Overcoming the Challenges of Tamarix Management with Diorhabda Carinulata Through the Identification and Application of Semioche

Home , Aphthona nigriscutis, Chrysomela lapponica, Diorhabda elongata, Gastrophysa viridula, Gnathotrichus, Myricaria

OVERCOMING THE CHALLENGES OF TAMARIX MANAGEMENT WITH

DIORHABDA CARINULATA THROUGH THE IDENTIFICATION AND

APPLICATION OF SEMIOCHEMICALS

Alexander Michael Gaffke

A dissertation submitted in partial fulfillment of the requirements for the degree

Doctor of Philosophy

Ecology and Environmental Sciences

MONTANA STATE UNIVERSITY Bozeman, Montana

May 2018

©COPYRIGHT

Alexander Michael Gaffke

2018

ACKNOWLEDGEMENTS

This project would not have been possible without the unconditional support of my family, Mike, Shelly, and Tony Gaffke.

I must thank Dr. Roxie Sporleder for opening my world to the joy of reading.

Thanks must also be shared with Dr. Allard Cossé, Dr. Robert Bartelt, Dr. Bruce

Zilkowshi, Dr. Richard Petroski, Dr. C. Jack Deloach, Dr. Tom Dudley, and Dr. Dan

Bean whose previous work with Tamarix and Diorhabda carinulata set the foundations for this research.

I must express my sincerest gratitude to my Advisor Dr. David Weaver, and my committee: Dr. Sharlene Sing, Dr. Bob Peterson and Dr. Dan Bean for their guidance throughout this project.

To Megan Hofland and Norma Irish, thanks for keeping me sane.

iii

TABLE OF CONTENTS

1. INTRODUCTION ...... 1

Tamarix ...... 1 Taxonomy ...... 1 Introduction and Spread ...... 2 Biology ...... 3 Ecology ...... 6 Impacts ...... 8 Control ...... 12 Classical Biological Control ...... 13 Diorhabda carinulata ...... 14 Evaluation and Selection...... 14 Life Cycle...... 16 Dispersal ...... 19 Release and Establishment ...... 20 Impacts on Tamarix ...... 21 Challenges and Conflicts ...... 23 Chemical Ecology and Insects ...... 29 Semiochemicals in the D. carinulata/Tamarix System ...... 33 Use of Semiochemicals to Resolve Challenges and Conflicts with D. carinulata ...... 35 References ...... 41

2. SEMIOCHEMICALS TO ENHANCE HERBIVORY BY DIORHABDA CARINULATA AGGREGATIONS IN SALTCEDAR (TAMARIX SPP.) INFESTATIONS ...... 62

Contribution of Authors and Co-Authors ...... 62 Manuscript Information Page ...... 64 Abstract ...... 65 Introduction ...... 65 Materials and Methods ...... 66 Field Experiment Location and Insect Source ...... 66 Lure and Chemicals ...... 66 Release Rate Analysis ...... 67 Field Study ...... 67 Statistical Analysis ...... 68 Results ...... 68 Release Rate Analysis ...... 68 Field Study ...... 68 High-density Site Reproductive Adult Response ...... 68 Low-density Site Reproductive Adult Response ...... 68 iv

TABLE OF CONTENTS CONTINUED

High-density Site Non-reproductive Adult Response ...... 68 Low-density Site Non-reproductive Adult Response ...... 68 High-density Larval Response ...... 69 Low-density Larval Response ...... 69 High-density Damage Rating ...... 69 Low-density Damage Rating ...... 69 Dieback Rating...... 70 Canopy Volume Measurements ...... 70 Discussion ...... 70 Conclusion ...... 73 Acknowledgements ...... 73 Supplemental Material ...... 73 References ...... 73

3. FIELD DEMONSTRATION OF A SEMIOCHEMICAL TREATMENT THAT ENHANCES DIORHABDA CARINULATA BIOLOGICAL CONTROL OF TAMARIX SPP...... 75

Contribution of Authors and Co-Authors ...... 75 Manuscript Information Page ...... 77 Abstract ...... 79 Introduction ...... 80 Materials and Methods ...... 84 Field Experiment Location and Insect Source ...... 84 Lure and Treatment ...... 84 Release Rate Analysis ...... 85 Field Study ...... 86 Statistical Analysis ...... 88 Results ...... 89 Release Rate Analysis ...... 89 Field Study ...... 89 Reproductive Adults ...... 89 Non-reproductive Adults ...... 90 Larval Densities ...... 90 Damage, Dieback, and Canopy Volume ...... 91 Discussion ...... 92 Acknowledgements ...... 99 References ...... 107

TABLE OF CONTENTS CONTINUED

4. ALLEE EFFECTS AND AGGREGATION PHEROMONES: NEW RELEASES OF DIORHABDA CARINULATA REMAIN LONGER IN THE PRESENCE OF PHEROMONE FORMULATIONS ...... 114

Contribution of Authors and Co-Authors ...... 114 Manuscript Information Page ...... 116 Abstract ...... 118 Introduction ...... 119 Materials and Methods ...... 124 Lures and Chemicals ...... 124 Insect Source ...... 125 Site Selection and Setup...... 126 Field Release Protocol ...... 127 Insect Sampling and Population Survey ...... 128 Statistical Analysis ...... 128 Results ...... 129 Immediate Dispersal ...... 129 Persistence of Individuals at a Release ...... 130 Discussion ...... 131 Acknowledgements ...... 135 Supplemental Material ...... 140 References ...... 143

5. EVALUATION AND FORMULATION OF REPELLENT SEMIOCHEMICALS FOR DISRUPTION OF DIORHABDA CARINULATA AGGREGATIONS ...... 149

Contribution of Authors and Co-Authors ...... 149 Manuscript information Page ...... 151 Abstract ...... 153 Introduction ...... 154 Materials and Methods ...... 159 Insects ...... 159 Hemolymph Collection ...... 159 Volatile Collection and Analysis ...... 160 Preparation of 4-oxo-(E)-2-hexenal for Bioassays ...... 161 Electrophysiology ...... 161 Two-choice Olfactometer Bioassay ...... 163 Statistical Analyses ...... 164 Results ...... 165 Volatile Collections ...... 165 vi

TABLE OF CONTENTS CONTINUED

EAG Response ...... 166 Repellency...... 166 Discussion ...... 168 Acknowledgements ...... 172 Supplemental Material ...... 178 References ...... 182

6. CONCLUSIONS...... 188

REFERENCES CITED ...... 193

APPENDIX A: Chapter Two Supplemental Material ...... 222

vii

LIST OF TABLES

Table Page

2.1 Mean ± SE (nghr-1) of semiochemicals emitted from 1g dollops of field aged dollops over a 31 day period ...... 68

4.1 Poisson regression analysis for mean densities per 20 sweeps, and mean counts per 3 mins ...... 139

S.4.1 Number of adult D. carinulata released and detected within caged release releases in 2016. Caged releases lost to O. stactogalus damage are designated with (---) ...... 142

viii

LIST OF FIGURES

Figure Page

2.1 Density expressed as mean ±SE capture of reproductive adult D. carinulata per three sweeps in (a) 2013 and (b) 2014 at the high density (HDS) or low density (LDS) site on plants receiving 1 g dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), or plant volatile (PL). Different letters above error bars denote significant differences between treatments...... 69

2.2 Density expressed as mean ±SE capture of non-reproductive adult D. carinulata per three sweeps in (a) 2013 and (b) 2014 at the high density (HDS) or low density (LDS) site on plants receiving 1 g dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), and plant volatile. Different letters above error bars denote significant differences between treatments...... 69

2.3 Density expressed as mean ±SE capture of larval D. carinulata per three sweeps in (a) 2013 and (b) 2014 at the high density (HDS) or low density (LDS) site on plants receiving 1 g dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), and plant volatile (PL). Different letters above error bars denote significant differences between treatments ...... 70

2.4 Mean ± SE damage rating for the high density site (HDS) in (a) 2013 and (b) 2014, and the low density site (LDS) in (c) 2013 and (d) 2014 on plants receiving 1 g dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), and plant volatile (PL). (a): BLa, PHb, PHPLb, PLb; (b): BLa, PHb, PHPLb, PLc; (c): BLa, PHb, PHPLb, PLa; (d): BLa, PHb, PHPLb, PLc. different letters behind each treatment denote significant differences between the means ...... 71

2.5 Mean ±SE dieback rating 2014-2015 at the (a) high density site (HDS) or (b) low density site (LDS) on plants receiving 1 g ix

LIST OF FIGURES CONTINUED

Figure Page

dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), and plant volatile (PL). Different letters above error bars denote significant differences between treatments ...... 72

2.6 Mean ± SE canopy volume estimated 2013-2015 for (a) the high density site (HDS) or (b) the low density site (LDS) on plants receiving 1 g dollops of treatment formulation. Treatments applied included blank (BL), pheromone (PH), pheromone and plant volatile (PHPL), and plant volatile (PL). (a): BLa, PHa, PHPLa, PLa; (b): BLa, PHb, PHPLb, PLa, different letters behind each treatment denote significant differences between the means ...... 72

3.1 Mean ± SE of (2E,4Z)-2,4-heptadien-1-ol emitted from 4-g dollops ...... 100

3.2 Weekly mean ± SE capture per three sweeps of reproductive adult Diorhabda carinulata on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL). Densities captured 2014-15 of P generation overwintered adults (a, c) and F1 first generation of current year adults (b, d) ...... 101

3.3 Figure 3: Weekly mean ± SE capture per three sweeps of non-reproductive adult Diorhabda carinulata on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL). Densities of the non-reproductive overwintering F2 generation of adults captured in 2014 (a), and 2015 (b) ...... 102

3.4 Weekly mean ± SE larval Diorhabda carinulata capture per three sweeps on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or 4g dollops of blank SPLAT (BL), in 2014 (a), and 2015 (b)...... 103

3.5 Weekly mean ± SE damage rating for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank x

LIST OF FIGURES CONTINUED

Figure Page

SPLAT (BL), in 2014 (a) and 2015 (b)...... 104

3.6 Mean ± SE dieback rating for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL) in 2015 and 2016 ...... 105 3.7 Change in canopy volume 2014-2016 for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL)...... 106

4.1 Sampling for D. carinulata adults for immediate dispersal 10-14 days after the initial release. Mean adult density ± SE per 25 sweeps (a, c) and mean adult density ± SE per 3 min count (b) of D. carinulata at release sites. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control) ...... 136

