 1993. A taxanomic study of Aristida stricta and A. beyrichiana. Rhodora 95 (881): 25-37. Peet, R. K and D. J. Allard 1993. Longleaf pine dominated vegettaion of the southern Atlantic and eastern gulf coastal region, USA. In: S. M. Hermann (Ed), Proc. Tall Timbers Fire Ecol. Conf. Tall Timbers Res. Stn., Tallahassee, FL, 18: 45-81. Perry, D. A. 1994. How biodiversity is created and maintained. In: Forest ecosystems. Baltimore. Johns Hopkins University Press, Baltimore. pp196-218. Platt, W. J., G. W. Evans, and M. M. Davis. 1988. Effects of fire season on flowering of forbs and shrubs in longleaf forests. Oecologia. 76: 353-363. Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, K. E. M. Galley, G. W. Tanner, J. L. Hardesty and L. A. Brennan. 2001. Effects of hardwood reduction techniques on longleaf pine sandhill vegetation in northwest Florida. Rest. Ecol. 9: 13-27. Ramsey, C. L., S. Jose and S. Ranasinghe. 2004. Manuscript in preparation. Richardson, D. R. 1985. Allelopathy and fire in the Florida scrub. Am. J. Bot. 72(6): 864. Shiver, B. D., J. W. Rheney, M. J. Oppenheimer. 1990. Site preparation method and early cultural treatments affect growth of flatwoods slash pine plantaions. South. J. Appl. For. 14 (4): 183-188.

PAGE 73

63 Shiver, B. D., S. A. Knowe, M. B. Edwards and W. N. Kline. 1991. Comparison of herbicide treatments for controlling common Coastal Plain flatwood species. South. J. Appl. For. 15: 187-193. Stork, P., and M. C. Hannah. 1996. A bioassay method for formulation testing and residue studies if sulfonylurea and sulfonanylide herbicides. Weed Res. 36: 271-281. Stout, I. J., and W. R. Marion. 1993. Pine flatwoods and xeric pine forests of the southern (lower) coastal plain. In: Martin, W. H., S. G. Boyce and A. C. Echternacht. Ed. Biodiversity of the Southeastern United States, Lowland Terrestrial Communities. John Wiley and Sons, Inc. New York. pp 373-446 Wilkins, R. N., W. R. Marion, D. G. Neary, and G. W. Tanner. 1993. Vascular plant community dynamics following hexazinone site preparation in the lower coastal plain. Canadian J. Forest Res. 23: 2216-2229.

PAGE 74

BIOGRAPHICAL SKETCH Sanjaya Ranasinghe was born on December 26 th 1975, in Colombo, Sri Lanka. In 1996 he came to the United States to pursue an undergraduate degree at Luther College Decorah, Iowa. Upon graduation in 2000 with a B.A. in Biology he was employed at Tidewater Inc., an environmental consultant company based in Columbia, Maryland. In 2001 he was accepted to pursue a masters degree with the School of Forest Resources and Conservation at University of Florida. 64


Permanent Link: http://ufdc.ufl.edu/UFE0001443/00001

Material Information

Title: Role of Herbicides in Longleaf Pine Flatwoods Restoration: Pine Growth, Understory Vegetation Response and Fate of Applied Herbicides
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001443:00001

Permanent Link: http://ufdc.ufl.edu/UFE0001443/00001

Material Information

Title: Role of Herbicides in Longleaf Pine Flatwoods Restoration: Pine Growth, Understory Vegetation Response and Fate of Applied Herbicides
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001443:00001


This item has the following downloads:


Full Text












ROLE OF HERBICIDES IN LONGLEAF PINE FLATWOODS RESTORATION: PINE
GROWTH, UNDERSTORY VEGETATION RESPONSE AND FATE OF APPLIED
HERBICIDES















By

SANJAYA RANASINGHE


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Sanjaya Ranasinghe

































To my beloved parents, sisters and Jenny.















ACKNOWLEDGMENTS

I would like to thank my supervisory committee members, Drs. Shibu Jose (Chair),

Debbie Miller and Alan Long, for their advice, time and support. In particular, I thank

Dr. Shibu Jose for providing me the opportunity to attend graduate school. I would like

to thank Dr. Debbie Miller for her advice concerning vegetation analysis procedures. Her

knowledge on longleaf ecosystems and its understory vegetation, was a valuable asset.

Dr. Long with his vast knowledge on fire ecology and longleaf ecosystems provided

invaluable improvements to my research. I would also like to thank Dr. Craig Ramsey

for his assistance with the project from its beginning. His contributions, from field work

to data analysis are much appreciated.

The efforts of Sara Merritt, Cathy Hardin, Leah McCue, Lewayne White, Chris

Atkins and Tim Baxley in data collection work are much appreciated. I would also like

to thank my fellow graduate students Diomides Zamora, Robert Wanvestrout, Maheteme

Gebremedhin and Andrew Ruth for their friendship and support throughout my study.

I would also like to thank the Florida Division of Forestry for providing land, field

assistance and funding for the project.

A special thank you to Jenny whose constant encouragement, love and support

provided me with the strength to strive harder. I would also like to thank my parents and

sisters for always believing in me and supporting me all throughout my education.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ....................................................... ............ .............. .. vii

LIST OF FIGURES ..................................................... .......... ................ viii

ABSTRACT .............. .......................................... ix

CHAPTER

1 LONGLEAF PINE ECOSYSTEMS AND THEIR RESTORATION
IN FL A TW O O D S ....................... .... ............................ ........ .............

Longleaf Pine Ecosystem s and Their Decline............................................................1
Role of Fire, Mechanical and Chemical Site Preparation as Restoration
T e c h n iq u e s ..................................................... ............... ................ .. 2
Longleaf Pine Flatw oods R estoration.................................... .................................... 5
Fate of A applied H erbicides..................................................................... ............... 8

2 LONGLEAF PINE SEEDLING AND UNDERSTORY VEGETATION
RESPONSE TO HERBICIDAL VEGETATION CONTROL AND
FATE OF APPLIED HERBICIDES IN FLATWOODS ..........................................10

Introduction ............... ....... ............................. ............................10
P ine F latw oods ..................................................................... 10
Vegetation Control and Pine Response .........................................................11
F ate of A applied H erbicides...................................................................... .. .... 12
M materials and M methods ....................................................................... .................. 14
Study Area D description .................................... ......................................14
Experimental Design and Treatments ...................................... ............... 14
M easurem ents ...............1... ...................5.............................
Pine survival and grow th ........................................................... ... .......... 15
Vegetation control assessment .............. .............................................. 15
H erbicide dissipation assessm ent..... .......... ....................................... 16
Statistical A analysis .......................... ...................... .............. .. ........... ..17
Pine survival and grow th ........................................................... ... .......... 17
Vegetation control .................. .................................... ...... ...........17
H herbicide persistence analysis ........................................ ............... 18
R e su lts ...................................... .................................................... 1 8









P in e S u rv iv a l ................................................................................................. 1 8
P ine R C D H eight and SV I...................................................................... ...... 18
V egetation C control .......................................... .................. ........19
H erbicide D issipation .................. ........................... ..................... 20
D isc u ssio n ............................................................................................................. 2 1
C conclusion ...................................................................................................... ....... 24

3 UNDERSTORY SPECIES DYNAMICS FOLLOWING HERBICIDAL PINE
RELEASE TREATMENTS ON A LONGLEAF FLATWOODS SITE ...................33

In tro d u ctio n .......................................................................................3 3
M materials and M methods ....................................................................... ..................36
Study A rea D description .................................... ....................................... 36
Experim ental Design and Treatm ents ...................................... ............... 37
M easurem ents ................................................................. ............38
Statistical A analysis ...................... ...... ............ ................. ...........38
R results ............................... ...................................................... ........ 39
Species Com position and D iversity................................... ...................... 39
U nderstory Shrub Cover and D ensity................................. ............................41
W iregrass C over ..................................................................... ......... 42
Discussion ............... ........ ......................... 42
C o n c lu sio n s........................................................................................................... 4 6

4 SUMMARY AND CONCLUSIONS.................................................................55

A PPEN D IX SPECIES LIST ................................................... ............................... 57

LIST OF REFEREN CE S .. ....... ............................................................. ............... 59

BIOGRAPH ICAL SKETCH ...................................................... 64
















LIST OF TABLES

Table pge

2-1 Soil description at four depths...................................................................... .. .... 30

2-2 Mean foliar cover for major understory species....................................................31

2-3 Stem num bers and stem heights..................................................................... ....32

3-1 Mean species evenness and richness of three vegetation surveys............................ 51

3-2 Percentage change in foliar cover ........................................ ........................ 52

3-3 M ean stem num bers ....................... .................... ................... .. ......53

3-4 Im portance values (IV ).................................................. ............................... 54
















LIST OF FIGURES


Figure page

2-1 M monthly rainfall data for year 2002 ............................................... .............. 25

2-2 Longleaf pine seedling survival and growth ............ ...........................26

2-3 Stem volume index oflongleaf seedlings ....................... .................... 27

2-4 M ean herbicide concentrations in part per billion............... .................. ...........28

2-5 Change in herbicide concentration over time across soil profile ...........................29

3-1 Canonical correspondence analysis (CCA) ordination ........................................48

3-2 M ean Shannon diversity index ...................................................... ..................50















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

ROLE OF HERBICIDES IN LONGLEAF PINE FLATWOODS RESTORATION: PINE
GROWTH, UNDERSTORY VEGETATION RESPONSE AND FATE OF APPLIED
HERBICIDES

By

Sanjaya Ranasinghe

December, 2003

Chair: Shibu Jose
Major Department: School of Forest Resources and Conservation

Longleaf pine ecosystems, once dominant in the southeastern Coastal Plain, have

been drastically reduced in acreage by excessive logging, land use changes and fire

suppression. Increased understory competition is a primary cause for inadequate

regeneration of longleaf pine. In pine flatwoods understory competition is heightened

due to the dense shrub understory. Effective vegetation control is achieved in flatwoods

with the use of intensive mechanical and chemical site preparation methods. Although

these methods are successful in increasing pine growth, they are also detrimental to

native plant diversity. The renewed interests in restoration of longleaf ecosystems are

often guided by multiple objectives. Protecting the rich biodiversity of these unique

ecosystems is of great importance. Therefore, intensive site preparation methods may not

be appropriate in vegetation control in areas concerned with restoration. As an

alternative low intensive site preparation followed by over the top application of

herbicides is suggested.









In this study the application of three commercially available herbicides

(hexazinone, sulfometuron and imazapyr) following site preparation by single drum

chopping and burning was investigated. Pine growth and survival were measured after

one growing season. Effect on understory vegetation was assessed three and nine months

after herbicide application. In addition, fate of applied herbicides was determined 16, 90

and 240 days after treatment at 0-15, 15-30 and 45-60 cm soil depths. A bioassay was

used to determine herbicide concentrations in soil.

Herbicide applications significantly reduced longleaf seedling survival compared to

control with the exception of hexazinone. However, imazapyr treatment significantly

increased pine seedling growth compared to the control. Sulfometuron treatment

displayed a stunting effect by reducing seedling growth compared to control. Imazapyr

treatment resulted in significant control of overall shrub species, with other treatments

exhibiting no effect. Species diversity increased with imazapyr application compared to

other treatments. However, sulfometuron significantly reduced species diversity

compared to control. Hexazinone and sulfometuron dissipated to very low concentrations

240 days after treatment. Imazapyr was found in very low concentrations 16 days after

treatment and indicated persistence, as it did not dissipate rapidly with time.














CHAPTER 1
LONGLEAF PINE ECOSYSTEMS AND THEIR RESTORATION IN FLATWOODS

Longleaf Pine Ecosystems and Their Decline

Longleaf pine (Pinuspalustris Mill) ecosystems once occupied an estimated 37

million ha in the south and southeastern United States (Frost 1993). These forests

dominated the Coastal Plain areas ranging from Virginia to Texas through central Florida

(Croker 1979, Landers, Van Lear and Boyer 1995), occupying a variety of sites ranging

from xeric sandhills to wet poorly drained flatwoods (Carter et al. 1998, Brockway and

Outcalt 2000, Boyer 1990). The extent of longleaf pine ecosystems has greatly declined

since European settlement. At present, they occupy less than 2 million ha, about 5% of

the original acreage (Kelly and Bechtold 1990, Busby et al. 1995). Excessive logging in

the early 20th century cleared vast areas of virgin longleaf forests with minimal effort in

regeneration. As the demand for timber increased, land that was once under longleaf pine

was converted to other faster growing pine species such as slash pine (Pinus elliottii

Engelm.) and loblolly pine (Pinus taeda L.). Conversion of forestland in to agriculture

reduced the longleaf acreage further more.

Throughout its natural range, longleaf ecosystems are faced with many

disturbances such as periodic fires and hurricanes, which play a very important ecological

role. Through these disturbances and varying site conditions, longleaf ecosystems are

comprised of a mosaic of community types that sustains a variety of flora and fauna

(Brockway and Outcalt 2000, Croker 1979). The canopy gaps and the open midstory









created by these disturbances allows more sunlight to reach the forest floor, thus inducing

understory vegetation growth.

Longleaf ecosystems have one of the richest species diversities outside the tropics

(Peet and Allard 1993). Although the overstory is dominated by one species, the

understory is host to a plethora of plant species. The diversity among the herbaceous

plants is the main contributor to its high biodiversity. The composition of the understory

is site specific, but is mainly dominated by grass species. In the western Gulf Coastal

Plain, the understory is comprised mainly ofbluestem grasses. In Florida and along the

Atlantic Coast wiregrass (Aristida beyrichiana) is dominant, with Aristida strict

occurring from central South Carolina through North Carolina (Peet 1993). One of the

significant causes for the reduction of longleaf regeneration was the interruption of

natural fire cycles in the understory.

Role of Fire, Mechanical and Chemical Site Preparation as Restoration Techniques

Prior to European settlement, periodic fires were a frequent phenomenon in

longleaf ecosystems and were ignited by lightning and native Americans (Croker 1979).

Such frequent fires promoted longleaf dominance limiting the less fire adapted hardwood

species to more mesic sites. Fire also created a forest structure with open midstory and a

savanna like understory. The understory, mainly comprised of grasses, provided the

necessary fuels and facilitated the spread of these fires.

Understanding the role of fire and the autecology of longleaf pine is vital for the

restoration of this ecosystem. Longleaf pine is a very intolerant pioneer species (Landers

et al. 1995) and does not compete well for site resources with other more aggressive

species (Brockway and Lewis 1997). With the removal of fire, the less fire adapted shrub

species were allowed to spread into the understory. The encroaching hardwoods compete









for site resources and light with the longleaf seedlings and hinder their growth and

regeneration. Research shows that simulating the natural fire regime by applying

frequent growing season burns increases the growth and survival of longleaf pine

seedlings (Grelen 1978, 1983). Longleaf seedlings undergo an extended stem-less phase

without height initiation. This phase, also known as the "grass stage", varies in length

depending on site resources and competition and may last as long as 10-15 years.

Increased competition from the shrubs hinders height initiation of regenerating longleaf

seedlings. The shrub species also outcompete the native grass and herbaceous species for

site resources such as light. Studies show a significant increase in species diversity and

richness on sites under a frequent fire regime (Brockway and Lewis 1997). In addition,

the dangers of catastrophic fires are increased with the encroachment of shrub and other

hardwood species. The shrubs and hardwoods occupy the once open midstory and

provide ladder fuels that could cause crown fires and damage entire forests (Brockway

and Lewis 1997). In addition to competition for light some hardwood scrub species

display allelopathy against native pines including longleaf pine and native herbaceous

species (Richardson 1985).

Compared to other pine species, longleaf pine is not a prolific seeder. Longleaf

pine seeds require over three years for their physiological development (Pederson et al.

1999). Thus good seed crops are infrequent and may arise once every 4-5 years. The

seeds are large and heavy and do not disperse a great distance (Landers et al. 1995). The

short dispersal of the seeds prevents longleaf pine from colonizing and establishing in

areas far from the seed source. Longleaf seeds require a seedbed of exposed mineral soil

free of surface litter. Fire exclusion results in accumulation of forest litter that hinders









proper germination of longleaf pine seeds (Croker 1975). The number of regenerating

seedlings is further lowered through predation by non-native feral hogs. One study done

on the impact of feral hogs demonstrated that exclusion of hogs resulted in 500 seedlings

in the fenced area compared to 8 seedlings in non fenced areas (Lipscomb 1989).

Due to its rapid and continuing decline, longleaf forests are considered an

endangered ecosystem. Therefore, there is heightened interest in developing techniques

to restore longleaf pine ecosystems. Better understanding of the silvicultural

requirements of longleaf pine has enabled foresters and landowners to successfully

establish and manage longleaf forests. With the alarming increase of endangered species,

there is also great amount of interest in maintaining the biodiversity of these ecosystems.

The use of prescribed fire has greatly enhanced longleaf restoration efforts. The

use of prescribed fire is suggested prior to seed fall to improve seedbed conditions. In

areas sensitive to burning, mechanical scarification methods are suggested (Croker 1975).

Fire is also used effectively to eliminate hardwood shrubs thus releasing regenerating

pines from vegetation competition. Unfortunately, prescribed fire may not be applicable

in some sites. The use of fire in forests near residential or commercial properties bears

the risk of damaging human lives and property. Prescribed fire may only be used under

favorable weather conditions and proper authorization. Forest fires also increase the

potential for soil erosion, and prescribed fire should be used with caution on sites prone

to erosion. With concern for catastrophic crown fires, sites with heavy fuel loads should

not be burned until fuel loads are reduced by other methods. Due to such limitations of

prescribed fire, alternative vegetation control methods are used.









Understory shrub control could also be achieved through mechanical site

preparation and vegetation removal. Intensive site preparation such as disking,

harrowing and bedding has been successful in temporarily reducing understory shrub

vegetation. On sites concerned with understory restoration such as wiregrass

establishment, intensive site preparation should be limited. Research shows that

intensive site preparation may adversely effect wiregrass populations (Clewell 1989).

Excessive soil disturbance through mechanical site preparation may also increase the risk

of soil erosion. Chemical treatments in the form of herbicides have been widely used to

control understory vegetation. Herbicides are found in many forms, which differ in their

mode of action, and target species. Therefore, herbicides can be used at different stages

of stand development to selectively control undesired species.

Longleaf Pine Flatwoods Restoration

Pine flatwoods constituted a major forest type in the southeastern Coastal Plain and

occupied nearly 50% of the Florida peninsula (Davis 1967). The soils within pine

flatwoods are poorly drained saturated sandy soils with seasonal flooding (Stout and

Marion 1993). Historically, this landscape was dominated by longleaf and slash pine

depending on site conditions. Longleaf pine dominated the upland and moderate to

poorly drained sites, frequented by fires. Slash pine was limited to the wetter sites where

fires were not prevalent (Stout and Marion 1993). Flatwoods ecosystems are also

characterized as having relatively dense understories, typically comprised of shrubs such

as gallberry (Ilex glabra), saw palmetto (Serenoa repens), runner oak (Quercus pumila),

fetterbush (Lyonia lucida), blueberry (Vaccinum myrsinites) and hairy wicky (Kalmia

hirsuta) (Huck 1987).









The predominantly shrub understory amplifies the challenges faced by young

longleaf seedlings due to the increased competition for site resources. Restoration

techniques such as prescribed fire have been widely used to control understory shrub

vegetation resulting in successful establishment of longleaf pine forests on many sites

(Provencher et al 2001). However, vegetation control by prescribed fire is short-lived on

flatwood sites due to the vigorous resprouting of the shrubs. Typical shrubs and

herbaceous species of flatwoods sites have extensive rootstock systems that enable the

vegetation to resprout rapidly following fire (Abrahamson 1984). Therefore, alternative

restoration tools are required for successful establishment of longleaf pine on flatwood

sites.

Alternative vegetation control methods such as mechanical and chemical site

preparation and pine release treatments have been widely studied on flatwood sites with

slash pine (Kline et al 1994, Shiver et al 1991, Shiver et al 1990, Burger et al 1988). For

example, Shiver et al (1990) reported significant growth increases in planted slash pine

seedlings with site preparation methods such as chopping and bedding. Burger et al

(1988) compared low intensity (chopping, burning) and high intensity bladingg and

harrowing) site preparation methods on flatwood sites and reported significant early

growth gains with intensive site preparation. A majority of the flatwood studies used

bedding as a standard site preparation method for better growth. Research done by

Shiver et al (1991) on the use of picloram-triclopyr mixtures at site preparation rates

reported 70 % control of gallberry, saw palmetto, wax myrtle and blueberry species. The

study also reported 70% or greater control of blueberry species and staggerbush with

imazapyr at site preparation rates. A similar study done by Kline et al (1994) using









triclopyr and imazapyr mixtures at site preparation rates achieved good control over a

majority of the shrub species. However, both studies found that even at high rates none

of the herbicides yielded equal control on all flatwood species.

The recent increase of interest in longleaf pine restoration is with multiple

objectives such as biodiversity protection and enhancement of wildlife habitat.

Therefore, the effect of restoration activities on native plant diversity is of concern.

Mechanical site preparation such as bedding, disking and harrowing causes extensive soil

disturbances. Although these methods are very effective in providing better growth gains

among planted pine seedlings, they are detrimental to the native plant diversity. For

instance, wiregrass a keystone species within longleaf pine ecosystems is very sensitive

to ground disturbance. According to Clewell (1989), wiregrass is a poor seed producer

and relies on vegetative propagation for its expansion. Wiregrass is also easily uprooted

and once uprooted it is rarely successful in propagation. Therefore, intensive site

preparation should not be used in areas concerned with understory restoration with

wiregrass.

Information regarding the effect of herbicides on the groundcover vegetation of

natural flatwoods and sandhill communities is scarce (Litt et al 2001). An extensive

literature review done by Litt et al (2001) on published research regarding herbicide

effects on groundlayer vegetation found only 3 and 7 studies, respectively, on natural

flatwoods and sandhill sites. A study done by Neary et al (1990) on a flatwood site in

Florida found nearly 75% reduction in species richness following an intensive weed

management regime with repeated application of sulfometuron, glyphosate and triclopyr

treatments and mowing. Brockway et al (1998) studied the effects of hexazinone applied









at 1.1 and 2.2 kg/ha with different application techniques on a sandhill site. They

reported significant decrease in the cover of oak species and shrub control at the higher

rate. Moreover, the higher rate application of hexazinone induced an 86% increase in

biomass of wiregrass in the first growing season after treatment. With the broadcast

application of hexazinone, a short-term reduction of forb species was also reported. The

continuance of herbicide effects appears to be dependent on the rate and method of

application. However, most treatment applications seem to cause only short-term

changes in species diversity and composition. A long-term study of herbicide effect on

understory plant diversity and richness was conducted by Boyd et al (1995) with one-

time broadcast applications of imazapyr, glyphosate, and hexazinone. The reported

results 7 years after treatment showed no significant herbicide effect on species diversity

and richness.

Fate of Applied Herbicides

The application of herbicides creates a contamination risk of ground and surface

water bodies. Extensive research has been conducted on the offsite movement of applied

herbicides and the risk of contamination. Studies done by Michael and Neary (1993)

with multiple herbicides on industrial forestry sites near a watershed reported levels of

contamination less than the Health Advisory Levels (HAL). On some sites, herbicide

concentrations exceeded HAL and drinking water standards when herbicide applications

were directed onto surface water bodies, but did not persist for an extended period of time

(Michael 2000, Neary et al. 1996). The groundwater level in flatwoods sites is prone to

seasonal fluctuations and may rise as high as the soil surface level. On such sites, the

application of herbicide may seem to involve high risk. In contrast, a study done on a









flatwoods site by Neary et al. (1989), reported low levels of offsite movement and

contamination of ground water using sulfometuron methyl.

Phytotoxic effects of herbicide residue on non-target species are also of concern.

Herbicide residue on agricultural fields has displayed phytotoxic effects on rotational

crops (Brewster and Appleby 1985). Movement and persistence of forestry herbicides

depend on weather, edaphic conditions, herbicide characteristics, method and time of

application and site characteristics (Norris 1981).

The majority of aforementioned research involved intensive mechanical and

chemical site preparation treatments. Such methods may not be appropriate on most

flatwoods sites with high water table and where multiple management objectives,

including understory species richness and diversity, are of great concern. As a shrub

control method in flatwoods, low-rate, over the top application of herbicide following

less intensive site preparation could serve s an alternative to high intensity site

preparation. In our study, three commonly used forestry herbicides, hexazinone,

sulfometuron and imazapyr were used and their efficacy was examined. These herbicide

treatments were applied following single drum chopping and prescribed burning. The

three major research objectives addressed in this study were:

* Determine the extent of pine seedling response

* Quantify the fate of applied herbicides in the soil

* Determine the effect of applied herbicides on understory species diversity and

composition














CHAPTER 2
LONGLEAF PINE SEEDLING AND UNDERSTORY VEGETATION RESPONSE TO
HERBICIDAL VEGETATION CONTROL AND FATE OF APPLIED HERBICIDES
IN FLATWOODS

Introduction

Pine Flatwoods

Pine flatwoods constituted a major forest type in the southeastern Coastal Plain

and occupied nearly 50% of the Florida peninsula (Davis 1967). These landscapes were

mostly dominated by longleaf pine (Pinuspalustris Mill.) or slash pine (Pinus elliottii

Engelm.) depending on site conditions. Longleaf pine dominated the upland and

moderate to poorly drained sites, frequented by fires. Slash pine was limited to the wetter

sites where fires were not prevalent (Stout and Marion 1993). Due to excessive logging

and inadequate regeneration, the extent of longleaf pine ecosystems including flatwoods

has greatly declined since European settlement. At present, longleaf pine forests occupy

less than 2 million ha, about 5% of the original acreage (Kelly and Bechtold 1990, Busby

et al. 1995).