4.2 Sampling for Mean adult density ± SE per 25 sweeps for release persistence after the initial release. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control). Post-release monitoring at 10-14 days after the initial release is designated as generation 0 on x-axis ...... 137

4.3 Sampling for D. carinulata adults for release persistence after the initial release. Mean adult density ± SE per 25 sweeps (a) and mean adult density ± SE per 3 min count (b) of D. carinulata at release sites. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control). Post-release monitoring at 10-14 days after the initial release designated as generation 0 on x-axis ...... 138

5.1 Gas chromatography profiles of volatiles collected from A) larval D. carinulata feeding on Tamarix; B) undamaged Tamarix; C) mechanically damaged Tamarix. *Peak identified as 4-oxo-(E)-2-hexenal ...... 173

5.2 Electroantennogram (EAG) response of D. carinulata to a negative control (dichloromethane) and 1,000 ng of 4-oxo-(E)-2-hexenal. xi

LIST OF FIGURES CONTINUED

Figure Page

Different letters below error bars denote statistical differences between the treatments ...... 174

5.3 Response (%) of adult D. carinulata in a two-choice olfactometer to 10 μl of larval hemolymph vs control (purified air). Different letters above the error bars denote statistical differences between the treatments ...... 175

5.4 Response (%) of reproductive adult D. carinulata in a two-choice olfactometer to four doses of 4-oxo-(E)-2-hexenal vs the control (dichloromethane). Different letters above the bars denote statistical differences between the treatments ...... 176

5.5 Response (%) of non-reproductive adult D. carinulata in a two-choice olfactometer to 1,000 ng of 4-oxo-(E)-2-hexenal vs the control (dichloromethane). Different letters above the bars denote statistical differences between the treatments ...... 177

S.5.1. 3rd instar larva induced to reflex bleed through pressure applied from forceps ...... 178

S.5.2. Excised D. carinulata head attached to recording electrodes of the electroanntenogram ...... 179

S.5.3. Electroantennogram (EAG) response of D. carinulata to a negative control (dichloromethane) and a positive control (1,000 ng of (E)-2-hexenal). Different letters below error bars denote statistical differences between the treatments ...... 180

S.5.4. Electroantennogram response ratio of D. carinulata adults to the ratio of positive control: negative control vs 4-oxo-(E)-2-hexenal: negative control. Different letters above error bars denote statistical differences between the ratios ...... 181

xii

ABSTRACT

The northern tamarisk beetle, Diorhabda carinulata (Desbrochers), is an approved and established classical biological control agent for saltcedars (Tamarix spp.). Adequate control of Tamarix has not yet been achieved in certain areas where D. carinulata has been released. Retaining beetle populations on sites where it has been released is problematic, and accurately monitoring D. carinulata populations to determine successful establishment is difficult. Negative, indirect impacts have also resulted from the agent’s establishment outside targeted treatment areas in the southwestern United States. Manipulation of D. carinulata spatial distribution with semiochemicals could potentially resolve or ameliorate these and other operational issues. Lures utilizing a specialized wax based matrix for the controlled release of semiochemicals were impregnated with a previously identified pheromone and/or behaviorally active host plant volatiles known to stimulate aggregation in D. carinulata. Emission of these compounds from the matrix was characterized using a push-pull volatile collection system, and quantified using gas chromatography-mass spectrometry. Observed release rates confirm that semiochemicals lures formulated with this matrix are a viable option for facilitating aggregation of D. carinulata under field conditions. The results of field-based assays indicate saltcedars treated with this semiochemical delivery system attracted and retained higher densities of D. carinulata than Tamarix that received a control (semiochemical free) lure. Higher densities of both adult and larval D. carinulata were recorded on treated plants. Semiochemically treated Tamarix plants also exhibited more damage, resulting in a greater decrease in canopy volume than control trees. The attraction and retention of D. carinulata to these species-specific semiochemicals on treated Tamarix plants also arrested the dispersal of newly released individuals, resulting in greater population growth. Repellent semiochemicals were also investigated for their potential to manipulate spatial distributions of D. carinulata in the field and behavioral assays conducted with reproductive adults demonstrated the ability of larval produced compounds to repel conspecific adults. These results indicate that semiochemical-impregnated media could be useful for detecting, retaining, and directing populations of D. carinulata. The use of semiochemicals could be used to potentiate low density populations, increase monitoring efficacy, retain adults on release sites, and repel D. carinulata from sensitive habitat. 1

CHAPTER ONE

INTRODUCTION

Tamarix

Taxonomy

In North America, ‘Tamarix’ is commonly used to refer to a complex of exotic shrubs, semi-shrubs, and trees, including four invasive species (Tamarix aphylla Karst, T. chinensis Lour, T. ramosissima Ledeb and T. parviflora DC) and their hybrids (Gaskin

2013; Gaskin and Schaal 2002). Kovalev (1995) speculated these species most likely arose during the Oligocene-Miocene aridization of Middle and Central Asian savannahs.

These ancient savannahs, which are now desertified, are considered recent centers of maximal species diversity for the Tamaricaceae. Phylogenetically young species such as

T. chinensis and T. ramosissima are thought to have arisen following the migration of the earliest Tamarix species, T. ericoides Rottler, from the center of evolutionary radiation of the genus into Central Asia (Kovalev 1995). These phylogenetically young species possess adaptations derived from the severe conditions of the northern Asian deserts, allowing for greater ecological flexibility. Adaptations to adverse conditions and challenging environments within the genus have also resulted in many species of Tamarix becoming weedy in introduced areas.

Two of the ten species of Tamarix that were introduced to North America, T. ramosissima and T. chinensis, have become especially invasive and are considered the most widespread and damaging of the Tamarix species (Kaufmann 2005). These species 2 are morphologically similar but geographically distinct; their allopatric distribution suggests that hybridization between T. ramosissima and T. chinensis is rare or absent in the native range (Gaskin and Schaal 2002). Tamarix chinensis occurs primarily in the northeast of China and into the Korean Peninsula (Baum 1978). Tamarix ramosissima is found from the Black Sea to the western portion of China (Baum 1978). With the introduction of these species to the United States, novel hybrids have been discovered between these and other Tamarix species (Gaskin and Schaal 2002; Gaskin and Shafroth

2005; Gaskin et al. 2012). It is thought that such hybridization events will lead to greater invasiveness because they can provide a means to increase genetic variation and introduce evolutionary novelty (Stebbins 1969; Williams et al. 2014). In many parts of

North America, Tamarix hybridization is so extensive that molecular testing seldom identifies individuals free of introgression from one or more additional congeners

(Gaskin et al. 2012). To simplify further discussion of this genus in North America, which is characterized by morphologically cryptic species and extensive hybridization, the Tamarix species and their hybrids will hereafter be referred to as Tamarix.

Introduction and Spread

Initial introductions of Tamarix to North America have been attributed to the horticultural industry. Tamarix was made commercially available for the first time in

1823 through the catalogue of New York City’s Old American Nursery (Horton 1964).

Tamarix subsequently became available to the western U.S. through Highland Nurseries of New York (Bowser 1957). Shortly after it became more widely available to the western states, the U.S. Department of Agriculture began mass cultivating Tamarix at a 3

Washington D.C. arboretum, specifically for re-distribution (Robinson 1965). This plant was mainly introduced to provide windbreaks and shade, stabilize stream banks, and to be grown as an ornamental. The first report of Tamarix escaping cultivation occurred in

Texas, in 1877, only 23 years after the plant had become available in the western states

(Robinson 1965).

For the next several decades, little attention was given to the increasing spread of

Tamarix (Brotherson and Field 1987). By the 1960s Tamarix had become widespread in

Arizona, New Mexico, Texas, Oklahoma, Kansas, Colorado and Utah (Robinson 1965).

Tamarix has now invaded riparian ecosystems across 17 western states, and is estimated to occupy 900,000 - 1,600,000 acres (Jarnevich et al. 2013; Zavaleta 2000). Di Tomaso

(1998) estimated that the total area infested by Tamarix has increased nationally 3-4% annually since the 1920s (approximately 44,500 acres per year). This genus has become the third most prevalent woody riparian plant complex in the western United States

(Friedman et al. 2005).

Biology

Adaptation of Tamarix to survive under desertification has resulted in the expression of diverse growth forms, allowing this genus to establish and thrive under broad environmental conditions (Kovalev 1995). Growth forms of Tamarix range from single-trunk trees to multi-stemmed shrubs, with a typical height  10 m (Baum 1978).

Growth forms are correlated with a mix of environmental (water availability, elevation) and genetic factors (Horton and Campbell 1974; Kerpez and Smith 1987; Everitt 1980).

Plants growing alone and undisturbed are rarely taller than 4m and produce a spherical 4 canopy, while plants growing under disturbed conditions are shorter, usually with a smaller canopy (DeLoach 1991). Seedling densities can reach 170,000 individuals m-2 on mud flats (Warren and Turner 1975). Dense seedling patches undergo extensive self- thinning, resulting in mature stand densities of less than one individual m-2 (DeLoach

1991). Surviving plants often form impenetrable stands. A tall (11m), upright morphotype with little foliage usually results when Tamarix grows in dense stands

(DeLoach 1991). Foliage for this growth form is restricted to the crown of the canopy, with up to 98% of the photosynthetic foliage located in the uppermost meter of the canopy (DeLoach 1991).

The leaves of Tamarix are small, scale-like, overlapping and range in size from 3-

9 mm (Brock 1994). Leaves comprise less than 50% of the photosynthetic area of a

Tamarix plant (DeLoach 1991). The majority of photosynthesis is conducted in cladophyll stems (Wilkinson 1966). Cladophyll stems are photosynthetic branches that function similarly to leaves. The leaves and cladophyll stems of Tamarix contain salt excretion glands (Kleinkopf and Wallace 1974; Dressen and Wangen 1981; Thomson et al 1969)

Tamarix root systems are very extensive with a deep tap root and expansive lateral roots (Horton 1960; Gasry 1963). Depth to ground water appears to be the main factor affecting maximum root length in Tamarix, with deep tap roots forming when the water table is greater than 1m below the soil surface (Gasry 1963). Extensive lateral growth of the root system occurs on sites where the water table is shallow (< 1 m) (Gasry

1963; Merkel and Hopkins 1957). 5

Tamarix flowering continues for an extended period from spring through summer, depending on available resources (Anderson and Nelson 2013). Small pink to white bisexual flowers are densely arranged along 2-5 cm racemes or in leaf axils (Lindgren et al. 2010; Quigley 2013). Fertilization reportedly occurs via generalist pollinators and wind, although wind pollination has been disputed (Stevens 1989). Tamarix attracts a surplus of invertebrate visitors (and potential pollinators) due to the small amount of nectar produced by its abundant blossoms (Anderson and Nelson 2013; Strudley and

Dalin 2013). Since Tamarix blooms from spring to fall, it can provide nectar resources to pollinators when other nectar sources disappear or become scarce (DeLoach 1991).