Due to the high demand for timber and the rise of plantation forestry, many

longleaf dominated sites were converted to slash pine and loblolly pine (Pinus taeda L)

(Croker 1979). Longleaf pine was replaced by other pines because of the slow early

growth characteristic of the species. It is a poor seed producer with infrequent seed crops

(Boyer 1990, Pederson et al. 1999) and requires scarified seedbed with exposed mineral

soil for adequate germination (Croker 1979). Once established, longleaf seedlings

exhibit a slow growth phase with little to no height initiation for several years. This slow









juvenile growth phase, also known as the "grass stage" is extended with increased

competition for site resources (Haywood 2000, Jose et al. 2003). Such challenges are

exacerbated on flatwoods sites due to the characteristic heavy understory of shrub

species. These understories are typically comprised of shrubs such as gall berry (Ilex

glabra L.), saw palmetto (Serenoa repens Bartr.), runner oak (Quercuspumila Walt.),

fetterbush (Lyonia lucida Lam.), blueberry (Vaccinum myrsinites Chapman.) and hairy

wicky (Kalmia hirsuta Walt.) (Huck 1987).

Vegetation Control and Pine Response

Prescribed fire has been widely used to control understory woody vegetation

resulting in successful establishment of longleaf pine forests on many sites (Provencher et

al. 2001). Vegetation control by prescribed fire is short-lived on flatwoods sites due to

the vigorous resprouting of the vegetation. Typical shrubs and herbaceous species of

flatwoods sites have extensive rootstock systems that enable the vegetation to resprout

rapidly following fire (Abrahamson 1984).

Alternative vegetation control methods such as mechanical and chemical site

preparation and pine release treatments have been widely studied on flatwoods sites with

slash pine (Kline et al. 1994, Shiver et al. 1990, 1991, Burger et al. 1988). For example,

Shiver et al. (1990) reported significant growth increases in planted slash pine seedlings

with site preparation methods such as chopping and bedding. Burger et al. (1988)

compared low intensity (chopping, burning) and high intensity (burning, blading and

harrowing) site preparation methods on flatwoods sites and reported significant early

growth gains with intensive site preparation. A majority of the flatwoods studies used

bedding as a standard site preparation method for better growth. Research done by

Shiver et al. (1991) on the use of picloram-triclopyr mixtures at site preparation rates









reported 70 % control of gallberry, saw palmetto, wax myrtle and blueberry species. The

study also reported 70% or greater control of blueberry species and staggerbush with

imazapyr at site preparation rates. A similar study done by Kline et al. (1994) using

triclopyr and imazapyr mixtures at site preparation rates achieved good control over

majority of the shrub species. However, both studies found that even at high rates none

of the herbicides yielded equal control on all flatwoods species. The importance of site

specific applications of herbicide treatments depending on the prevalent shrub species

was emphasized by both studies.

The aforementioned research on methods of vegetation control conducted on

industrial forestry sites with slash pine may be applicable to the establishment of longleaf

pine stands in flatwoods. However, the recent interests in restoring longleaf pine

ecosystems are with multiple objectives in addition to increased seedling growth. In

areas concerned with native biodiversity, sensitive plant populations and groundwater

quality, intensive site preparation and high-rate herbicide applications may not be

applicable. For example, wiregrass is reported as very sensitive to site disturbance.

Therefore, on sites concerned with wiregrass regeneration intensive site preparation

methods should be avoided (Clewell 1989).

Fate of Applied Herbicides

The offsite movement and persistence of applied herbicides and the risk of

contamination has been widely studied. Movement and persistence of forestry herbicides

depend on weather, edaphic conditions, herbicide characteristics, method and time of

application and site characteristics (Norris, 1981). Studies done by Neary and Michael

(1989) with sulfometuron methyl on a flatwoods site reported low levels of off-site

movement and contamination of ground water. The study also reported that the high









acidity (average pH = 4) of flatwoods sites further impedes the mobility of herbicides

such as sulfometuron methyl. Michael and Neary (1993) reported on the off site

movement of hexazinone, imazapyr, picloram and sulfometuron on industrial forestry

sites in the south. Results from 23 studies showed levels of contamination less than the

Health Advisory Levels (HAL) for all herbicides tested. On some sites, herbicide

concentrations exceeded HAL and drinking water standards when herbicide applications

were directed onto surface water bodies, but did not persist for an extended period.

Phytotoxic effects of herbicide residue on non-target species are also of concern.

Herbicide residue on agricultural fields has displayed phytotoxic effects on rotational

crops (Brewster and Appleby 1985). Thus, persistent herbicides may impact seasonal

variation in plant communities. Due to site specificity of herbicide behavior, care should

be taken in estimating herbicide movement and persistence based on data from other sites

(Michael 2000).

Research on herbicidal vegetation control on flatwoods sites has focused primarily

on pre-plant site preparation treatments. Information regarding the effect of over-the-top

herbicide applications on longleaf pine seedlings and the understory vegetation of

flatwoods sites is scarce. This study examined use of three commercially available

herbicide treatments (hexazinone, sulfometuron and imazapyr) following low-intensive

site preparation. The specific questions researched were:

* What is the response of longleaf pine seedling survival and growth to the applied

treatments?

* What is the impact of herbicide application on major understory species foliar

cover and density?









What is the mobility and persistence of the applied herbicides in flatwoods soils?



Materials and Methods

Study Area Description

This study was conducted on a flatwoods site at the Point Washington State

Forest in Walton County, Florida (30020' 16.04" N, 860 4' 19.22" W). Average annual

high and low temperatures were 25.50C and 120C respectively. Annual precipitation was

about 1640mm (2002) with most received in the late summer months (Figure 2-1). Soils

were of low pH (pH <5) and sandy texture with low nutrient content (Table 2.1). The

study area soils were mapped as sandy, siliceous, thermic aeric alaquods belonging to the

Leon series, which is characterized by deep, poorly to very poorly drained soils. Soils of

the flatwoods pinelands are formed on sandy quaternary formations derived from marine

deposits (Stout and Marion 1993).

Prior to study establishment the overstory was a planted slash pine stand with

intermittent longleaf pine saplings. The average age of the stand was 26 years with a

basal area of 1.85 m2 and an average dbh of 19.1 cm. The understory was comprised

mainly of species such as gallberry, saw palmetto, runner oak, dangleberry (Gaylussacia

frondosa L.), hairy wicky, wiregrass (Aristida beyrichiana Trin.and Rupr.) and bluestem

grasses. Dormant season fires were used on a three-year burn cycle to mitigate fuel build

up.

Experimental Design and Treatments

Prior to site preparation, the overstory was harvested (August 2001). The harvest

debris and the understory were roller chopped once and prescribed burned in October

2001. A randomized complete block design was used to examine the effects of herbicidal









vegetation control methods on pine seedling growth and survival. The study incorporated

six blocks with five treatment plots within each block. All treatment plots were 36.6m x

24.4 m, including a > 3m buffer strip between plots. In December 2001, one year old

containerized longleaf pine seedlings were hand-planted at 3.1m x 1.8m spacing.

Seedlings were planted in rows to facilitate the application of treatments. Each treatment

plot included 100 seedlings amounting to 3000 seedlings for the entire study. In March

2002, four herbicide treatments [Sulfometuron (0.26 ai kg/ ha), Hexazinone (0.56 ai kg/

ha), Sulfometuron (0.26 ai kg/ ha) + Hexazinone (0.56 ai kg/ ha) mix, Imazapyr (0.21 ae

kg/ ha)] were applied in a 1.2 m band over the top of seedlings via a knapsack sprayer. In

each block, one treatment plot was kept herbicide free as a control.

Measurements

Pine survival and growth

Pine survival was monitored six and 12 months (respectively, June 2002 and Dec

2002) after planting. Pine growth was measured after the first growing season (Dec

2002). Seedling height and root collar diameters (RCD) were measured as growth

parameters on all planted seedlings. Seedling height was measured using a ruler, from

soil surface to the top of the bud and RCD was measured using a digital caliper. Stem

volume index (SVI) was calculated with the measured RCD and height data. Initial RCD

of seedlings were also recorded prior to planting as baseline data. Within our study, post-

planting burial was observed among the majority of the dead seedlings. In addition to

growth measurements, the extent of post planting burial was estimated for each seedling.

Vegetation control assessment

A preliminary vegetation survey was conducted (June 2001) prior to overstory

harvest and site preparation to assess the initial presence and percent cover of understory









species. Following site preparation and herbicide application, two vegetation surveys

were conducted. These surveys were done six and eleven months (June and Nov 2002)

after pine planting. In each plot six randomly selected 1m2 quadrats were sampled in the

herbicide treated bands and were revisited for subsequent surveys. Percent cover was

ocularly estimated for all species using the modified Daubenmire scale (Daubenmire

1959). Stem number and average stem heights were recorded for all shrub species.

Herbicide dissipation assessment

Soil was sampled over time in the hexazinone, sulfometuron and imazapyr plots

in three blocks to monitor the persistence and movement of the herbicides. Soil samples

were collected 16, 90 and 240 days after treatment. In each treatment plot, six random

points within treated bands were sampled and the general location was revisited for

subsequent sampling. Samples were collected at 0-15, 15-30 and 45-60cm depths. In

each block, soil collected within a treatment plot was composite by depth to a single

sample. Samples were kept frozen until the time of analysis. A bioassay was used to

determine the herbicide concentration in soil samples (Ramsey et al 2004). Standard

curves were established for each herbicide by subjecting brown top millet (Panicum

ramosum L.) seeds to a known concentration of each herbicide. Three herbicide

concentration ranges, 1-10 ppb at 1 ppb interval, 10-100 ppb at 10 ppb interval and 100-

1000 ppb at 100 ppb interval were tested. The seeds were grown in a growth chamber

with 16 hours of daylight (1800 mrnol m-2) at 3010C and 8 hours of darkness at 151C.

Growth chamber conditions were set as described by Hemandez-Sevillano et al. (2001).

Plants were harvested after 13 days and dry plant weight was measured. The field soils









was thawed at room temperature and brown top millet seeds were grown under the same

growing conditions as the standards.

Statistical Analysis

Pine survival and growth

Pine survival, RCD, height and SVI data after one growing season were analyzed

using analysis of variance (ANOVA) within the framework of a randomized complete

block design (RCBD) using SAS version 8.0 (SAS 2000). Significant treatment effects

(a=0.1) were separated using Duncan's multiple range test. The initial RCD and the

extent of post planting burial were used as covariates in the ANOVA model.

Vegetation control

The effect of treatments on stem counts and heights of major shrub species were

analyzed using ANOVA for a randomized complete block design. Pre harvest uniformity

of stem counts for each shrub species were tested prior to analysis. Only those shrub

species with uniform stem counts prior to treatment application were considered for

analysis. Percent control on shrub stem number and heights were calculated. Significant

differences (a=0.1) between treatments were detected by Duncan's multiple range test.

ANOVA was not used in the analysis of percent cover data, as it did not conform

to the assumption of normality. Therefore, the Kruskal-Wallis test was used with PROC

NPAR1WAY as a nonparametric alternative to ANOVA. Pre harvest percent cover

uniformity of shrubs, grasses, forbs and major understory species were tested. Treatment

effects on overall percent cover changes in shrubs, grasses and forbs were analyzed for

both survey dates. Five major understory species that showed pre harvest uniform

percent cover were used to test for treatment effects on percent cover of individual

species. Percent control was calculated for each vegetation class and major species.









Treatment plots were compared to the untreated control plots to establish significant

differences (a=0.1).

Herbicide persistence analysis

A power model was developed to predict herbicide concentration (Y) using the

plant dry weight (X) as follows:

Y= aXb

The quality of fit of the model was tested for each herbicide by calculating R2

value. Dry whole plant weight from seeds grown in the herbicide contaminated field soil

was used in the power model to predict the herbicide concentrations in the field soil. The

predicted concentrations were analyzed using ANOVA for differences in concentrations

between sampling days and depths. Duncan's multiple range test was used to detect

significant differences between days and depths at a significance level of a=0.1.

Results

Pine Survival

Analysis of variance revealed a significant treatment effect on pine seedling

survival after one growing season (P<0.0001). The control and hexazinone treatments

had the highest survival (85.3% and 85% respectively) and were significantly greater

than sulfometuron (78.3%), sulfometuron+hexazinone mix (67.8%) and imazapyr

(65.6%) treatments (Fig 2-2a).