Seeds are produced by Tamarix throughout the growing season whenever soil moisture is adequate to support active plant metabolism (Jacobs and Sing 2007). Mature trees produce hundreds of thousands of seeds (Browser 1957; Warren and Turner 1975).

The minute seeds (0.17mm diameter; 0.45mm length) are contained within capsules

(Merkel and Hopkins 1957). Each seed has an apical pappus, a tuft of hairs arranged to form an umbrella or parachute which aids in wind and water dispersion (Baum 1966).

The seeds shed the pappus and can germinate as soon as 24 hr. after coming into contact with moisture (Horton 1960). The seeds often germinate as spring flooding subsides and silt bars and beaches become available for colonization. Warren and Turner (1975) found seedling densities reaching 170,000 individuals/m2 on a reservoir mud flat. The seeds have a high rate of germination, 57% within 24 hours, but are only viable for a few weeks in summer conditions due to desiccation (Brock 1994). Seed viability dramatically decreases after three to four months (Merkel and Hopkins 1957). In addition to the 6 copious amounts of seed produced, Tamarix can also reproduce vegetatively. If a plant is buried during a flooding event, the buried stems can produce roots and new shoots

(DeLoach 1991; Everitt 1980).

Ecology

The extensive root system allows Tamarix to acquire water from both the phreatic and vadose zone, which classifies Tamarix as a facultative phreatophyte. It can tolerate periods of drought and inundation, allowing for colonization both of riparian areas and upper terraces far from water sources (Smith et al. 1998; Natale et al. 2010).

Water use by Tamarix has long been blamed for the changes in riparian hydrology in the western United States (Robinson 1965; Johnson 1987; Di Tomaso 1998). Early studies of Tamarix labeled the plant as an excessive water consumer capable of drying up sources of water (Brotherson and Field 1987). A commonly cited estimate of Tamarix water use is 757 Lday-1 for a single plant (DiTomaso 1998). This estimate has been attributed to unsupported statements from earlier studies (Owens and Moore 2007).

However, new technological developments that allow for unobtrusive temporal and spatial measurements of evapotranspiration by Tamarix are calling into question this species’ status as a hyper-consumer of water (Glen et al. 2007; Nagler et al. 2009).

Recent revised estimates of daily water use by Tamarix derived from sap flux measurements put the value at approximately 122 Lday-1 (Owens and Moore 2007).

Water use by Tamarix at this revised rate does not differ significantly from other native riparian plants (Dahm et al. 2002; Glen and Nagler 2005). These lower rates of water use call into question the management practice of controlling Tamarix along riparian 7 corridors to salvage water, and may explain the lack of surface water savings realized from Tamarix removal programs (Owens and Moore 2007; McDonald 2010).

While Tamarix stands have evapotranspiration rates similar to other riparian plants, Tamarix can grow and invade broader areas of the floodplain than native riparian species, invading even upper riparian terraces, which can be as dry as non-riparian zones

(Cleverly 2013; Whitcraft et al. 2007; Hultine and Bush 2011; Reichenbacher 1984).

Invasion of these upper terraces can shift resident plant community composition away from non-riparian species to domination by riparian Tamarix, resulting in increased upper terrace evapotranspiration rates, which can have a major impact on downstream water availability (Hultine and Bush 2011).

Tamarix can survive in the presence of high soil salt levels and can utilize saline ground water by secreting excess salt through the glands on its leaves and cladophyll stems (Kleinkopf and Wallace 1974; Dreesen and Wangen 1981; Thomson et al. 1969;

Shafroth et al. 1995). The salt glands excrete a broad array of salts and minerals including chlorine, carbonates, sodium, potassium, bromine, calcium, nitrate, magnesium, and sulfates (Waisel 1960; Hem 1967; Berry 1970). Leaf fall in the autumn results in a major pulse of salts to the soil surface, altering soil and ground water salinity (Carman and

Brotherson 1982; Ladenburger et al. 2006; Salinas et al. 2000). Deposition of salts by

Tamarix has been referred to as elemental allelopathy (Brock 1994; Di Tomaso 1998;

Morris et al. 2009).

The spread and persistence of Tamarix stands throughout the southwestern United

States is correlated with frequent and intense fires in riparian areas (Busch 1995). 8

Alteration of fire regimes can have devastating effects on native riparian plants, as these native species are not typically fire adapted (Ellis 2001; Rood et al. 2007). While select native plants have evolved a few minor fire tolerance adaptations, Tamarix, in contrast, is well adapted to fire (Busch and Smith 1993). When above ground biomass is consumed and removed by fire, Tamarix readily re-sprouts from unaffected root crowns below the soil surface that are protected from burning (Busch and Smith 1993). Salt and drought tolerance, plus its extensive and robust root system allows for more rapid recovery of

Tamarix compared to native plants in burned areas (Carreira and Niell 1992). As fire frequencies increase in riparian areas in the western United States, Tamarix abundance and dominance is maintained with continued decline and local extirpation of native species (Ohmart and Anderson 1982).

Impacts

Riparian corridors of the arid and semi-arid western United States historically supported novel and diverse flora and fauna, especially in the southwest (Patten 1998).

Native plant communities in these areas have evolved under predictable disturbance from yearly flooding events and are dependent on this source and type of disturbance to maintain biodiversity, optimal age class distributions, and community equilibrium (Poff et al. 1997). These diverse communities are now threatened by the invasion of Tamarix.

The biology of Tamarix allows it to utilize ground water and surface water, which increases soil salinity, and increases wildfire return rates and intensity. Increased soil salinity, detrimental fire effects and dense canopy cover decrease the competitiveness of native plants. Tamarix is readily capable of replacing 50% of native riparian vegetation, 9 commonly displacing resident vegetation to form large monocultures (Horton and

Campbell 1974). In riparian zones, Tamarix commonly displaces native cottonwoods

(Populus spp.) and willows (Salix spp.), and on upper terraces replaces stands of mesquite (Prosopis spp.) (Frasier and Johnsen 1991; Cleverly et al. 1997).

By the 1950s, the majority of riparian areas along major waterways in the western

United States were occupied by Tamarix. Rapid Tamarix invasion in the United States coincided with the construction of major dams in the Desert Southwest, and with riparian vegetation control programs that removed native plants in an effort to conserve water (Di

Tomaso 1998, DeLoach et al 2000). When the dams were built, they eliminated most of the seasonal flooding that was common in the Southwest. This elimination of cyclical seasonal flooding events favored the propagation of Tamarix over native species (Everitt

1980; Irvine and West 1979; Petranka and Holland 1980).

The displacement of native plants by Tamarix is promoted by anthropogenic habitat alterations, but does not account for the spread of Tamarix into unaltered habitat.

Tamarix is now invading undisturbed tributaries, small streams and desert springs

(Lovich and DeGouvenain 1998; Barrows 1998). It is also spreading rapidly into undisturbed natural habitat in Montana and North Dakota (Pearce and Smith 2003; Pearce and Smith 2007). A study conducted by Cleverly et al. (1997) found that the displacement of native plants by Tamarix was due to its superior drought tolerance; periodic, recurring droughts over a fifty year period diminished the population of native plants, with little or no negative effects on Tamarix populations 10

Colonization by Tamarix has degraded the quality of native plant communities and consequently, wildlife habitat. This is primarily due to the replacement of native plant species and assemblages with the monocultures of Tamarix (Dudley et al. 2000).

Tamarix’ ability to replace native vegetation has contributed to the decline of many native bird species (Dudley and DeLoach 2004). Anderson and Ohmart’s (1984) study of the avian use of Tamarix concluded that Tamarix was the identified variable most negatively correlated with avian diversity and density. One reason for such poor assemblages of birds in Tamarix stands is depauperate associated insect assemblages.

Very few insects feed on Tamarix in North America (Liesner 1971). The lack of insects greatly reduces the food sources of many of the insectivorous avian species common in these riparian areas.

On the Rio Grande River in western Texas and the Pecos River of New Mexico, populations of furbearers and rodents were lower in areas dominated by Tamarix compared to areas dominated by native vegetation (Engel-Wilson and Ohmart 1978). In

Big Bend National Park, beavers have been nearly eliminated because of Tamarix invasion (Boeer and Schmidly 1977). It has been shown that springfed ecosystems dominated by Tamarix have decreased populations of Ash Meadow pupfish (Cyprinodon nevadensis mionectes Miller) and Ash Meadow speckled dace (Rhinichthys osculus nevadensis Girard), two endangered species endemic to the Mojave Desert (Kennedy et al. 2005). DeLoach and Tracy (1997) reviewed 51 threatened or endangered (T&E) species or candidate T&E species that occur in areas infested by Tamarix. Forty of the fifty one T&E species were negatively impacted by Tamarix. As Tamarix displaces more 11 and more native vegetation, suitable habitat for these T&E species will become scarcer, resulting in further decreases in already imperiled populations (Dudley et al. 2000).

There are also significant economic losses associated with Tamarix invasion, estimated to cost the western United States between $133-185 million each year (Duncan et al. 2004; Zavaleta 2000). These losses are mostly realized in decreased ecosystem services and functions. Tamarix also decreases the recreational use of parks for camping, hunting, and fishing, boating, birdwatching, and photography. This occurs due to the impenetrable stands formed by Tamarix, accompanied by decreases in desirable species.

However, not all Tamarix characteristics are detrimental, and some detrimental effects have been questioned. Tamarix is important for minimizing soil erosion, specifically along stream banks. This erosion control was needed to maintain and support agricultural operations in the southwestern United States (Chew 2009). Large monocultures of Tamarix can be used as maintenance plants for honeybee production

(DeLoach et al. 2004). It has been speculated that in many areas highly or chronically disturbed by humans, native vegetation may never fully recover (Lovich and Bainbridge

1999). Since the disturbed area cannot support native vegetation, Tamarix can provide basal functions of the original ecosystem. An obvious example of this is the use of

Tamarix as nesting habitat for the endangered southwestern willow flycatcher

(Empidonax traillii extimus), a bird that predominately nests in native willows (DeLoach et al. 2004).