Pine RCD, Height and SVI

Analysis of first growing season RCD data revealed a significant treatment effect

on RCD growth (p<0.0001). Initial RCD and extent of post-planting burial were

significant as covariates (p <0.0001). Imazapyr treatment yielded the highest RCD

(12.17mm) and was significantly larger than all other treatments. The RCD in









hexazinone (11.75mm) and sulfometuron-hexazinone mix (11.25mm) treatments showed

significant differences between the two treatments but did not differ from the control

(11.65mm). Sulfometuron treatment resulted in the lowest RCD growth (10.80mm)and

was significantly lower than all other treatments (Fig 2-2b).

Analysis of height data showed significantly taller seedlings in the imazapyr

(1.35cm) and sulfometuron+hexazinone mix (1.29cm) compared to other treatments.

Hexazinone (1.18cm) did not significantly affect seedling height compared to the control

(1.10cm). Sulfometuron (1.04cm) again resulted in lower growth and had the lowest

height compared to all other treatments (Fig 2-2c).

The SVI showed a similar trend as RCD and height. Imazapyr treatment resulted

in the largest mean SVI of 2.46cm3 and was significantly greater than all other

treatments. Sulfometuron (1.59cm3) yielded lowest SVI compared to all other treatments.

Hexazinone and sulfometuron+hexazinone mix treatment did not significantly affect SVI

compared to the control (Fig 2-3).

Vegetation Control

Imazapyr resulted in significant overall reduction of the percent cover of shrubs in

both 3 MAT and 9 MAT surveys (66% and 59% control respectively). None of the other

herbicide treatments was effective in reducing the overall percent cover of shrubs.

Overall, percent cover of grasses and forbs were not reduced by any of the treatments.

Analysis of major flatwoods species showed no treatment effect on wiregrass and

saw palmetto percent cover. Imazapyr was successful in significantly reducing the

percent cover of gallberry in the June 2002 survey (78% control) and showed sustained

control through November 2002 survey (70% control). Gallberry percent cover was

unaffected by the other treatments except in sulfometuron+hexazinone mix treatment









where it increased in both surveys. None of the treatments reduced the percent cover of

runner oak by June 2002 survey. However runner oak exhibited a delayed response to

hexazinone (37% control), imazapyr (69% control) and sulfometuron+hexazinone mix

(46% control) treatments as its percent cover was significantly reduced in the Nov 2002

survey. Bluestem species were not adversely affected by the applied treatments. The

percent cover of bluestem species increased in the sulfometuron treatment plots by the

Nov 2002 survey.

Shrub stem counts and heights were analyzed for gallberry, runner oak and saw

palmetto species (table 2-3). Gallberry and runner oak stem counts and heights were

unaffected by the herbicide treatments except for imazapyr. Imazapyr significantly

reduced (58%) the amount of resprouting stems of gallberry in June 2002 survey. An

increased amount of control of gallberry was found (79%) in Nov 2002 and a significant

reduction in the number of runner oak stems (65%). The effect on stem heights followed

a similar trend with only imazapyr yielding reduced stem heights. In the June 2002

survey, imazapyr significantly reduced the stem heights of gallberry and runner oak

species (64% and 40% respectively). The significant reduction in stem heights of

gallberry and runner oak species by imazapyr was sustained in the Nov 2002 survey

(37% and 29% respectively). None of the herbicide treatments significantly affected the

stem counts or heights of saw palmetto species.

Herbicide Dissipation

Three predictive power models were developed for imazapyr, sulfometuron and

hexazinone with pseudo R2 values of 0.89, 0.65 and 0.60 respectively. The analysis of

the predicted concentrations for the herbicide contaminated soils revealed no significant

difference among soil depths and days (Fig 2-4). Herbicide concentrations of the entire









sampled soil profile were analyzed over the three sampling dates. Hexazinone and

sulfometuron showed significant decrease in herbicide concentrations between 16 DAT

and 240 DAT. Imazapyr concentrations were not significantly different among the three

dates (Fig 2-5).

Discussion

Successful establishment, survival and improved growth of pine seedlings on

flatwoods sites rely on the efficacy of applied vegetation control methods. The majority

of the studies reporting improved survival and growth were associated with intensive site

preparation and high rate pre-plant herbicide applications which provide significant

release from competition stress (Shiver et al. 1991, Kline et al. 1994).

Within our study, overall survival of seedlings was moderate at best including the

control. Post planting burial observed among the majority of the dead seedlings may

have contributed to the overall seedling mortality. Metcalfe and Cantrell (1986) studied

the effect of sulfometuron and sulfometuron+hexazinone treatments on first year survival

of longleaf seedlings planted on a sandy soil in Florida. In their study with higher rates

of sulfometuron and sulfometuron+hexazinone mix treatments, no significant effect on

seedling survival was found compared to the control. Our results are contrary to the

above findings as sulfometuron and sulfometuron+hexazinone mix treatments

significantly lowered seedling survival compared to the control. Although imazapyr and

sulfometuron+hexazinone mix treatments resulted in lowest survival, they both

significantly increased seedling height with imazapyr also increasing RCD growth. Our

results reveal an evident "trade off" in benefits of using herbicide applications such as

imazapyr against control, providing forest managers the choice between increased

survival of pine seedlings or increased growth depending on their silvicultural objectives.









Application of sulfometuron displayed a stunting effect on seedling growth by yielding

the lowest RCD and height. Our findings agree with previous research done by Gjerstad

et al. (1983) where they reported detrimental effects such as stunting of pines by

sulfometuron in sandy soils.

Analysis of understory vegetation revealed significant reduction in foliar cover of

shrubs by imazapyr treatment. However, imazapyr did not reduce the foliar cover of

grasses and forb species. Runner oak displayed a delayed response to hexazinone,

sulfometuron+hexazinone mix and imazapyr treatments. The benefits of runner oak

percent cover reduction may be realized the following growing seasons. Although

imazapyr had poor control over grasses and forbs, it resulted in increased growth of

longleaf pine seedlings by controlling the shrub component alone. This result suggests an

important competitive relationship between the flatwoods understory shrub species and

planted longleaf seedlings. Our observations reiterate the importance of effective shrub

control for improved growth of longleaf seedlings on flatwoods sites. Imazapyr was the

only treatment to effectively reduce the resprouting of major flatwoods shrub species

such as gallberry and runner oak by reducing the stem number and their heights.

Interestingly, none of the treatments reduced the percent cover of wiregrass. Thus, in

flatwoods areas concerned with restoration of longleaf-wiregrass ecosystems, the tested

herbicide treatments at the specified rates may be applied for longleaf pine establishment

without any detrimental effects on wiregrass.

Bioassays are widely used to detect herbicide concentrations in soil. With the

existing limitations of analytical methods due to inefficient residue extraction methods,

bioassays provide a good alternative (Hernandez-Sevillano et al. 2001). They are









effectively used to detect biologically active levels of herbicide residue at lower

concentrations in soil (Anon 2000). Therefore, a bioassay is best suited for our study due

to the relatively low herbicide rates applied. A study done by Hollaway et al. (1999)

comparing different detection methods reported that bioassays were more accurate than

high performance liquid chromatography (HPLC) when used to detect residues at lower

concentrations. Many studies have also reported on the efficacy of bioassays in detecting

herbicide residue concentrations as small as 1 ppb (Stork and Hannah 1996, Hollaway et

al. 1999, Hemandez-Sevillano et al. 2001).

The three predictive models developed through the bioassays had relatively high

pseudo R2 values indicating good predictive strength. However, bioassay results did not

show significant differences in herbicide concentrations between soil depths for all

herbicides. Although nonsignificant, the results reveal a trend in herbicide movement as

concentrations increase in lower depths with time (fig 2.4). Lack of significance may be

attributed to the large amount of variation encountered within the bioassay. For future

studies, additional field sampling and increased replications within the bioassay are

suggested to account for high variation. Herbicide dissipation analysis over the entire

sampled soil profile reveals a significant decrease in herbicide concentrations of

hexazinone and sulfometuron between 16 DAT and 240 DAT. Imazapyr was persistent

in the soil at low concentrations for an extended period of time. Soil adsorption of

imazapyr is improved with increased organic matter and low pH in soil (Dickens and

Wehtje 1986). The accumulation of debris from the site preparation bum and low pH of

the site may have contributed to the low predicted imazapyr concentrations by increasing









its adsorption to soil. Herbicide concentrations less than 50 ppb were found for all three

herbicides 240 DAT.

Conclusion

Within the first growing season after treatment application imazapyr treatment

resulted in a significant decrease in seedling survival. However, it resulted in greater

overall growth by significantly increasing RCD and height of seedlings. Imazapyr also

resulted in significant reduction of shrub foliar cover, stem density and stem heights. The

decrease in understory dominance of shrubs with imazapyr induced increased growth of

longleaf seedlings. Application of hexazinone, sulfometuron and

sulfometuron+hexazinone mix treatments failed to yield significant vegetation control

and improve overall pine seedling growth and survival. Sulfometuron resulted in lowest

RCD growth and its application is not recommended on sandy flatwoods soils.

Applied herbicides did not significantly impact the understory grasses and forb

species. More importantly, none of the herbicide treatments was detrimental or

augmented wiregrass foliar cover. However, with the decline of shrub dominance the

potential for improvement in wiregrass cover is increased.

Hexazinone and sulfometuron do not seem to persist for extended periods.

Imazapyr was present in very low concentration soon after application and exhibited a

slower dissipation. However all three herbicides dissipated to very low concentrations

less than 75ppb, 240 days after treatment.

Overall, imazapyr provided the best-desired results with significant increase in

pine growth and better control of shrub species with no significant effect on other

understory species.










400

350

300
E 250
E
200

150

100

50
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
month

Figure 2-1. Monthly rainfall data for year 2002 at Point Washington State Forest, Walton
County, Florida.











(a)
100
a a
80

E 60

O 40

20
imz hexa sulfo sulfo+ hexa Con


(b)
12.5

12.0 i b

E 11.5 c

0d
11.0

10.5

10.0 ,
irz hexa sulfo sulfo+ hexa con

(c)
1.6
a

1.2 c b


0.8


0.4 ,
imz hexa sulfo sulfo+ hexa Con
treatment



Figure 2-2. Longleaf pine seedling survival and growth. (a) mean survival, (b) mean root
collar diameter (RCD), (c) mean seedling height by treatment. Means
followed by the same letter are not significantly different (ca=0.1). (imz:
imazapyr; hexa: hexazinone; sulfo: sulfometuron; sulfo+hexa:
sulfometuron+hexazinone mix; con: control)














3.5
3.3
3.0
2.8 a
^ 2.5-
E ab ab
12.3 b
2 2.0
1.8 i
1.5
1.3
1.0
imz hexa sulfo sulfo+hexa con
treatment

Figure 2-3. Stem volume index of longleaf seedlings after one growing season by
treatments. Means associated with the same letters are not significantly
different (a=0.1). (imz: imazapyr; hexa: hexazinone; sulfo: sulfometuron;
sulfo+hexa: sulfometuron+hexazinone mix; con: control)












hexazinone
250

S200 a

C 150 T a

100 -
loo-oo

o 50

0

7.5 22.5 52.5


irmazapyr
60
E16 DAT
50 0
o 40 a
0a a

5 20 a




7.5 22.5 52.5


sulfometuron
900
800 16 DAT a
S700
600
:-,
500
400
S300
Saa a
200 a a a
o 100
0
7.5 22.5 52.5

soil depth (cm)





Figure 2-4. Mean herbicide concentrations in part per billion (ug kg -1) by depth. Values
associated with the same letters are not significantly different (a=0.1).
Significance established with duncan's multiple range test.











300
a
250_ 16 DAT
250
O 90 DAT
C. 240 DAT
a. 200
a
% 150 ab
S aa b
a
o 100
o b
50 -
b a a a


hexazinone sulfometuron imazapyr
herbicide

Figure 2-5. Change in herbicide concentration over time across soil profile (60 cm).
Concentration values followed by different letter are significantly different
from other dates for the respective herbicide. Significance declared at ca=0.1
using Duncan's multiple range test.






30


Table 2-1. Soil description at four depths of Point Washington State Forest, Walton
County, Florida.
Soil property Soil depths (cm)
0-15 15-30 30-45 45-60
Texture sandy sandy sandy sandy
Soil pH 4.40 4.60 4.80 4.60
CEC (meq/100g) 3.80 2.20 2.50 2.50
Organic matter (%) 1.38 0.62 0.58 0.37
Phosphorus (kg ha-1) 16.67 4.44 4.44 4.44
Potassium (kg ha-1) 42.23 14.44 10.00 8.89
Magnesium (kg ha-) 60.02 12.22 10.00 8.89
Calcium (kg ha-) 330.11 50.01 44.46 33.33










Table 2-2. Mean foliar cover for major understory species and the total foliar cover of
vegetation types by treatment.