Control

Even though the importance of the few benefits Tamarix provides can be debated, the ecological damage caused by Tamarix far outweighs the benefits it provides

(DeLoach et al. 2004). The damage caused by Tamarix has necessitated the development of diverse management strategies and implementation of massive control programs

(Robinson 1952). The purpose of these control program is to restore the native component of riparian communities, conserve water and enhance water conveyance, and protect habitat for endangered species (McDaniel and Taylor 2003). By the 1940s

Tamarix was being managed with herbicide applications, ground water pumping, and mechanical removal with bulldozers and mowers (Robinson 1952; Robinson 1965;

Weeks et al. 1987). Despite these control programs, populations of Tamarix continued to expand, with concurrent expansion of the scope and size of the programs to control it

(Douglass et al. 2013).

One common weed management tactic for Tamarix is the mechanical removal of plants. This approach removes above ground growth of the plants by mowing, cutting, plowing, and crown removal (Brock 1994; Smith et al. 2002). Mechanical removal of the above ground biomass alone is not an effective control measure since Tamarix can re- sprout from underground buds (Robinson 1952). However, mechanical removal of the plant can result in canopy reduction for several years, allowing native plants to re- establish and revegetate the area (Graf 1979). If plant mortality is the desired outcome from mechanical removal, it can be achieved if the plant crown is removed or the root system is severed (McDaniel and Taylor 2003). Discing of Tamarix seedlings coupled 13 with above ground biomass removal has resulted in mortality rates of 99% (Smith et al.

2002). While mechanical removal can be an effective means of controlling Tamarix, the labor requirements and cost to treat large areas is prohibitive (McDaniel and Taylor

2003).

Herbicides are also commonly employed in an attempt to control Tamarix (Brock

1994). Historically, herbicidal control of Tamarix has been unsatisfactory (Duncan and

McDaniel 1998). Herbicides are commonly applied to the foliage, the basal bark of the plants, and freshly cut stumps (Douglass et al. 2013). Imazapyr is the most commonly used and effective herbicide for chemically controlling Tamarix (Duncan and McDaniel

1998). However, imazapyr is persistent in the soil and therefore has long term residual activity. Glyphosate can be used to control Tamarix, but is not very effective when used alone (Fick and Geyer 2010). Tank mixes of imazapyr and glyphosate resulted in 90-99% mortality of treated plants (Duncan and McDaniel 1998). However, this method is cost prohibitive, especially if complete control of the stand is not achieved and re-colonization of Tamarix occurs.

Classical Biological Control

Tamarix was targeted for a classical biological control program because of its widespread establishment, invasive status and the extensive economic and environmental damage it has caused in the western United States (Zavaleta 2000; Friedman et al. 2005).

Few insects utilize Tamarix in the introduced range, while in its native range, Tamarix supports a large number of host specific and damaging insects (Kovalev 1995; DeLoach et al. 2003). The fact that there are no native or economically important plant species 14 closely related to Tamarix in North America also makes Tamarix an excellent candidate for biological control, specifically classical biological control (McFayden 1998; DeLoach et al 2000; Pemberton 2000). Classical biological control is the intentional introduction and release of an exotic pest’s coevolved natural enemy for the desired outcome of suppressing the pest’s population below economically and ecologically damaging levels.

In its native range, Tamarix is capable of dominating riparian areas but rarely to the extent observed in the United States (Dudley et al. 2000). The infrequent dominance of riparian areas in the native range could be due to suppression by insect herbivores

(Gerling and Kugler 1973; Kovalev 1995). A biological control program for Tamarix was therefore initiated in the United States to replicate contributions to Tamarix population regulation attained in the native range by specialist insect herbivores (DeLoach 2003;

Tracy and Robbins 2009). One specialist herbivore with potential for Tamarix control is the northern tamarisk leaf beetle, Diorhabda carinulata Desbrochers (Coleoptera:

Chrysomelidae).

Diorhabda carinulata

Evaluation and Selection

Host range testing of D. carinulata began in North America in 1992 in U.S.

Department of Agriculture, Agricultural Research Service containment facilities located in Temple, Texas and Albany, California (DeLoach 2003). Host range was initially evaluated on six species and three hybrids of Tamarix, followed up by host specificity testing on 58 additional plant species (DeLoach 2003). Plants used to characterize the 15 host range of D. carinulata were selected based on the centrifugal phylogenetic system developed by Wapshere (1974) and Harris and Zwölfer (1968).

In larval no-choice tests, D. carinulata was only able to develop on the Tamarix species, the Tamarix hybrids, plus the following related species: Frankenia jamesii

Torrey, Frankenia johnstonii Correll, Frankenia salina Molina, and Myricaria germanica

Royle (DeLoach 2003). These results indicated that when D. carinulata was presented with the option of foliar consumption or starvation, feeding occurred only on four plants that were not members of the target weeds’ family, Tamaricaceae. However, since these

Frankenia species occur in North America, further studies were needed to determine the risk D. carinulata posed to these nontarget shrubs. To address these concerns, Lewis et al. (2003a) conducted laboratory, greenhouse and outdoor cage studies to determine if D. carinulata larva could develop on Frankenia spp. and if adults are attracted to, fed or oviposited on Frankenia spp. Larvae were found to be capable of completing development on Frankenia spp.; however, adult beetles were not attracted to the plants and this allowed for the eventual open release of D. carinulata in North America (Lewis et al. 2003a; Dudley and Kazmer 2005).

After thorough testing of host preference and survival of D. carinulata in the

United States, it was determined that the host range of D. carinulata was sufficiently limited and it was approved by the United States Department of Agriculture –Animal and

Plant Health Inspection Service (USDA-APHIS) for release in California, Nevada, Utah,

Colorado, Wyoming and Texas (U.S. Department of Agriculture Animal and Plant Health

Inspection Service 1999). The first beetles released in the U.S. were collected in Fukang, 16

China and Chilik, Kazakhstan; adults were released into cages at 10 locations in 1999

(Lewis et al. 2003b). Initial releases made into cages to optimize the success of these first attempts at establishment were followed by open field releases beginning in 2001

(Lewis et al. 2003b).

In addition to D. carinulata, three other species of Diorhabda were tested and approved for release. These species were tested and released because of the lack of establishment of D. carinulata below the 38th North parallel in North America. Release of the additional Diorhabda species was also a potential solution for the lack of host plant association between invasive Tamarix parviflora and D. carinulata. The three additional

Diorhabda species released were D. sublineata Lucas, D. elongata Brullé, and D. carinata Faldermann. Diorhabda sublineata became established in Texas; D. carinata established in California and D. elongata became established in both Texas and

California. These populations are located in areas where D. carinulata initially failed to establish (Carruthers et al. 2008; Dalin et al. 2009).

Life Cycle

Females place eggs singly or in tight clusters of up to 25 eggs on foliage, with each female producing approximately 200 eggs (Lewis et al. 2003b). The eggs are spherical and light tan in color. Eggs are more commonly deposited in clusters, and hatch about five to seven days after oviposition (Lewis et al. 2003b).

Diorhabda carinulata has three larval instars that can be distinguished by stage- specific coloration. The first instar is completely black; the second instar is black with yellow spots around each spiracle, while the third instar is grey with a yellow thoracic 17 shield (Lewis et al. 2003b). All instars feed on the foliage of Tamarix (DeLoach et al.

1996). The first two larval instars feed primarily on the terminal leaves and stems while the larger third instar larvae typically move inward toward the host plant’s trunk or primary stem(s), consuming more mature foliar tissues after the supply of younger terminal foliage has been exhausted (Lewis et al. 2003b). At the completion of the third instar, larvae cease feeding and drop from the canopy of the plant to the leaf litter below, where they form a pupal cell (Lewis et al. 2003b). The mean time to progress from one instar to the next is approximately 3-7 d. After metamorphosis begins, 5-8 d pass before adults emerge from the soil (Lewis et al. 2003b). Generation time varies from approximately 37-41 d and populations can increase from 2-25 fold for each successive generation (Lewis et al. 2003b).

Pupation is completed entirely in the soil litter, after which the newly emerged adults feed for several days before becoming reproductively active (Bean et al. 2007a).

Reproductively mature adults mate several times before laying eggs (Bean et al. 2013). In

North America, D. carinulata populations are, multivoltine, with the northernly distributed populations typically having two adult generations each year (Lewis et al.

2003b). First generation adults are reproductively active, while second generation adults enter reproductive diapause at the onset of autumn (Lewis et al. 2003b; Bean et al. 2007b;

Milbrath et al. 2007). The occurrence of reproductive and non-reproductive generations is controlled by day length (Bean et. al 2007a). The photoperiod that induces diapause occurs when daylight hours are shorter than 14 hours and 39 minutes, requiring 5-20 consecutive days spent at those day length conditions to induce diapause (Bean et al. 2007a; Bean et al. 2012). Adult D. 18 carinulata that are in or entering reproductive diapause display altered behavior compared to reproductive adults. Adults in reproductive diapause do not exhibit dispersal related behaviors, nor do they aggregate (Cossé et al 2006; Bean 2007b). Reproductive diapause triggers increased feeding, followed by movement of adults into the soil (Lewis et al.

2003b). Once favorable temperatures return, adults emerging from the soil become reproductively active (Bean et al. 2007a, Bean et al. 2013). Emergence of D. carinulata in the spring is independent of day length, but controlled by temperature (Lewis et al. 2003b). Spring emergence coincides with bud break in host plants (Lewis et al. 2003b).

The day length required to induce diapause is critical to winter survival in the native range; this also restricts the distribution of D. carinulata in North America. At latitudes south of 38° N, day length is never greater than the critical photoperiod that induces diapause. Below this latitude D. carinulata becomes univoltine and enters diapause in early summer (Bean et al. 2007a). Adults in reproductive diapause are exposed to hot and dry conditions in the American Southwest. High temperatures during diapause increase mortality rates in the adult beetles, likely caused by an associated increase in metabolic rate that depletes dietary reserves (Thompson and Davis 1981).

Winter field conditions kill adults with depleted metabolic reserves. Mortality rates in southerly populations, beginning with starvation, are further increased because entry into the soil during the summer exposes D. carinulata to predatory ants, a significant mortality factor in D. carinulata populations (Bean et al. 2007a).

Rapid evolution was evident in North American populations of D. carinulata within five years after initial introduction (Bean et al. 2012). Critical day length inducing diapause decreased significantly. Populations of D. carinulata near Artesia, New Mexico entered diapause at a day 19 length 44 min shorter than previously recorded (Bean et al. 2012). This post-release phenological adaptation resulted in greater synchrony between the life cycle of D. carinulata and the host plant under field conditions at more southerly locations. The decrease in critical day length required to induce diapause also increased the reproductive period for D. carinulata by delaying overwintering behavior in the autumn. Increased reproductive activity allowed for greater population growth, counteracting the influence of predation (Dudley and DeLoach 2004). Delayed onset of overwintering behavior also reduced stress caused by high ambient temperature during diapause.

The decrease in critical day length required to trigger diapause was key to the expansion of the southern range of D. carinulata.

Dispersal

Reproductive D. carinulata readily disperse over both short and longer distances with hill topping and swarming behaviors frequently observed (Cossé 2005; Bean et al.

2013). This behavioral pattern usually results in swarms of beetles moving from a few meters to 20-30 m, causing relatively limited dispersal and population movement even when food is abundant. Observations made in Texas recorded beetle population dispersal at a two year total of 320 m (Knutson and Muegge 2008). An alternative behavior that has been observed in the field is long distance mass dispersal of beetles when food sources become limiting, including if their preferred host is not present (DeLoach et al.

2004; Hudgeons et al. 2007a).

Rapid population growth resulting in defoliation of large swaths of Tamarix by D. carinulata caused dispersal rates much greater than anticipated (DeLoach et al 2004;

Pattison et al. 2011; Nagler et al. 2014). Beetle populations originally anticipated to 20 disperse at a rate of 1-2 km per year have been observed dispersing 25 km per year

(Nagler et al. 2014). These alternating behaviors of short and long distance dispersal allow for the buildup of extremely high population densities of D. carinulata in areas heavily infested with Tamarix, with subsequent mass dispersal of the population to new areas after defoliation of local host vegetation. This strategy is commonly referred to as levy flights, where a mix of short and long dispersals are used to maximize the search for food (Reynolds and Rhodes 2009). Key to the ability of D. carinulata to maintain high densities when moving both short and long distances is the utilization of aggregation pheromones enhanced by Tamarix-produced semiochemicals (Grevstad and Herzig 1997;

Fernandez and Hilker 2007; Wood 1982; Cossé 2005, 2006).

Release and Establishment

The original releases of D. carinulata were made in field cages in the United

States in 1998, 1999, and 2000. In 2001, beetles that had survived and established in cages were collected and open field releases of 400 adults were made at seven locations

(DeLoach et al. 2004). Adults destined for the first open field releases were initially held for 1-2 weeks in sleeve cages on branches of Tamarix plants, which resulted in mating and egg laying on the caged foliage. Once oviposition was observed, the sleeves were removed and the adults were free to disperse into the surrounding vegetation (DeLoach et al. 2004). Openly released beetles were observed to rapidly disperse (Carruthers et al.

2008). Along with initial open releases of 400 adults, supplemental releases were made throughout the summer of 2001 to increase the likelihood of establishment (DeLoach et al. 2004). 21

Very little beetle activity was observed during the first two years of open field releases (2001-2002). Detectable populations of D. carinulata were recorded at most of the release sites but required monitoring over large areas to locate beetles (Carruthers et al. 2005). Extensive defoliation by third instar larvae was first observed at the end of

August 2002 (Carruthers et al. 2008). Rapid population increases continued, with one site experiencing 10,000 hectares of defoliated Tamarix by 2004 (DeLoach 2004; Carruthers et al. 2008; Pattison et. al 2011). By 2003, it was apparent that releases made at

Lovelock, NV; Delta, UT; Lovell, WY; Pueblo, CO and Schurz, NV had become established (DeLoach 2004; Carruthers et al. 2005).

Releases of D. carinulata made in California and Texas did not establish. The failure of this species to establish in locations south of the 38th North parallel was attributed to short day length triggering early diapause and depletion of metabolic reserves to levels inadequate for winter survival (Bean et. al. 2007a; Milbrath et al. 2007).

Impacts on Tamarix

Larval and adult D. carinulata feed almost exclusively on Tamarix, with the exception of Myricaria germanica and Frankenia spp. as mentioned above. Defoliation is the major impact of D. carinulata feeding on Tamarix. Larvae and adults feed by rasping and scraping the leaves and photosynthetic stems (Dudley and Kazmer 2005). This ultimately results in leaves and stems desiccating and dying (Snyder et al. 2010).

Desiccation and death of photosynthetic tissues is termed a defoliation event despite limited biomass removal by feeding (Snyder et al 2010). 22

High reproductive capacity, coupled with successful aggregation and dispersal of beetle populations results in extensive defoliation of Tamarix by D. carinulata.

(Carruthers et al. 2008; Lewis et al. 2003b). With vigorous establishment in many states, expansions of beetle populations and extensive defoliation of large areas have been observed (DeLoach et al. 2004). Large, multivoltine populations can defoliate Tamarix over the course of a season, but defoliation from a single generation of beetles has also been observed (DeLoach 2004; Carruthers et al. 2008; Pattison et al. 2011). Defoliation events cause extensive stress to the Tamarix plants (Hultine et al. 2011; Snyder et al.

2010). Repeated defoliation events can reduce levels of water soluble sugars from 9-14% to less than 2%, and deplete root crown nonstructural carbohydrates (Hudgeons et al.

2007b). These carbohydrate reserves are vital for Tamarix survival and persistence

(Loescher et al. 1990; Sala et al. 2012). A compromise in the supply of adequate carbohydrate reserves also affects Tamarix recovery and regrowth after defoliation, resulting in reduced leaf production and root mass in the following year (Hudgeons et al.

2007b).

Defoliation by D. carinulata also causes water loss in Tamarix plants. Rasping and chewing by feeding adults and larvae disrupts leaf and stem epidermal cells, rendering Tamarix unable to effectively regulate water (Snyder et al. 2010). The inability of defoliated plants to maintain water balance results in desiccation (Snyder et al. 2010).

Because desiccation occurs rapidly, plants are unable to translocate nutrients before the leaves and photosynthetic stems die (Snyder et al. 2010; Conrad et al. 2013). Rapid loss 23 of leaf area greatly decreases photosynthesis. A modest level of D. carinulata foliar damage (30%) can reduce annual stem and root growth (Williams et al. 2014).

Tamarix can recover from defoliation events, and regrowth commonly occurs both during the current season and in the following spring (Dudley and DeLoach 2004).

Frequently, plants that were defoliated and have regrown will subsequently be avoided by

D. carinulata (Dudley et al. 2012). Avoidance can be sustained over multiple years, allowing plants to avoid mortality in the short term (Dudley and DeLoach 2004). Tamarix can recover from limited defoliation, but repeated defoliations generally result in plant mortality (Hudgeons et al. 2007b). Field observations show that two years of defoliation can result in a 25-33% reduction in plant canopy area (Dudley and DeLoach 2004). One of the more extreme examples of impact due to defoliation was the reported reduction of plant volume by 93% over a four year period in Nevada (Pattison et al. 2011). The repeated defoliation observed at this site not only reduced canopy size, but 40% of trees with four continuous years of defoliation died (Dudley et al. 2006). Mortality rates depend on many interacting factors such as stand age, number and timing of defoliation events, plant genetics, availability of resources and occurrence of other stress inducing events like fires and drought (Hultine et al. 2010; Williams et al. 2014; Kennard et al.

2016; Drus et al. 2014; Hultine et al. 2015). The rates of mortality in Tamarix stands can be highly variable, ranging from 0-70% (Kennard 2016; Bateman et al. 2010).

Challenges and Conflicts

While many of the initial releases of D. carinulata were successful, the methods used were labor intensive, expensive, and in the some cases relatively ineffective for 24 establishing sustainable populations. Beetles had to be contained in field or sleeve cages, with constant collection and redistribution of the adults (Knutson and Muegge 2008). For open field releases, 5,000 individuals were recommended to ensure establishment

(Knutson and Muegge 2008). The total number of adults used in the initial open field release approached 60,000 individuals. This number represents an extraordinary investment by biological control practitioners to rear this many individuals, as most biological control programs have limited numbers of agents available for making initial releases (Shea and Possingham 2000; Memmott et al. 1996). Of the 60,000 individuals available to practitioners between 2001 and 2003, almost half (27,000) were released at

Lovell, WY (DeLoach et al. 2004). This represents a common challenge many biological control practitioners face: is it better to make a single large release to ensure establishment by increasing the odds of successful reproduction, or multiple smaller releases to minimize the risk of local extinction events (Memmott et al. 1996; Hopper and

Roush 1993; Grevstad 1999)? Establishment was achieved at the Lovell, WY site, but D. carinulata numbers at this location continue to be limited compared to sites in Nevada and Colorado initiated with much smaller release densities (DeLoach et al. 2011).

Diorhabda carinulata is known to rapidly disperse from its release sites

(Carruthers et al. 2008). This adds another layer of complexity to the process of making a release, as dispersal of the adults from the release location can result in loss of fitness due to small effective population size (Hopper and Roush 1993). The loss of fitness due to low or small population density is known as an Allee effect. Accumulation or intensification of Allee effects can quickly result in population extinction (Boukal and 25

Berec 2002). When local releases of new agents are made at densities designed to minimize Allee effects, regional or national rapid expansion and distribution of effective agents is sacrificed, slowing the progress of many biological control programs. Even when large releases are made, establishment is not guaranteed. An extreme example of a large release failure are the multiple releases of thousands of beetles in Montana, specifically the releases made on Fort Peck Reservoir. Repeated releases of D. carinulata were made but establishment was unsuccessful (DeLoach et al. 2011).

While Lovell, WY was the site of the largest releases of D. carinulata, initial establishment was limited and little increase in beetle numbers was observed. By 2004, beetles were present and damaging Tamarix, but no large scale defoliation had been observed (DeLoach et al. 2004). It was not until 2007 that a large D. carinulata defoliation event was observed at Lovell, WY (DeLoach et al. 2011). The delay in observable impact could be due to many factors, such as overwintering mortality, predation, and population crashes due to climatic events (Dudley et al. 2006). All of these factors could prevent population increases necessary to control Tamarix. The lack of impact at Lovell is in direct contrast to D. carinulata releases on Boysen Reservoir, also in Wyoming. Tamarix at Boysen Reservoir has experienced significant D. carinulata mediated dieback and mortality in the years following the initial 2005 releases (Nancy

Pieropan, personal communication).