Control


Species
Wiregrass
June 2002
Nov 2002


Bluestem spp
June 2002
Nov 2002

Gallberry
June 2002
Nov 2002

Runner oak
June 2002
Nov 2002

Saw palmetto
June 2002
Nov 2002

Veg-type
All shrub
June 2002
Nov 2002

All grasses
June 2002
Nov 2002

All forbs
June 2002
Nov 2002


Sulfometuron


4.50+ 0.9 6.02+ 1.1
11.5+2.5 13.97+2.7


0.96+ 0.2
4.75+ 1.8


9.15+ 1.7
14.20+ 2.1


19.09+ 4.6
29.21+ 4.2


11.22+ 4.1
10.83+ 3.9


50.7913.6
68.6510.1


11.462.2
26.002.7


1.05+ 0.3
2.04+ 0.5a


10.32+ 1.2
15.21+ 2.2


21.60+ 4.9
32.20+ 5.4


8.04+ 2.3
11.50+ 2.4


44.5811.6
67.91+12.4


8.081.7
18.633.9


4.601.5 3.020.6
8.364.7 7.101.0


Hexazinone


6.5+ 1.4
13.65 2.6


1.08+ 0.2
3.80+ 0.6


10.911.5
17.13+ 1.9


12.55+ 3.3
18.50+ 3.2a


8.70+ 2.3
9.62+ 2.4


34.284.9
57.26.1


8.991.0
24.925.8


4.221.2
6.221.2


Sulfo+Hexa1


Imazapyr


5.00+ 1.0 5.81+ 1.7
12.86+2.1 12.40+ 1.4


1.00+ 0.2
5.16+ 1.3


13.34+ 1.9 a
19.12+ 2.4 a


10.89+ 1.5
15.96+ 2.0 a


7.94+ 4.1
13.64+ 4.5


36.936.8
60.426.7


7.931.7
20.653.6


3.101.3
6.502.3


0.50
3.83+ 1.2


2.10+ 0.4 a
4.31+ 0.9 a


6.12+ 1.6
9.25+ 2.0 "


11.30+ 3.6
10.75+ 2.1


23.264.2a
34.41+2.6"


7.51.4
22.483.4


5.522.3
12.324.1


Note: Values followed by letter "a" are significantly different from the control along
rows at a =0.1.
1 Sulfometuron+Hexazinone mix treatment












Table 2-3. Stem numbers and stem heights for three understory shrub species by treatment.
Shrub stem numbers (stems/ m2)
Species Control Sulfometuron Hexazinone Sulfo+Hexa1 Imazapyr


Gallberry
Runner oak
Saw palmetto


Gallberry
Runner oak
Saw palmetto

Species


Gallberry
Runner oak
Saw palmetto


18.37+ 2.6 20.60+ 2.7
31.40+ 5.3 31.29+ 5.9
2.20+ 0.5 2.45+ 0.4


34.50+ 4.9
52.07+ 7.8
2.75+ 0.5

Control


3 MAT
16.93+ 1.5
23.14+ 3.9
2.00+ 0.4
9 MAT


39.80+ 5.3 38.33+ 3.7
59.20+ 12.1 37.57+ 6.4
3.44+ 0.8 2.00+ 0.6
Shrub heights (cm)
Sulfometuron Hexazinone


18.89+ 1.1 18.64+ 0.7
22.92+ 1.9 a b 22.66+ 1.6 a
36.00+ 5.5 30.45+ 5.0


3 MAT_
20.57+ 1.0
19.40+ 1.4 b
51.20+ 6.3
9 MAT


Gallberry 25.24 1.5 26.71 1.4 28.16 1.5
Runner oak 29.92 2. a 26.37 2.2a b 26.00+ 1.7a b
Saw palmetto 24.81+ 4.2 26.11+ 5.9 35.75+ 9.0
Note: values followed by different letters are significantly different along rows (ca=0.1)
1 Sulfometuron+Hexazinone mix treatment


16.60+ 1.8
20.32+ 2.5
2.55+ 0.6


43.00+ 5.9
36.57+ 4.7
3.28+ 1.3

Sulfo+Hexa1


19.52+ 1.1
18.00+ 0.8 b
25.11 4.8


27.66+ 1.6
23.82+ 1.3b
30.85+ 6.1


7.60+ 0.9 b
17.62+ 3.6
2.90+ 0.6


7.03+ 1.6 b
18.04+ 3.3 b
3.33 0.8

Imazapyr


6.72 0.4 b
13.66+ 0.2c
36.10+ 4.0


15.85+ 1.6 b
21.16+ 1.8 c
30.25+ 5.4














CHAPTER 3
UNDERSTORY SPECIES DYNAMICS FOLLOWING HERBICIDAL PINE
RELEASE TREATMENTS ON A LONGLEAF FLATWOODS SITE



Introduction

Prior to European settlement in the United States, the southeastern coastal plain

was dominated by large tracts of longleaf pine forests, covering an estimated 37 million

ha (Frost. 1993). These forests ranged from southeastern Virginia to eastern Texas

through central Florida (Croker 1979, Landers et al. 1995), occupying a variety of sites

ranging from xeric sandhills to wet poorly drained flatwoods (Boyer 1990, Carter et

al.1998). At present, longleaf acreage has been reduced to less than 5% of its former

range with an estimate of less than 2 million ha remaining (Kelly and Bechtold 1990,

Busby et al. 1995).

The longleaf pine range within the southeastern Coastal Plain is prone to many

natural disturbances such as tropical storms and forest fire (Delcourt et al 1993).

Coupled with these disturbances and varying site conditions, longleaf ecosystems

included a mosaic of community types that sustained a high diversity of flora and fauna

(Croker 1979, Brockway and Outcalt 2000). Longleaf pine ecosystems are well adapted

to such natural calamities and were dependent on them to maintain their structure and

functions. The frequent fires that swept through these forests prevented the

encroachment of less fire-adapted hardwood species and maintained an open midstory.

These natural disturbances coupled with competition intolerance of longleaf pine created









a widely spaced overstory. Such structural characteristics of these forests facilitated

increased light transmitance facilitating understory plant growth.

Longleaf ecosystems have one of the richest species diversities outside the tropics

(Peet and Allard 1993). Although the overstory is mostly monospecific, the understory is

host to a plethora of plant species. The composition of the understory is site specific and

mainly dominated by grass species (Outcalt 2000). In the western gulf Coastal Plain, the

understory is comprises mainly ofbluestem grasses. In Florida and along the Atlantic

Coast wiregrass (Aristida beyrichiana Trin and Rupr) is dominant with Aristida strict

Michx., occupying from central South Carolina through North Carolina (Peet 1993).

Some of the common forb species include Asclepias spp., Aster spp., Liatris spp., Rhexia

spp., Carphephorus spp. and Eupatorium spp. among others. (Platt et al. 1988, Huck

1987). On the more mesic flatwoods sites the understory is comprised of a variety of

shrub species such as gallberry (Ilex glabra L.), runner oak (Quercuspumila Walt.),

fetterbush (Lyonia lucida Lam.), saw palmetto (Seronoa repens Bartr.), blueberry

(Vaccinum myrsinites Chapman.) and hairy wicky (Kalmia hirsuta Walt.) (Huck. 1987).

The forests of the southeastern Coastal Plain also include a large number of rare plants

(Collins et al. 2001).

With the decline of these forests and changes in land use and development

activities, some of the rare and native plant populations were fragmented and eliminated

reducing genetic diversity (Collins et al. 2001). Due to the rapid loss of these forests and

the associated highly diverse understory, its restoration has gained importance. In

restoration efforts focus on controlling the understory shrub dominance has gained

importance. Shrub dominance increases resource competition stress of longleaf pine









seedlings and suppresses the ground-layer understory vegetation. On flatwoods sites, the

competition for site resources and shrub dominance is elevated due to the characteristic

dense understory. Prescribed fire has been widely used to control understory vegetation

resulting in successful establishment of longleaf pine forests on many sites (Provencher et

al. 2001). Vegetation control by prescribed fire is short-lived on flatwoods sites due to

the vigorous resprouting of many species. Typical shrub and herbaceous species of

flatwoods sites have extensive rootstock systems that enable them to resprout rapidly

following fire (Abrahamson 1984).

The uses of mechanical and chemical site preparation and release treatments are

suggested as alternatives to reduce the understory dominance of shrub species. Although

intensive mechanical site preparation provides good shrub control (Burger et al. 1988), it

also causes extensive ground disturbance that may be detrimental to desired native

species. For instance, wiregrass, one of the keystone understory species of longleaf pine

ecosystems, is reported to be very sensitive to ground disturbance (Clewell 1989).

Chemical vegetation control methods are widely used as site preparation and pine

release treatments. Herbicide applications at high rates (site preparation) were studied by

Kline et al. (1994) and Shiver et al. (1991) on industrial flatwoods sites, with significant

reduction of the understory shrub component. Information regarding the effect of

herbicides on the ground-layer vegetation of natural flatwoods and sandhill communities

is scarce (Litt et al. 2001). An extensive literature review done by Litt et al. (2001)

regarding herbicide effects on groundlayer vegetation found only 3 and 7 studies

respectively, on natural flatwoods and sandhill sites. A study done by Neary et al. (1990)

on a flatwoods site in Florida found nearly 75% reduction in species richness following









an intensive weed management regime with repeated applications of sulfometuron,

glyphosate and triclopyr mowing. Brockway et al. (1998) studied the effects of

hexazinone applied at 1.1 and 2.2 kg ha'1 with different application techniques on a

sandhill site. They reported significant decreases in the cover of oak species and

increases in shrub control at the higher rate. Moreover, the higher rate application of

hexazinone induced an 86% increase in biomass of wiregrass in the first growing season

after treatment. With the broadcast application of hexazinone, a short-term reduction of

forbs species was also reported. A long-term study of herbicide effect on understory

plant diversity and richness was conducted by Boyd et al (1995) on sandhill and

piedmont site with one-time broadcast applications of imazapyr, glyphosate, and

hexazinone. The reported results seven years after treatment showed no significant

herbicide effect on species diversity and richness.

Our study was conducted to examine the effects of three common forestry

herbicides (imazapyr, hexazinone and sulfometuron) on the understory species of a

flatwoods site. The specific questions addressed were:

* The effect of herbicides on the overall vegetation composition and species

diversity;

* The effect of herbicides on cover and density of major understory shrub species;

* The effect of herbicides on the percent cover of wiregrass.



Materials and Methods

Study Area Description

This study was conducted on a flatwoods site in Point Washington State Forest,

Walton County, Florida (30020' 16.04" N, 860 4'19.22" W). Average annual highest and









lowest temperatures were 25.50C and 120C, respectively. Annual precipitation was about

1640mm (2002) with most received in the late summer months (Figure 2.1). Soils are of

low pH (pH <5) and sandy texture with low nutrient content (Table 2.1). The study area

soils were mapped as sandy, siliceous, thermic aeric alaquods belonging to the Leon

series, which is characterized by deep, poorly to very poorly drained soils. Soils of the

flatwoods pinelands are formed on sandy quaternary formations derived from marine

deposits (Stout and Marion 1993).

Prior to study establishment the overstory was a stand dominated by slash pine with

the occasional longleaf pine saplings. The average age of the stand was 26 years with a

basal area of 1.85m2 and an average dbh of 19.1cm. The understory was comprised

mainly of species such as gallberry, saw palmetto, runner oak, dangleberry, hairy wicky,

wiregrass and bluestem grasses. Dormant season fires were used on a three-year burn

cycle to mitigate fuel build up.

Experimental Design and Treatments

Prior to site preparation, the overstory was harvested (August 2001). The harvest

debris and the understory were roller chopped once and prescribed burned in October

2001. A randomized complete-block design (RCB) with five replicates was used to

examine the effect of four herbicidal vegetation control methods on the understory

vegetation of the site. The study included five treatment plots within each block. All

treatment plots were 26.6m x 24.4 m, and included a > 3m buffer strip between plots. In

December 2001, one year old containerized longleaf pine seedlings were hand-planted at

3.lm x 1.8m spacing. Seedlings were planted in rows to facilitate the application of

treatments. In March 2002, four herbicide treatments [Sulfometuron (0.26 ai kg ha-1),

Hexazinone (0.56 ai kg ha-1), Sulfometuron (0.26 ai kg ha-1) + Hexazinone (0.56 ai kg









ha-1) mix, Imazapyr (0.21 ae kg ha-l)] were applied in a 1.2 m band over the top of

seedlings using a knapsack sprayer. In each block, one treatment plot was kept herbicide

free as a control plot.

Measurements

A preliminary vegetation survey was conducted (June 2001) to assess the initial

species composition and diversity, prior to harvesting and site preparation. The

experimental blocks and treatment plots were demarcated and within all plots, three

randomly selected 1m2 sampling quadrats were sampled. After study establishment and

herbicide application, two more vegetation surveys were conducted. These surveys were

done three and nine (June and Nov 2002) months after herbicide treatment (MAT) and

sampling was done on the herbicide applied strips along pine seedlings rows. In each

treatment plot six randomly selected 1M2 quadrats were sampled and the same location

was revisited for subsequent survey. In every survey, all plants found within a quadrat

were assigned to vegetation class of shrubs, grass, forbs, vines and ferns. Percent cover

was ocularly estimated for all species using the modified Daubenmire scale (Daubenmire

1959). Stem number and average stem heights were recorded for all shrub species.

Statistical Analysis

Uniformity of percent cover of understory species and stem counts of major shrub

species was tested within the pre-harvest survey data. Only species that displayed

uniformity in percent cover and stem counts were used in comparison with post-treatment

survey data. Percent cover data did not conform to the assumption of normality even

after data transformation. Therefore, comparison of percent cover between treatments

and the control plots for each survey date was performed with Kruskal-Wallis test at a

significance level of ca=0.05. Treatment effects on stem counts were analyzed using









analysis of variance (ANOVA) techniques within the framework of a randomized

complete block design (RCBD) using SAS version 8.0 (SAS 2000). Significant treatment

effects were separated using Duncan's multiple range test. Importance value (IV) was

calculated for shrubs, grasses and forbs. Shrub species IV was calculated as an average

of relative frequency, relative cover and relative density. Average of relative frequency

and relative cover was used as IV for grass and forb species. IV values were log

transformed before analysis. Differences of IV values between treatments were tested

with Duncan's multiple range test at a significance level of a=0.05.

Species diversity was calculated using Shannon-Wiener diversity indices, which

incorporate species richness and evenness of the species. Species diversity was

calculated for all three vegetation surveys using percent cover data and were analyzed

using Multivariate Statistical Package (MVSP) version 3.1 (Kovach 1999).

Canonical Correspondence Analysis (CCA) was used to examine the overall

distribution of species among the study plots. CCA is a direct gradient analysis which

constraint the species distribution along specified environmental variables (Palmer 1993).

The applied treatments were used as nominal environmental variables and time since

treatment was used as a quantitative environmental variable in CCA analysis. CCA

analysis was performed for each survey separately. Monte Carlo permutation test was

used to detect significant (a=0.05) environmental variables.

Results

Species Composition and Diversity

CCA analysis with pre-harvest survey and control plots revealed a change in

species composition with time without the effect of herbicides (Figure 3-1 a). Monte









Carlo permutation tests showed time (p= 0.015) as a significant factor. The

compositional change may be attributed to the seasonal variation within the plant

community and the effects of site preparation. Analysis of 3 MAT vegetation survey data

showed no significant changes in the species composition and distribution between the

treatments (Fig 3-1 b). Monte Carlo permutation tests also revealed that 3 MAT none of

the treatments were significant. However, 9 MAT data showed a distinct separation

between herbicide treatment and the control (Fig 3-2 c). The first axis was highly

correlated with absence of herbicides and the second axis although nonsignificant was

correlated with imazapyr treatment. Monte Carlo permutation revealed absence of

herbicide (P= 0.04) to be the most significant factor for the separation of sample plots

within the ordination. The permutation tests did not reveal any significant herbicide

treatment effect.

A total of 81 species were recorded within the duration of the study. A larger

number of species was found within post-treatment surveys compared to the survey done

prior to harvest. The applied herbicides had no significant effect on species richness 3

and 9 MAT.

Within post-treatment surveys, overall species evenness increased across all

treatments (Table 3-1). However, in both surveys none of the herbicides significantly

changed species evenness among the treatments. Although non-significant, imazapyr

treatment resulted in the highest evenness in both surveys with sulfometuron resulting in

the lowest evenness values.

Species diversity increased 3 MAT across all the treatments compared to pre-

harvest vegetation (Figure 3-2). However, none of the herbicide treatments had a









significant effect on the understory species diversity 3 MAT (Figure 3-2). Further

increase in diversity was observed across all treatments 9 MAT. The largest increase

occurred with imazapyr treatment and was significantly greater than all other treatments,

including control. Sulfometuron markedly reduced species diversity 9 MAT and was

significantly lower than imazapyr, hexazinone treatments and control (Figure 3-2).

Overall, shrub foliar cover was not significantly affected in both 3 and 9 MAT

surveys in all treatments except imazapyr (Table 3-2). Imazapyr resulted in a reduction

of 65% of shrub foliar cover 3 MAT and 58% reduction 9 MAT compared to control

(Table 3-2). Overall grass and forb foliar cover were not significantly affected by any of

the herbicides (Table 3-2).

Understory Shrub Cover and Density

Prior to harvest and site preparation, the understory was dominated by gallberry,

runner oak, and saw palmetto. These species were abundantly found with percent

frequency of 76, 56 and 42, respectively. These species were used to examine herbicide

effects. Among the major understory shrub species, saw palmetto cover was not

significantly affected by any herbicides in both post-treatment surveys (Table 3-2).

Runner oak cover was not affected by the herbicides 3 MAT (Table 3-2). However, it

showed a delayed response to imazapyr, hexazinone and sulfometuron+hexazinone mix

treatments with a foliar cover reduction of 67%, 36% and 45%, respectively, compared to

the control, 9 MAT (Table 3-2). The greatest reduction of cover occurred in gallberry in

the imazapyr treatment with 77% reduction 3 MAT and 70% reduction 9 MAT.

Gallberry cover was not affected by any other treatments (Table 3-2).

Shrub density differences between treatments were analyzed for gallberry, runner-

oak and saw palmetto, which showed uniform density across the study site prior to









treatment application. Saw palmetto density was not affected by any of the applied

treatments. The density of runner oak was not affected by any herbicides 3 MAT but

showed delayed response to imazapyr with a 65% reduction of stem density 9 MAT

(Table 3-3). Gallberry density was unaffected by the applied treatments except imazapyr.

Within the imazapyr treatment gallberry density reduced by 58% and 79% 3 and 9 MAT,

respectively.

Analysis of mean importance values revealed no significant change due to applied

herbicides except imazapyr (Table3-4). Imazapyr treatment significantly reduced the

importance value of gallberry 3 MAT (> 50%) and was further reduced (>60%) 9 MAT.

Wiregrass Cover

Prior to harvest, wiregrass was the dominant grass species with 63 percent

frequency. Interestingly none of the herbicide treatments negatively affected the foliar

cover of wiregrass.

Discussion

Ecological disturbance such as periodic fires, hurricanes and plant mortality due to

pathogens or insects influence the species diversity within an ecosystem by changing

dominance, interspecific competition and initiating succession (Perry 1994). Herbicide

application through its selective activity may also influence dominance of species,

interspecific competition and species composition. Thus, with herbicide application

changes in species diversity, richness and composition can be expected.

CCA ordination on pre-harvest vegetation and the control plots revealed a change

in species composition with time. The difference in compositional change may be a

response to site disturbance and seasonal variation within the plant community. For

instance, an increase in number of forb species such as Aster tortifolius, Liatris tenuifolia









and Sabatia brevifolia were found in the latter survey conducted in late fall. The

ordination analysis for 9 MAT vegetation data showed a significant separation of

treatments due to the absence of herbicide. The separation between herbicide treatments

and the control reveals a herbicide effect on overall species composition.

In our study, post harvest vegetation surveys included a higher number of species

compared to the pre harvest vegetation. The increase in species richness occurred in all

treatment plots and may be attributed to the disturbances brought forth by site preparation

treatments and prescribed burning. Similarly, Brockway et al. (1997) found increases in

species richness following disturbances such as prescribed fire. Species richness did not

change between the herbicide treatments indicating no apparent effect of treatments on

species richness. This is contrary to published research, which shows a decline in species

richness within the first growing season after herbicide application. A study done by

Wilkins et al. (1993) with hexazinone on a flatwoods site reported significant reduction in

species richness with increasing application rates within the first year. A similar study by

Brockway et al. (1998) on a sandhill site with a broadcast application of hexazinone at

1.1 ai kg/ha reduced the species richness by 28%. It should be noted that both the above

studies used site preparation rates of hexazinone and used broadcast application methods.

The relatively lower rate application of herbicides and the banded application method

used in our study did not significantly affect species richness.

The increase in species diversity across the study site is a characteristic response to

site disturbances. Usually a decrease in species diversity within the first growing season

after herbicide treatments is common. Brockway et al. (1998) reported an overall

reduction in species diversity on a sandhill site treated with high rates of hexazinone.









With the exception of imazapyr and sulfometuron, the herbicide treatments used in our

study did not significantly impact understory species diversity at the applied rates.

Imazapyr was the only treatment to significantly increase species diversity 9 MAT. In

addition, an increase in species evenness was also observed. Species evenness measures

the similarity of abundance between species found within a given area. Higher evenness

values signify a lack of dominance by a fewer species and reveals similar proportions of

abundance. The significant reduction of shrub foliar cover and density by imazapyr may

have reduced the dominance of shrub species, thereby increasing resource availability to

other species. The absence of shrub dominance may have facilitated similar distribution

of abundance among other species increasing species evenness and in turn improving

species diversity. In contrast to imazapyr, sulfometuron resulted in a significant decrease

in species diversity. The decrease in diversity is attributable to the lower species richness

and evenness found in the sulfometuron treatment.

Although imazapyr showed significant control over the dominant shrub species, it

did not cause a reduction in species richness within the treated area. This reveals the

selectivity of imazapyr on its target species and its inability to severely damage or

eliminate them from the understory at the applied rate. The herbicide application method

used in our study may also have contributed to the lack of severity of herbicide effect on

species diversity and richness. In contrast to broadcast application, the band application

method provides a herbicide free buffer between treatment bands. The buffer strips

provide a suitable seed source for re-colonization within the treated plots. Studies done

with broadcast and spot application treatments have reported greater decline in species









richness and diversity with the broadcast application of herbicides compared to the spot

application method (Brockway et al. 1998).

One of the important concerns with longleaf pine restoration techniques is their

effect on wiregrass regeneration and growth. Clewell (1989) reported on the detrimental

effects of soil disturbance on wiregrass. In his research 85% reduction in wiregrass

density was found with soil disturbance in flatwoods. In addition, the negative impact of

site preparation methods such as chopping was also reported. In contrast to the above

findings, the site preparation chopping performed in this study did not have any negative

effect on wiregrass foliar cover. However, the damaging effects of soil disturbances is

lessened along an increasing moisture gradient (Clewell 1989). Thus, the recovery of

wiregrass following site disturbances may be site specific. A study done on a sandhill

site reported an increase in wiregrass cover in a linear relationship with increasing

application rates of hexazinone within the first growing season (Wilkins et al. 1993).

They also reported a significant reduction in the dominant oak species that may have led

to the increase in wiregrass cover. A similar study done by Brockway et al. (1998)

reported an 86% increase in wiregrass cover following a 93% reduction in dominant

shrub cover. The above research indicates a positive response of wiregrass to reduction

of shrub dominance within the first growing season. In our study, the herbicide

treatments did not significantly effect the cover of wiregrass. However, imazapyr

resulted in a significant reduction in gallberry cover and runner oak showed delayed

significant reduction to hexazinone, imazapyr and sulfometuron+hexazinone mix

treatments. Thus, the reduced shrub competition may yield greater wiregrass cover in the









following years. Wiregrass is also reported to experience a lag time and requires at least

two growing seasons to respond to reduced competition (Brockway et al. 1998).

Overall, imazapyr yielded the most desirable results by reducing the shrub

dominance while improving the understory biodiversity. Imazapyr was also successful in

yielding improved growth of planted pine seedlings. Hexazinone yielded a delayed

control of runner oak with no significant impact on the understory biodiversity. The

benefits of hexazinone with its delayed control of runner oak may be realized in the

coming years. However after one growing season hexazinone did not yield greater

benefits compared to control. Sulfometuron yielded poor results by reducing the

biodiversity with no significant control over shrub dominance. The application of

sulfometuron was also found detrimental to planted longleaf pine seedlings on flatwoods

sites as described in chapter two. The sulfometuron+hexazinone mix treatments did not

result in any improved benefits.

Conclusions

Within the first growing season, imazapyr treatment significantly increased the

species diversity while increasing evenness. In contrast, sulfometuron decreased the

species diversity and reduced species evenness. None of the other herbicides

significantly impacted the understory plant diversity. Imazapyr treatment resulted in an

overall reduction of shrub foliar cover with a significant reduction of gallberry and runner

oak. However, shrub species such as saw palmetto and huckleberry species displayed

resistance to imazapyr at the applied rate. Imazapyr caused a higher mortality rate of

pine seedlings compared to control, but resulted in a significant increase in growth.

Hexazinone was not effective in the control of major understory shrub species with the

exception of runner oak. Runner oak displayed a delayed response to hexazinone by






47


significant reduction of foliar cover 9 MAT. None of the other herbicide treatments was

effective in reducing shrub dominance. Therefore, on flatwoods sites with an understory

dominated by gallberry and runner oak, application of imazapyr seems to yield the best

control of shrub dominance one-year after treatment. The use of sulfometuron at the

above-applied rates is not recommended on flatwoods sites as it decreased understory

species diversity while hindering longleaf pine seedling growth.