Monitoring D. carinulata at low densities can be very difficult as this species tends not to be evenly distributed in stands of Tamarix (Cossé et al. 2005). After the initial U.S. releases of the agent in 2001, extensive monitoring was required to detect 26 small, low density populations in 2002 (Carruthers et al. 2005). In instances where beetle densities are low, the likelihood of returning a false negative when monitoring is high.

False negative observations can lead to mismanagement of the system (Yoccoz et al.

2001). For example, if monitoring results return a false negative, then the land manager cannot adopt strategies that could boost low density populations of an agent because they are unaware that it is present. A monitoring strategy should be developed to minimize the chances of reporting false negatives when looking for successful D. carinulata establishment.

Along with the above challenges of establishing populations, impacting the target weed, and accurately monitoring for presence of the agent, biological control of Tamarix with D. carinulata has come into conflict with the conservation and management of avian species that occur in the southwestern United States (Paxton et al. 2011; Dudley and

DeLoach 2004). The most prominent case of a detrimental interaction between D. carinulata and an avian species is the interaction with the endangered subspecies of the southwestern willow flycatcher Empidonax traillii extimus (Dudley and Bean 2012).

In 1992, the southwestern willow flycatcher was petitioned for endangered status by the U.S. Fish and Wildlife Service, and subsequently federally listed as an endangered species in 1995 (Suckling et al. 1992; US Fish and Wildlife Service 1995). These conservation actions not only succeeded in gaining protected status for the bird, but also documented the use of Tamarix as nesting habitat by the southwestern willow flycatcher

(Sogge et al. 2013). In 2005, Tamarix was formally listed as critical habitat for the southwestern willow flycatcher (Federal Register 2005). 27

The major concern of land management agencies about the interaction between D. carinulata and the southwestern willow flycatcher was that defoliation of Tamarix trees hosting nesting flycatchers could result in reduced nest canopy cover. Canopy reduction could increase the risk of nest exposure to lethal temperatures, desiccation, and elevate contact with predators and parasitic cowbirds (Sogge et al. 2003, 2008; DeLoach 2003).

Along with concerns over habitat degradation associated with the potential reduction of

Tamarix canopy cover by D. carinulata feeding, habitat loss was also feared, in the event that defoliation by the beetle succeeded in killing Tamarix trees before native plant communities could recover and provide suitable nesting habitat for the southwestern willow flycatcher. This would result in a gap in the availability of nesting habitat, worsening the situation for this species (Sogge et al. 2008).

To minimize and abate fears of detrimental interactions between D. carinulata and the southwestern willow flycatcher, a moratorium was put into effect preventing release of the beetle within 320 km of any southwestern willow flycatchers known to be nesting in Tamarix (Carruthers et al. 2008; Dudley and Bean 2012). This 320 km buffer prevented the release of D. carinulata in the entire state of Arizona, southern portions of

Nevada, Utah, and Colorado, the western portion of New Mexico, and much of southern

California. The majority of Tamarix nesting southwestern willow flycatchers occur in

Arizona and New Mexico (Sogge et al. 2006). This buffer was thought to minimize the risks of D. carinulata interacting with the southwestern willow flycatcher, especially since the beetles appeared to be incapable of establishing south of the 38th North parallel

(Bean et al. 2007a). The critical habitat for the southwestern willow flycatcher is south of 28 the 38th North parallel. However, in 2006, beetles were moved in an unauthorized release into the flycatcher zone, prompting a lawsuit against USDA-APHIS and US-FWS

(Center for Biological Diversity 2009). This lawsuit resulted in USDA-APHIS’ termination of interstate transport permits for Diorhabda spp., and the issue of a moratorium on biological control of Tamarix using any “species, subspecies or ecotypes of the D. elongata complex” which if contravened through unauthorized collection, transport or release of the beetle could result in criminal punishment or fines under provisions of the Endangered Species Act and Plant Protection Act (APHIS 2010). The termination of interstate transport affected all D. carinulata programs, even in areas where the southwestern willow flycatcher does not occur. The termination of the biological control program for Tamarix, while reducing potential risks D. carinulata poses to avian species, also reduces the protection and conservation of natural ecosystems that Tamarix biological control could have provided, especially considering Tamarix biological control is the method most resilient to secondary invasion of other noxious weeds (González et al. 2017a; González et al. 2017b).

Contrary to predictions based on physiological limits (e.g., day length, ambient temperature during reproductive diapause, etc.) observed from the initial U.S. releases of the agent, D. carinulata was able to successfully establish, build to high densities and spread throughout the southwest willow flycatcher zone. Diorhabda carinulata’s ability to establish south of the 38th North parallel is the result of rapid evolution of the photoperiod requirement for diapause initiation (Bean 2012). Populations of D. carinulata are now established throughout much of the northern portion of the 29 southwestern willow flycatcher distribution (Dudley and Bean 2012). Because the southward expansion of D. carinulata populations introduces a potential source of risk to vulnerable populations of the southwestern willow flycatcher, D. carinulata continues to be an ongoing concern for federal, state and private wildlife management agencies and organizations (Sogge 2008). Furthermore, these concerns are not limited to the fate of the southwestern willow flycatcher, but may be expanded to include a number of regional avian conservation targets (Paxton 2011).

Chemical Ecology and Insects

Chemical ecology is the study of compounds used in communication by organisms. A major aspect of chemical ecology research is the identification of these compounds, referred to as semiochemicals. Semiochemicals are compounds that convey a signal from one organism to another. These communications can be intraspecific and interspecific. When a semiochemical produces intraspecific communication, it is referred to as a pheromone. In insects the chemosensory neurons are contained in sensory hairs called sensillae. Olfactory sensillae are contained within two sensory organs on the head of insects, the antennae and maxillary palps, while gustatory sensillae are present across the body of the insect, most notably in the mouthparts, legs, wings and ovipositor

(Dahanukar et al. 2005).

While gustatory stimuli can play a critical and informative role in insect chemical ecology (Chapman 2003), olfactory signals are the focus of this research. Olfaction is the detection of an airborne volatile chemical. Odors play a role for insects in the perception 30 of the environment around them, allowing them to determine availability of food, oviposition sites, mates, and inducing aggregation and dispersal responses (Bruce et al.

2005; Jermy and Szentesi 1978; Dethier 1948). Many of the compounds that mediate these behaviors are volatile compounds that facilitate communication by organisms over long and short distances (Borges et al. 1987; Wall and Perry 1987; Zhang et al. 2003).

The quantity of compound required to elicit a response can be very small. For example, female cabbage loopers (Trichoplusia ni Hubner) produce 1-2 ng of sex pheromone in order to attract a mate (Shorey and Gaston 1965). In order to maintain competition and reproductive isolation, many insects respond to sex pheromones that represent a very specific blend and narrow range of compounds (Wyatt 2003; Cardé and Haynes 2004).

Pheromones produced by insects can be subdivided into five functional groups.

(1) Sex pheromones are chemicals that attract conspecifics of the opposite sex and elicit a response leading to mating. Sex pheromones can be produced by either males or females to attract the other sex (Kelsey 1966; Chuman et al. 1987; Hardee et al. 1967). (2)

Aggregation pheromones are semiochemicals that attract individuals from both sexes, unlike sex pheromones which only attract the opposite sex. Aggregation pheromones are commonly used by insects to overcome Allee effects, predation, overwhelm host plants and for orientation to other individuals (Raffa et al. 1993; Rohlfs and Hoffmeister 2004;

Butler and Simpson 1967), which may result in mating. (3) Alarm pheromones are produced by organisms in response to perceived danger (Verheggen et al. 2010). Alarm pheromones are well documented in social insects, but several asocial insects also utilize alarm pheromones (Blum and Brand 1972; Nault et al. 1973). (4) Trail pheromones are 31 compounds that orient or lead individuals to a target location, commonly a food source or the colony. Trail pheromones are another example of pheromones that are common in social insects, especially ants (Morgan 2009; Traniello and Robson 1995). (5) Epideictic pheromones, commonly referred to as spacing pheromones or anti-aggregation pheromones, are pheromones utilized to prevent overcrowding, usually overcrowding of a food resource or available space. As Douglas fir beetles (Dendroctonus pseudotsugae

Hopkins) colonize a host tree, they release aggregation pheromones to attract more and more individuals; when the host tree has been fully utilized, a threshold is reached where both males and females produce epideictic compounds that will stop the aggregation phase (Wood 1982). Aphids are also known to utilize epideictic compounds to repel conspecifics from host plants already heavily infested by the species (Quiroz et al. 1997).

Semiochemicals produced by plants also play an important role in insect communication (Visser 1988). Host plant odors are critical for host plant location by phytophagous insects (Bruce and Pickett 2011). Insects do not necessarily respond to species-specific compounds produced by their host plant, but are thought to respond to ratio-specific odors (Knight and Light 2001; Krasnoff and Dussourd 1989; Baoyu et al.

2001). One class of compounds commonly used as semiochemical cues in phytophagous insects are green leaf volatiles. Green leaf volatiles are a six-to-nine carbon containing aldehydes, alcohols, and esters (Zainal and Ismail 2015). These compounds are important in plant defense and for priming plant defenses, and are commonly used by insects to orient towards their hosts (Zainal and Ismail 2015; Visser and Avé 1978; Dickens et al.

1992; Whitman and Eller 1992). 32

Historically, extremely large amounts of material were required for chemical ecologists to identify semiochemicals. To identify the sex pheromone produced by the male boll weevil, Tumlinson et al. (1969) required 67,000 males, 4.5 million mixed sex weevils, and 54.7 kg of weevil feces. Since then, rapid progress in the identification and collection of semiochemicals has been made, specifically gas chromatography coupled to electroantennography (GC-EAG), usually matched with gas chromatography coupled to a mass spectrometer (GC-MS).

In order to identify semiochemicals, the volatile compounds produced by the plants and insects must be collected. Descriptions of several systems for volatile collection have been published. Adsorbents used in volatile collection vary from glass wool, activated charcoal, and Porapak Q, among others (Baker et al. 1981; Heath and

Manukian 1992; Byrne et al. 1975). Volatiles collected in the adsorbent can then be used for identification and to determine the behavioral activity of the compound.