48


Figure 3-1.Canonical correspondence analysis (CCA) ordination for understory
vegetation. (a) CCA ordination with pre harvest and post treatment control
plots. (b) CCA ordination of vegetation 3 months after treatment (MAT). (c)
CCA ordination of vegetation 9 MAT. The tested nominal variables are
plotted as centroids.

















CCA case scores


0 pre-treatment June 01

Control June 02

* control Nov 02


June 2001


November 2002


CCA case scores


Ssulfo+hexa
hexa
V sulfo
cont
i Imaz


44 -35 -26 -18 09



A -2
2


Axis 1



CCA case scores


xa


2 8


--21
imaz #
1 4
14





-35 -28 -2sulfo+hexa -07
*A 007'

hexa
*-1 4


r I cornt )

SCsU 4 2o1 28 t 35





control


-28

-3 5

Axis 1


V

V


sulfo+hex
A hexa
V sulfo

I cont

SImaz


Vector scalina 333






50





2.5
-K- control
-- sulfometuron c
--A- hexazinone
-- sulfo+hexa
2 --- imazapyr a
> 2 ab

0 b




1.5
pre-harvest 3 MAT 9 MAT
Date

Figure 3-2. Mean Shannon diversity index for each treatment by date. Shannon's index
values for each treatment followed by a different letter is significantly
different from other treatments within each date. Significance established
through Duncan's multiple range test at ca=0.1. (sulfo+hexa:
sulfometuron+hexazinone mix treatment)









Table 3-1. Mean species evenness and richness of three vegetation surveys
Treatment Evenness Richness

Pre harvest survey 0.70 10.40

3 MAT2 survey
Control 0.76 14.80
Sulfometuron 0.74 11.60
Hexazinone 0.78 13.00
Sulfo+Hexa1 0.77 12.40
Imazapyr 0.81 13.80

9 MAT2 survey
Control 0.83 13.00
Sulfometuron 0.77 11.20
Hexazinone 0.83 12.60
Sulfo+Hexa1 0.81 12.80
Imazapyr 0.87 14.80
Sulfometuron+Hexazinone mix treatment
2Months after treatment











Table 3-2. Percentage change in foliar cover (%
vegetation classes by treatment


cover) of major understory species and


Species/ Sulfometuron Hexazinone Sulfo+Hexa Imazapyr
Vegetation
class


Wiregrass
June 2002
November 2002

Bluestem spp
June 2002
November 2002

Gallberry
June 2002
November 2002

Runner oak
June 2002
November 2002

Saw palmetto
June 2002
November 2002

All shrub
June 2002
November 2002

All grasses
June 2002
November 2002

Allforbs
June 2002
November 2002


33.86 +
21.53 +


9.27 +
56.93 a


12.80 +
7.16+


13.17+
10.26 +


28.29 -
6.18+


12.22 -
1.07-


29.49 -
28.34-


34.34-
15.07-


44.44 +
18.69 +


12.14+
19.83 -


19.27 +
20.65 +


34.22 -
36.66 a


22.45 -
11.12-


32.50-
16.67-


21.55-
4.15 -


8.26-
25.59-


11.11 +
11.82 +


3.51 +
8.77 +


45.79 +
34.68 +


42.93 -
45.34 a


29.19-
25.97 +


29.29 +
7.82 +


48.24 -
19.29-


77.04 -a
69.61 -a


67.91 -
68.33 -a


0.71 +
0.73 +


27.28 54.20 a
11.98- 49.87-a


30.80-
20.57-


32.60 -
22.24 -


34.55-
13.53 -


20.00 +
47.36 +


Isulfometuron+hexazinone mix treatment
+ and indicate increase and decrease of foliar cover in relation to control
a significantly different from control within rows. Significance established with Kruskal-
Wallis test at a = 0.1









Table 3-3. Mean stem numbers for three major understory shrub species by treatment
Shrub stem numbers (stems/m2)
Species Pre harvest

Gallberry 22.422.1
Runner oak 32.163.8
Saw palmetto 2.220.2
Control Sulfometuron Hexazinone Sulfo+Hexa Imazapyr


3 MAT1
Gallberry 18.372.6 20.602.7 16.931.5 16.601.8 7.600.9 a
Runner oak 31.405.3 31.295.9 23.143.9 20.322.5 17.623.6
Saw palmetto 2.200.5 2.450.4 2.000.4 2.550.64 2.900.6


9 MAT1
Gallberry 34.504.9 39.805.3 38.333.7 43.005.9 7.031.6 a
Runner oak 52.077.8 59.2012.1 37.576.4 36.574.7 18.043.3 a
Saw palmetto 2.750.5 3.440.8 2.000.6 3.281.3 3.330.8
1 Months after treatment


a significant differences between treatment within rows.
Duncan's multiple range test at c = 0.1


Significance established by












Table 3-4. Importance values (IV) for major understory species in three vegetation surveys by treatment.
Species1
Treatment Blue Arbe Gadu* Ilgl* Kahi* Limi* Pani Qupu* Sere*

Pre harvest 0.79 0.60 0.99 1.05 1.02 1.11 0.70 1.05 0.954


3 MAT
Control 0.70 0.54 1.65 1.22 1.65 1.68 0.88 1.32 1.32
Sulfometuron 0.85 0.48 1.51 1.26 1.68 0.23 0.71 1.12 1.23
Hexazinone 0.92 0.62 1.50 1.21 0.40 1.54 0.17 1.13 0.78
Sulfo+Hexa2 0.73 0.52 0.92 1.27 0.39 0.89 0.65 1.06 0.95
Imazapyr N/A3 0.45 N/A3 0.45a N/A3 1.06 0.29 0.63 1.20


9 MAT
Control 0.60 0.47 1.59 1.12 2.38 1.22 1.04 1.41 1.11
Sulfometuron 0.53 0.58 1.61 1.24 N/A3 N/A3 0.46 1.24 1.45
Hexazinone 0.60 0.65 1.03 1.35 N/A3 1.39 0.28 1.09 0.73
Sulfo+Hexa2 0.77 0.63 0.57 1.42 0.37 1.13 0.10 1.07 1.05
Imazapyr 0.71 0.56 1.26 0.36a 1.07 1.06 0.34 0.51 1.18
1Blue:Bluestem spp, Arbe:Wiregrass, Gadu:Dwarf huckleberry, Ilgl:Gallberry, Kahi:Hairy wicky, Limi:Gopher apple, Pani:Panicum
spp, Qupu:Runner oak, Sere:Serenoa repens
2 Sulfometuron+Hexazinone mix treatment
3N/A: Not present in treatment plots at the time of survey
* IV values calculated as an average of relative frequency, cover and density
a Significantly different along rows. Significance established with Duncan's multiple range test at o = 0.1














CHAPTER 4
SUMMARY AND CONCLUSIONS

Successful restoration of longleaf ecosystems and their associated understory may

depend on the control of competing vegetation. On flatwoods sites, the need for

vegetation control is magnified due to the dense shrub understory that smothers young

longleaf seedlings and the herbaceous vegetation. Although prescribed fire is an

effective restoration tool in many longleaf sites, in flatwoods its shrub control is short

lived due to the vigorous sprouting of the shrub vegetation (Abrahamson 1984).

Intensive mechanical site preparation, while yielding enhanced growth of seedlings

(Burger et al. 1988), may cause extensive damage to the understory vegetation (Clewell

1989).

In our study, we examined the effects of over the top application of herbicides

following site preparation with chopping and burning. The use of herbicides such as

imazapyr exhibited an evident trade off between non-application and application of

herbicides. Herbicide treatments such as imazapyr, which resulted in significantly higher

growth within the first growing season, also resulted in significantly lower survival.

Where as in control, higher survival was observed but seedling growth was significantly

lower compared to imazapyr treatment. On the other hand hexazinone had no effect on

survival and did not improve longleaf pine seedling growth. The use of sulfometuron on

flatwoods is not recommended as it significantly lowered survival and growth of planted

longleaf pine seedlings. The sulfometuron+hexazinone mix treatment did not yield any

additional benefit. Imazapyr was the only treatment that yielded significant control of the









overall shrub foliar cover with no effect on the herbaceous component of the understory.

The significant increase in seedling growth by controlling the shrub component alone

reiterates the importance of shrub vegetation control on flatwoods sites. Hexazinone

yielded delayed control of runner oak which was one of the dominant shrub species and

may yield desirable results in the coming years.

The understory species composition exhibited a change without the influence of

herbicide treatments. These changes could be attributed to the seasonal changes in the

vegetation and the site preparation activities. However, the herbicide treatments as a

whole showed a distinct influence on the overall species composition. The understory

species diversity was increased across all treatments compared to pre harvest vegetation.

The herbicide treatments did not have a significant effect on the species diversity with the

exception of imazapyr and sulfometuron. Imazapyr increased species diversity 9 MAT

with nonsignificant increases in species evenness and richness. In contrast, sulfometuron

decreased species diversity while resulting in lowered evenness and richness.

Contrary to published research, the site preparation conducted in our study did not

significantly affect the foliar cover of wiregrass. The herbicide treatments also had no

significant affect on wiregrass.

In conclusion, imazapyr yielded the best-desired effects with increased longleaf

pine seedling growth, significant shrub control and improved understory plant diversity.

In contrast, sulfometuron resulted in lower survival, growth and species diversity.

Therefore, application of sulfometuron for shrub control on flatwoods sites is not

recommended.














APPENDIX
SPECIES LIST


Scientific name
Shrubs
Asimina incana
Cyrilla racemiflora
Gaylussacia dumosa
Gaylussacia frondosa
Ilex cariacea
Ilex glabra
Ilex vomitoria
Kalmia hirsuta
Licania michauxii
Lyonia lucida
Magnolia virginiana
Myrica cerifera
Photinia pyrifolia
Quercus pumila
Seronoa repens
Stillingia sylvatica
Vaccinium spp

Grasses
Bluestem spp
Arisitida beyrichiana
Calamovilfa curtissii
Ctenium aromaticum
Cyperus
Eragrostis spectabilis
Panicum
Panicum erectifolium
Panicum laxiflorum
Scleria
Xyris caroliniana

Forbs
Asclepias viridula
Aster adnatus
Aster eryngiifolius
Aster reticulatus
Aster tortifolius


Code


Common name


Asin
Cyra
Gadu
Gafr
Ilca
Ilgl
Ilvo
Kahi
Limi
Lylu
Mavi
Myce
Phpy
Qupu
Sere
Stsy
Vacc


Blue
Arbe
Cacu
Ctar
Cype
Ersp
Pani
Paer
Pala
Scle
Xyca


Asvi
Asad
Aser
Asre
Asto


Wooly paw paw
Ti ti
Drawf huckleberry
Dangleberry
Large gallberry
Gallberry
Yaupon
Hairy wicky
Gopher apple
Fetterbush
sweet bay
Wax myrtle
Red choke berry
Runner oak
Saw palmetto
Queens delight
Blueberry spp


Bluestem grasses
Wiregrass
Curtis sandgrass
Toothache grass
Sedge spp
Purple lovegrass
Panicun spp
Erect leaf witchgrass
Velvet Witchgrass
Nutrush spp
Yellow eyed grass


Southern milkweed
Scaleleaf aster
Thistleleaf aster
White top aster
Dixie aster









Species list continued
Carphephorous
pseudoliatris
Carphephorus
odoratissimus
Chrysopsis
Conyza canadensis
Coreopsis linifolia
Desmodium rotundifolium
Drosera capillaris
Elephantopus
tomentosus
Eupatorium capillifolium
Eupatorium compositifolium
Eupatorium mohrii
Eupatorium pilosum
Euthamia graminifolia
Gelsemium sempervirens
Gratiola hispida
Hypericum hypericoides
Hypoxis sessilis
Hypoxis spp
Lachnanthes caroliana
Lechea
Lechea pulchella
Liatris gracilis
Liatris tenuifolia
Mimosa quadrivalvis
Oenothera fruticosa
Opuntia humifusa
Pityopsis graminifolia
Pterocaulon
pycnostachyum
Rhexia alifanus
Rhexia petiolata
Sabatia brevifolia
Seymeria cassioides
Smilax laurifolia
Smilax pumila
Solidago odora
Stylisma patens
Tragia urens
Verbena brasiliensis
Viola septemloba
Vitis rotundifolia


Caps


Bristleleaf chaffhead


Caod Deer tounge


Chry
Coca
Coli
Dero
Drca
Elto

Euca
Euco
Eumo
Eupi
Eugr
Gese
Grhi
Hyhy
Hyse
Hypo
Laca
Lech
Lepu
Ligr
Lite
Miqu
Oefr
Ophu
Pigr
Ptpy

Rhal
Rhpe
Sabr
Seca
Smla
Smpu
Sood
Stpa
Trur
Vebr
Vise
Viro


Silkgrass spp
Canadian horseweed
Texas tickseed
Tricklyfoil
Pink sundew
Devils grandmother

Dog fennel
Yankee weed
Mohr's thoroughwort
Rough Boneset
Flat top goldenrod
Yellow jessamine
Rough Hedgehyssop
St. Andrews cross
Glossyseed yellow stargrass
Stargrass spp
Carolina redroot
Pineweed spp
Leggett' s pineweed
Slender gayfeather
Shortleaf gayfeather
Sensitive brier
Evening primrose
Prickly pear
Silkgrass
Blackroot

Meadow beauty
Fringed meadow beauty
Shortleaf Rosegentian
Yaupon Blacksenna
Laurel green brier
Green brier
goldenrod
Coastal plain dawn flower
Wavyleaf noseburn
Brazilian vervain
Blue violet
Muscadine
















LIST OF REFERENCES


Anon. 2000. Guidance document on residue analytical methods, European Commission,
Directorate General Health and Consumer Protection. SANCO/825/00 rev 6, 20
June 2000. 16p.

Abrahamson, W. G. 1984. Post-fire recovery of Florida Lake wales ridge vegetation. Am.
J. Bot. 71 (1): 9-21.

Boyd, R. S., J. D. Freeman, J. H. Miller and M. B. Edwards. 1995. Forest herbicide
influences on floristic diversity seven years after broadcast pine release treatments
in central Georgia, USA. New Forests. 10: 17-37.

Boyer, W. D. 1990. Pinus Palustris, Longleaf pine. In: Burns, R. M. and B. H. Honkala.
Silvics of North America, Vol. 1. Conifers. USDA Forest Serv. Washington, DC,
pp 405-412.

Brewster, B. D. and A. P. Appleby. 1983. Response of wheat (Triticum aestivum) and
rotation crops to chlorosulfuron. Weed Sci. 31: 801-805.

Brockway, D. G., and C. E. Lewis. 1997. Long-term effect s of dormant-season
prescribed fire on plant community diversity, structure and productivity in a
longleaf pine wiregrass ecosystem. Forest. Ecol. Manage. 96: 167-183.

Brockway, D. G., and K. W. Outcalt. 2000. Restoring longleaf pine wiregrass
ecosystems: Hexazinone application enhances effects of prescribed fire. Forest.
Ecol. Manage. 137: 121-138.

Brockway, D. G., K. W. Outcalt, and R. N. Wilkins. 1998. Restoring longleaf pine
wiregrass ecosystems: Plant cover, diversity and biomass following low rate
hexazinone application on Florida sandhills. Forest. Ecol. Manage. 103: 159-175.

Burger, J. A., and W. L. Pritchett. 1988. Site preparation effects on soil moisture and
available nutrients in a pine plantation in the Florida flatwoods. Forest. Sci. 34: 77-
87.

Busby, R. L., C. E. Thomas, R. E. Lohrey, K. H. N. Le, and M. B. Edwards. 1995.
Rapidly restoring large diameter longleaf pine ecosystems at a low cost: A
preliminary appraisal. Proc. of the Eighth Bienn. South. Silvic. Res. Conf, Auburn,
Alabama, 1994. Gen. Tech. Rep. South. Res. Stn. US Dept. Agric. Forest. Serv.
SRS-1, 192-196.









Carter, R. E., M. D. Mackenzie, D. H. Gjerstad and T. A. Waldrop. 1998. Ecological land
classification of longleaf pine ecosystems in the Southern Loam Hills of South
Alabama. Proc. 9th Bienn. South. Silvic. Res. Conf. Gen. Tech. Rep. US Dept.
Agric. Forest. Serv. SRS 20, 81-85.

Clewell, A. F. 1989. Natural history of wiregrass (Aristida stricta). Nat. Areas J. 9, 223-
233.

Collins, B., P. S. White and D. W. Imm. 2001. Introduction to ecology and management
of rare plants of the southeast. Nat. Areas J. 21(1): 4-11.

Croker, T. C., Jr., 1975. Seedbed preparation aids natural regeneration of longleaf pine.
USDA. For. Ser. Res. Pap, South. Forest Exp. Stn. No. SO-112.

Croker, T. C., Jr., 1979. Longleafpine. The longleafpine story. J. Forest History. 23(1):
32-43.

Daubenmire, R. F. 1959. Canopy coverage method of vegetation analysis. Northwest Sci.
33: 43-64.

Davis, J. H. 1967. General map of natural vegetation of Florida. Inst. Food and Agric.
Sci. Circ. S-178, University of Florida, Gainesville.

Delcourt, P. A., H. R. Delcourt, D. F. Morse and P. A. Morse. History, evolution and
organization of vegetation and human culture. In: Martin, W. H., S. G. Boyce and
A. C. Echternacht. Ed. Biodiversity of the Southeastern United States, Lowland
Terrestrial Communities. John Wiley and Sons, Inc., New York pp 47-80

Dickens, R., and G. Wehtje. 1986. Mobility and soil solution characteristics of imazapyr
and sulfometuron methyl. Proc. South. Weed Sci. Soc. Conf 39: 368.

Frost, C. C. 1993. Four centuries of changing landscape patterns in the longleaf pine
ecosystem. In: Hermann, S. M. (ed), Proc. Tall Timbers Fire Ecol. Conf. Vol. 18.
Tall Timbers Res. Stn., Tallahassee, FL, pp 17-43.

Gjerstad, D. H., L. R. Nelson. 1983. New herbaceous weed control in pine outplantings.
Proc. South. Weed Sci. Soc. Conf. 36: 245.

Grelen, H. E. 1978. May burns stimulate growth of longleaf pine seedlings. US Forest
Serv. Res. Note, South. Forest Exp. Stn. No. SO-234.

Grelen, H. E. 1983. May burning favors survival and early height growth of longleaf pine
seedlings. South. J. Appl. For. 7 (1):, 16-20.

Haywood, J. D. 2000. Mulch and hexazinone herbicide shorten the time longleaf pine
seedlings are in the grass stage and increases height growth. New Forests. 19: 279-
290.









Hernandez-Sevillano, E., M. Villarroya, J. L. Alonso-Prados and J. M. Garcia-Baudin.
2001. Bioassay to detect MON-37500 and triasulfuron residues in soils. Weed
Technology. 15: 447-452.

Hollaway, K. L., R. S. Kookana, D. J. McQuinn, M. R. Moerkerk, D. M. Noy and M. A.
Smal. 1999. Comparison of sulfonylurea herbicide residue detection in soil by
bioassay, enzyme-linked immunosorbent assay and hplc. Weed Res. Oxford. 39(5):
383-397.

Huck, R. B. 1987. Plant communities along an edaphic continuum in a central Florida
watershed. Florida scientist. 50: 111-128.

Jose, S., S. Merritt and C. L. Ramsey. 2003. Growth, nutrition, photosynthesis and
transpiration responses of longleaf pine seedlings (Pinuspalustris Mill.) to light,
water and nitrogen. Forest Ecology and Management 180:335-344.

Kelly, J. F., and W. A. Bechtold. 1990. The longleaf pine resource. In: Farrar, R. M. (Ed).
Proc. Symp. Manage. Longleaf pine. USDA Forest Serv. South. Forest Exp. Stn.
Gen. Tech. Rep. SO-75, New Orleans LA, pp. 11-22.

Kline, W. N., J. L. Troth and B. D. Shiver. 1994. Maximizing pine yields in the southern
flatwoods using Garlon herbicide. Down to Earth. 49: 14-22.

Kovach, W. L., 1999. MVSP-A multivariate statistical package for windows, ver. 3.1.
Kovach Computing Services, Pentraeth, Wales, UK.

Landers, J. L., D. H. Van Lear, and W. D. Boyer. 1995. The LongleafPine Forests of the
Southeast: Requiem or Renaissance? J. For. 93: 39-44.

Lipscomb, D. J. 1989. Impacts of feral hogs on longleaf pine regeneration. South. J.
Appl. For. 13(4): 177-181.

Litt, A. R., B. J. Herring and L. Provencher. 2001. Herbicide effects on ground layer
vegetation in southern pine lands (USA): A review. Nat. Areas J. 21 (2): pp 177-
188

Metcalfe, C. S., and R. L. Cantrell. 1986. Secondry screening for herbaceous weed
control in longleaf pine plantations-Oust registration data. Auburn University
Silvic. Herbicide Cooperative. Res. note 86-1.

Michael, J. L. 2000. Pesticides used in forestry and their impacts on water quality. Proc.
South. Weed Sci. Soc. 53: 81-87.

Michael, J. L., E. C. Webber, Jr., D. R. Bayne, J. B. Fischer, H. L. Gibbs and W. C.
Seesock. 1999. Canadian J. Forest Res. 29: 1170-1181

Michael, J. L., and D. G. Neary. 1993. Herbicide dissipation studies in southern forest
ecosystems. Environ. Toxicology and Chemistry. 12: 405-410.









Neary, D. G., and J. L. Michael. 1989. Effect of sulfometuron methyl on groundwater and
stream quality in coastal plain forest watersheds. Water Resour. Bull. 25 (3): 617-
623.

Neary, D. G., J. L. Michael, J. A. Griffith. 1996. Herbicides- Protecting long term
sustainability and water quality in forest ecosystems. Second Int. Conf. Forest
vegetation manage. 26: 1-2, 241-264.

Norris, L. A. 1981. The movement, persistence, and fate of the Phenoxy herbicides and
TCDD in the forest. Residue Rev. 80: 65-135.

Outcalt, K. W., 2000. The longleaf pine ecosystem of the south. Native Plants J. 1 (1):
43-51.

Palmer, M. W. 1993. Putting things in even better order: The advantages of canonical
correspondence analysis. Ecology. 74: 2215-2230.

Pederson, N., J. S. Kush, R. S. Meldahl, W. D. Boyer, J. D. Haywood. 1999. Longleaf
pine cone crops and climate: a possible link. Gen. Tech. Rep. South. Res. Stn,
USDA Forest Serv. No.SRS-30, 255-258.

Peet, R. K., 1993. A taxanomic study of Aristida strict and A. beyrichiana. Rhodora 95
(881): 25-37.

Peet, R. K and D. J. Allard 1993. Longleaf pine dominated vegettaion of the southern
Atlantic and eastern gulf coastal region, USA. In: S. M. Hermann (Ed), Proc. Tall
Timbers Fire Ecol. Conf. Tall Timbers Res. Stn., Tallahassee, FL, 18: 45-81.

Perry, D. A. 1994. How biodiversity is created and maintained. In: Forest ecosystems.
Baltimore. Johns Hopkins University Press, Baltimore. ppl96-218.

Platt, W. J., G. W. Evans, and M. M. Davis. 1988. Effects of fire season on flowering of
forbs and shrubs in longleaf forests. Oecologia. 76: 353-363.

Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, K. E. M. Galley, G. W.
Tanner, J. L. Hardesty and L. A. Brennan. 2001. Effects of hardwood reduction
techniques on longleaf pine sandhill vegetation in northwest Florida. Rest. Ecol. 9:
13-27.

Ramsey, C. L., S. Jose and S. Ranasinghe. 2004. Manuscript in preparation.

Richardson, D. R. 1985. Allelopathy and fire in the Florida scrub. Am. J. Bot. 72(6): 864.

Shiver, B. D., J. W. Rheney, M. J. Oppenheimer. 1990. Site preparation method and early
cultural treatments affect growth of flatwoods slash pine plantaions. South. J.
Appl. For. 14 (4): 183-188.









Shiver, B. D., S. A. Knowe, M. B. Edwards and W. N. Kline. 1991. Comparison of
herbicide treatments for controlling common Coastal Plain flatwood species. South.
J. Appl. For. 15: 187-193.

Stork, P., and M. C. Hannah. 1996. A bioassay method for formulation testing and
residue studies if sulfonylurea and sulfonanylide herbicides. Weed Res. 36: 271-
281.

Stout, I. J., and W. R. Marion. 1993. Pine flatwoods and xeric pine forests of the southern
(lower) coastal plain. In: Martin, W. H., S. G. Boyce and A. C. Echternacht. Ed.
Biodiversity of the Southeastern United States, Lowland Terrestrial Communities.
John Wiley and Sons, Inc. New York. pp 373-446

Wilkins, R. N., W. R. Marion, D. G. Neary, and G. W. Tanner. 1993. Vascular plant
community dynamics following hexazinone site preparation in the lower coastal
plain. Canadian J. Forest Res. 23: 2216-2229.















BIOGRAPHICAL SKETCH

Sanjaya Ranasinghe was born on December 26th 1975, in Colombo, Sri Lanka. In

1996 he came to the United States to pursue an undergraduate degree at Luther College

Decorah, Iowa. Upon graduation in 2000 with a B.A. in Biology he was employed at

Tidewater Inc., an environmental consultant company based in Columbia, Maryland. In

2001 he was accepted to pursue a master's degree with the School of Forest Resources

and Conservation at University of Florida.