When volatiles pass over insect antennae, it causes firing of the neurons in the specialized sensillae. The connection of microelectrodes at the base of an insect’s head and to the distal end of the antenna makes it possible to record electrical depolarizations in the antenna. The detection and measurement of this depolarization is referred to as electroantennography (EAG) (Schneider 1957). This system can then be coupled with gas chromatography. Gas chromatography is a common type of chromatography that is used for separating and analyzing volatile compounds. Testing individual compounds present in the volatile sample can be achieved when EAG is coupled with gas chromatography.

The ability to couple GC-EAG resulted in an early powerful analytical tool for detecting 33 insect semiochemicals (Wadhams 1990). However, while GC-EAG allows for the detection of potential semiochemicals, it does not allow for any inference on the chemical structure or behavioral activity of the compound.

In order to identify semiochemicals, the structure of the compound must be determined. Gas chromatography coupled to a mass spectrometer (GC-MS) has become a critical tool for the identification of semiochemical components (Pickett 1990). The utilization of GC-MS couples the ability of gas chromatography to separate out individual compounds and the ionization of these compounds into fragments for mass spectrometry.

The identity of the ion fragments can be determined by mass-to-charge ratio. The precise structure of the original compound can then be determined from the array of ionized fragments. After the compound that is causing antennal depolarization is identified, either isolated or synthetic sourced compound should be deployed in behavioral assays to determine the potential role of the semiochemical (Fontan et al. 2002; Ward 1981).

Semiochemicals in the D. carinulata/Tamarix System:

Two male-produced aggregation pheromones have been identified for D. carinulata: (2E,4Z)-2,4-heptadienal and (2E,4Z)-2,4-heptadien-1-ol. These compounds are also present in Tamarix foliage and are released when beetle feeding injures the foliage (Cossé et al. 2005). A small amount of these pheromones is associated with female D. carinulata, but is thought to originate from ingestion of host foliage, and not through endogenous production (Cossé et al. 2005). Conversely, male beetles appear to control the rate and timing of emission of these compounds. Male D. carinulata emit 8 to 34

18 times the amount of (2E,4Z)-2,4-heptadienal and (2E,4Z)-2,4-heptadien-1-ol emitted by females, which suggests that the males use pheromone biosynthesis and do not depend on acquisition of compounds from foliage (Cossé et al. 2005). In many Chrysomelidae, male-produced aggregation pheromones are common (Cossé et al. 2002; Peng and Weiss

1992; Dickens et al. 2002; Zilkowski et al. 2006). Adult males held under a diapause inducing day length displayed reduced pheromone emission rates that were consistent with amounts released from feeding on the foliage (Bean et al. 2007a).

Green leaf volatiles (GLVs) are defensive and signaling compounds commonly produced in response to herbivory by plants, and can serve as signaling compounds between plants and insects (Watanabe 1958; Murray et al. 1972; Visser and Ave 1978).

These volatiles are commonly used by herbivores to locate host plants (Bruce et al. 2005;

Visser and Ave 1978). Research by Cossé et al. (2006) indicated that D. carinulata can detect GLVs released by Tamarix. Specifically, D. carinulata responds to: (Z)-3-hexenal,

(E)-2-hexenal, (Z)-3-hexen-1-ol, and (Z)-3-hexenyl acetate. Together, these four compounds facilitate aggregation of adult D. carinulata males and females.

Cossé et al. (2005) determined that (2E,4Z)-2,4-heptadien-1-ol was much more attractive than (2E,4Z)-2,4-heptadienal and was as attractive as a 1:1 blend of the two pheromones. The (2E,4Z)-2,4-heptadien-1-ol component captured numbers of insects equivalent to a paired combination of the two aggregation pheromone components in field trials using baited sticky traps (Cossé et al. 2005). The aggregation pheromone component (2E,4Z)-2,4-heptadien-1-ol attracted equivalent numbers of males and females, suggesting that it functions as an aggregation pheromone and not as a sex 35 pheromone. Cossé et al. (2006) demonstrated that capture of adults using (2E,4Z)-2,4- heptadien-1-ol was synergized by the four GLVs and captured the greatest number of adults. Synergism between attractive host plant volatiles and aggregation pheromones is well documented in the scientific literature (Dickens 1989; Light et al. 1993; Deglow and

Borden 1998).

Sticky traps baited with the four GLVs were deployed in defoliated or intact stands of Tamarix; male beetles constituted the majority of the adults captured, at 77 to

97% (Cossé et al. 2006). Traps were deployed in the afternoon and counted the following morning. The limited exposure of the lures to adult activity and skewed ratio of adults captured in this study suggests that males initiate dispersal flights in search of host plants. It has been hypothesized that a new plant colonized by males releases the four

GLVs in response to feeding injury, and as colonizing males begin to release the aggregation pheromone, more beetles are attracted from the surrounding area to further colonize the plant. This results in aggregation on the foliage with subsequent mating, egg laying, and defoliation (Cossé et al. 2006; Bean et al. 2013). At some point, either another potential semiochemical cue disrupts the aggregation, or eventually the adults disperse after the plant is defoliated (Peng and Weiss 1992; Rostás and Hilker 2002; Bean et al.

2013).

Use of Semiochemicals to Resolve Challenges and Conflicts with D. carinulata:

Shortcomings and concerns about D. carinulata as a biological control agent can potentially be addressed through developing the following: (1) a strategy to potentiate 36 populations, (2) a monitoring system that can detect very low density populations, (3) a strategy to retain populations in a target area to facilitate establishment, and (4) a strategy to disperse D. carinulata from areas in conflict with avian species. These strategies could be developed by combining tactics based on chemical ecology and integrated pest management (IPM). The use of semiochemicals to manipulate insect behavior to manage insects has become a very common practice (Gaston et al. 1967; Cardé and Minks 1995;

Hokkanen 1991). By combining the behaviorally active stimuli identified by chemical ecologists and the concepts of IPM, a strategy to manipulate insect behavior can be developed. Many tactics have already been developed, and include methods such as mass trapping, attract and kill, mate disruption, trap cropping and monitoring (Foster and

Harris 1997; Wall 1990). These methods are critically dependent on stimuli, usually semiochemicals, to either trigger or inhibit specific behavior (Borden 1989; Cook et al.

2006; Foster and Harris 1997; Vité and Baader 1990).

The strategies mentioned above are examples from agriculture or forestry and focus on manipulating the pest species to minimize the impact of herbivory by pests, which ultimately results in protection of the crop plants or trees. However, in classical weed biological control with insects, the goal is to conserve and propagate the insect to kill the plant. The question then arises, could the concepts developed to kill or manage insect pest numbers and protect plants be used to protect insects and kill plants? Could the challenges of efficacy and ecological conflicts in biological control with D. carinulata be resolved by implementation of common semiochemically-based pest management strategies? The goal of this dissertation is to determine if semiochemicals 37 can be used to advantageously manipulate the spatial distributions of weed biological control agents. I am proposing that the issues associated with biological control using D. carinulata can be addressed through the use of semiochemicals to behaviorally manipulate D. carinulata in the field.

Considering how field releases of D. carinulata are made, the deployment of the aggregation pheromone at the time of the release could result in greater retention of individuals on release sites, resulting in decreased Allee effects and increased population growth, and ultimately increasing the likelihood of establishment. It has been hypothesized that that successful establishment of new releases is being guided by naturally-occurring semiochemicals, specifically aggregation pheromones (Cossé et. al

2006; Hudgeons et al. 2007a). If there is an inadequate amount of pheromone being produced by the males to maintain aggregations at the initial release point, it is likely that the beetles will disperse. Following releases of D. elongata in Texas, it was observed that adults dispersed from the release location and formed an aggregation next to a nearby field cage housing conspecifics (Hudgeons et al. 2007a). The observed dispersal and subsequent reformation of the aggregation was believed to be based on attraction to chemical cues originating from conspecifics in the field cage. The beetles released in the presence of the field cage established, while larger releases of D. elongata made with no nearby cage failed (Hudgeons et al. 2007a). The openly released, uncaged population that successfully established was initiated with 200 individuals, a much smaller number than the releases of 5,000 individuals suggested in the original recommendation (Knutson and

Muegge 2008). Thus, by deploying the aggregation pheromone in the field, multiple 38 smaller releases of D. carinulata could be made with similar confidence of success as for larger releases. By maintaining aggregations of beetles when a smaller release is made, biological control practitioners can resolve the conflict between making a single large release which will likely minimize Allee effects, or making multiple smaller releases that are susceptible to Allee effects but decrease the risk of that a localized catastrophic event will result in extinction, as would be the case with a single large release (Memmott et al.

1996; Hopper and Roush 1993; Grevstad 1999).

The lack of significant impact or successful control of Tamarix by D. carinulata in some locations could be addressed by deployment of the aggregation pheromone or the behaviorally active plant volatiles, or both. Extremely small populations of Douglas-fir beetles can be aggregated using semiochemicals (Borden et al. 1983; Borden 1989). This allows for beetle damage to be concentrated on a relatively small number of trees

(Borden et al. 1983; Borden 1989). Martel et al (2005) determined that application of synthetic host volatiles to potato resulted in increased densities of the Colorado potato beetle on semiochemically treated plants, compared to an untreated control. The purpose for concentrating the Douglas-fir beetles and the Colorado potato beetle in aggregations was to minimize the area where damage occurred, and to direct the damaging population away from a specific area. The possibility of using these methods to concentrate populations of D. carinulata to maximize their impact on a target set of plants is very apparent. It is possible that if the population of D. carinulata at Lovell, WY, which had underwhelming results in 2004, was exposed to targeted placement of aggregation forming semiochemicals in the field, the established but low number of beetles could 39 have been concentrated to defoliating levels. These increased densities can also impact subsequent generations and population growth, since insects that have been aggregated will be more likely to find mates and reproduce. The purposeful manipulation of a low density, underperforming population of D. carinulata could diminish the lag time until impactful population increases occur, thereby facilitating rapid population growth.

Semiochemicals are commonly used to assess pest populations to inform management decisions (Elkinton and Cardé 1981; Howse et al. 1998; Minks and

Voerman 1973). Through the use of semiochemical treated traps, emerging, small populations that were previously below detectable thresholds could be confirmed as present (Wall 1989; Billings and Upton 2010; Borden 1989). Deployment of lures emitting the aggregation pheromone and host plant volatiles could greatly aid managers in detecting low density populations of D. carinulata, allowing them to make better informed management decisions.

One method that could be used to address the ongoing concern about conflict between endangered avian species and D. carinulata is to develop a tactic that could be used to repel D. carinulata from areas of concern. It is possible that a semiochemical eliciting an anti-aggregation or repellent response could be used to deter adult feeding and oviposition in areas where D. carinulata could negatively impact avian nesting habitat. Many examples exist where managers have successfully protected plants from insect pests through the deployment of anti-aggregation/repellent chemicals (Furniss et al. 1974; Furniss et al. 1977; Ross and Daterman 1995; Cook et al. 2006). While the

‘pull’ of the aggregation pheromone could focus the feeding of D. carinulata on Tamarix 40 stands not already being used for nesting sites, as local beetle populations increased, all

Tamarix in the area would be defoliated regardless of semiochemical treatment (Gaffke et al. 2018a).

An initial investigation for potential repellent/anti-aggregation semiochemicals could be made using larval D. carinulata. Larvae of several species in the leaf beetle family (Coleoptera: Chrysomelidae) are known to elicit intraspecific behavioral responses in conspecific adults (Schindek and Hilker 1996; Gross and Hilker 1994;

Hilker 1989). Semiochemicals produced by larval D. carinulata could be used to disperse or disrupt aggregation. These repellent semiochemicals could provide the critical

‘push’ function, protecting avian habitat. The investigation of an anti-aggregating compound for D. carinulata should be prioritized, in the interest of conserving Tamarix currently providing critical habitat to imperiled native organisms.

Spatial manipulation of D. carinulata populations could conceivably be accomplished utilizing aggregation and repellent semiochemicals. The application of aggregation causing semiochemicals could result in better monitoring techniques for D. carinulata, faster establishment of newly released agents, and potentiate both low density releases and established populations. The application of repellent semiochemicals could provide a strategy for land managers to disperse D. carinulata from areas of specific avian conservation interest. The use of these semiochemicals has the potential to resolve many existing challenges and conflicts associated with the release of D. carinulata.

References

Andersen, D.C. and Nelson, S.M. 2013. Floral ecology and insect visitation in riparian Tamarix sp. (saltcedar). Journal of Arid Environments 94: 105-112.

Anderson, B.W. and Ohmart, R.D. 1984. Vegetation management study for the enhancement of wildlife along the lower Colorado River. Final report to USA Bureau of Reclamation, Contract No. 7-07-30-V0009

APHIS 2010. USDA APHIS PPQ Moratorium for biological control of saltcedar. https://www.aphis.usda.gov/plant_health/ea/downloads/SaltcedarMemo-2010.pdf

Baker, T.C., Gaston, L.K., Pope, M.M., Kuenen, L.P.S. and Vetter, R.S. 1981. A high- efficiency collection device for quantifying sex pheromone volatilized from female glands and synthetic sources. Journal of Chemical Ecology, 7: 961-968.

Baoyu, H., Zhongning, Z. and Yuling, F. 2001. Electrophysiology and behavior feedback of diamondback moth, Plutella xylostella, to volatile secondary metabolites emitted by Chinese cabbage. Chinese Science Bulletin, 46: 2086-2088.

Barrows, C. 1998. The case for wholesale removal. Restoration and management notes 16: 130-134

Baum, B. 1966. Monographic revision of the genus Tamarix. Final research report for the USDA. Department of Botany, Hebrew University, Jerusalem, Israel.

Baum, B. 1978. The genus Tamarix. Academy of Science and Humanities, Jerusalem, Israel.

Bean, D., Dudley, T. and Hultine, K. 2013. Bring on the Beetles. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 377-403

Bean, D.W., Dalin, P. and Dudley, T.L. 2012. Evolution of critical day length for diapause induction enables range expansion of Diorhabda carinulata, a biological control agent against tamarisk (Tamarix spp.). Evolutionary Applications, 5: 511- 523.

Bean, D.W., Dudley, T.L. and Keller, J.C. 2007a. Seasonal timing of diapause induction limits the effective range of Diorhabda elongata deserticola (Coleoptera: 42

hrysomelidae) as a biological control agent for tamarisk (Tamarix spp.). Environmental Entomology, 36: 15-25.

Bean, D.W., Wang, T., Bartelt, R.J. and Zilkowski, B.W. 2007b. Diapause in the leaf beetle Diorhabda elongata (Coleoptera: Chrysomelidae), a biological control agent for tamarisk (Tamarix spp.). Environmental Entomology, 36: 531-540.

Berry, W.L. 1970. Characteristics of salts secreted by Tamarix aphylla. American Journal of Botany, 57: 1226-1230.

Blum, M.S. and Brand, J.M. 1972. Social insect pheromones: their chemistry and function. American Zoologist, 12: 553-576.

Boeer, W.J. and Schmidly, D.J. 1977. Terrestrial mammals of the riparian corridor in Big Bend National Park. In Importance, preservation and management of riparian habitat: a symposium. General Technical Report to the U.S. Department of Agriculture, RM-43: 212-217

Borden, J.H. 1989. Semiochemicals and bark beetle populations: exploitation of natural phenomena by pest management strategists. Ecography, 12: 501-510.

Borden, J.H., Chong, L.J. and Fuchs, M.C. 1983. Application of semiochemicals in post- logging manipulation of the mountain pine beetle, Dendroctonus ponderosae (Coleoptera: Scolytidae). Journal of Economic Entomology, 76: 1428-1432.

Borges, M., Jepson, P.C. and Howse, P.E. 1987. Long‐range mate location and close‐ range courtship behaviour of the Green Stink Bug, Nezara viridula and its mediation by sex pheromones. Entomologia Experimentalis et Applicata, 44: 205- 212.

Bowser, C.W. 1957. Introduction and spread of the undesirable tamarisks in the Pacific southwestern section of the United States and comments concerning the plants' influence upon the indigenous vegetation. American Geophysical Union Transaction, 38: 415-416.

Brock, J.H. 1994. Tamarix spp. (salt cedar), an invasive exotic woody plant in arid and semi-arid riparian habitats of western USA. Ecology and management of invasive riverside plants. Wiley, New York, NY: 27-44.

Brotherson, J.D., and Field, D. (1987). Tamarix: impacts of a successful weed. Rangelands, 9: 110-112. 43

Bruce, T.J. and Pickett, J.A. 2011. Perception of plant volatile blends by herbivorous insects–finding the right mix. Phytochemistry, 72: 1605-1611.

Bruce, T.J., Wadhams, L.J. and Woodcock, C.M. 2005. Insect host location: a volatile situation. Trends in Plant Science, 10: 269-274.

Busch, D.E. 1995. Effects of fire on southwestern riparian plant community structure. The Southwestern Naturalist 40: 259-267.

Busch, D.E. and Smith, S.D. 1993. Effects of fire on water and salinity relations of riparian woody taxa. Oecologia, 94: 186-194.

Butler, C.G. and Simpson, J. 1967, December. Pheromones of the queen honeybee (Apis mellifera L.) which enable her workers to follow her when swarming. Proceedings of the Royal Entomological Society of London, 42: 149-154

Byrne, K.J., Gore, W.E., Pearce, G.T. and Silverstein, R.M. 1975. Porapak-Q collection of airborne organic compounds serving as models for insect pheromones. Journal of Chemical Ecology, 1: 1-7.

Cardé, R.T. and Minks, A.K. 1995. Control of moth pests by mating disruption: successes and constraints. Annual Review of Entomology, 40: 559-585.

Carman, J.G. and Brotherson, J.D. 1982. Comparisons of sites infested and not infested with saltcedar (Tamarix pentandra) and Russian olive (Elaeagnus angustifolia). Weed Science, 30: 360-364.

Carreira, J.A. and Niell, F.X. 1992. Plant nutrient changes in a semi‐arid Mediterranean shrubland after fire. Journal of Vegetation Science, 3: 457-466.

Carruthers, R.I., Anderson, G., DeLoach, CJ., Knight, .B., Ge, S. and Pong, P. 2005 Monitoring saltcedar biological control impact. Proceedings Monitoring Science and Technology Symposium, US Forest Service, RMRS P-37CD, Fort Collins, Colorado.

Carruthers, R.I., DeLoach, C.J., Herr, J.C., Anderson, G.L. and Knutson, A.E. 2008. Saltcedar areawide pest management in the western USA. Areawide Pest Management Theory and Implementation. CAB International, Wallingford: 271- 299.

Center for Biological Diversity 2009. Lawsuit filed to save endangered songbird; southwestern willow flycatcher threatened by release of imported beetle. 44

http://www.biologicaldiversity.org/news/press_releases/2009/southwestern- willow-flycatcher-03-27-2009.html.

Chapman, R.F. 2003. Contact chemoreception in feeding by phytophagous insects. Annual Review of Entomology, 48: 455-484.

Chew, M.K. 2009. The monstering of tamarisk: how scientists made a plant into a problem. Journal of the History of Biology, 42: .231-266.

Chuman, T., Guss, P.L., Doolittle, R.E., McLaughlin, J.R., Krysan, J.L., Schalk, J.M. and Tumlinson, J.H. 1987. Identification of female-produced sex pheromone from banded cucumber beetle, Diabrotica balteata LeConte (Coleoptera: Chrysomelidae). Journal of Chemical Ecology, 13: 1601-1616.

Cleverly, J. 2013. Water use by Tamarix. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 85-98.

Cleverly, J.R., Smith, S.D., Sala, A. and Devitt, D.A. 1997. Invasive capacity of Tamarix ramosissima in a Mojave Desert floodplain: the role of drought. Oecologia, 111:12-18.

Conrad, B., Acharya, K., Dudley, T.L. and Bean, D.W. 2013. Impact of episodic herbivory by the tamarisk leaf beetle on leaf litter nitrogen and stem starch content: A short communication. Journal of Arid Environments, 94: 76-79.

Cook, S.M., Khan, Z.R. and Pickett, J.A. 2006. The use of push-pull strategies in integrated pest management. Annual Review of Entomology, 52: 375-400.

Cossé, A.A., Bartelt, R.J. and Zilkowski, B.W. 2002. Identification and electrophysiological activity of a novel hydroxy ketone emitted by male cereal leaf beetles. Journal of Natural Products, 65: 1156-1160.

Cossé, A.A., Bartelt, R.J., Zilkowski, B.W., Bean, D.W. and Andress, E.R. 2006. Behaviorally active green leaf volatiles for monitoring the leaf beetle, Diorhabda elongata, a biocontrol agent of saltcedar, Tamarix spp. Journal of chemical ecology, 32: 2695-2708.

Cossé, A.A., Bartelt, R.J., Zilkowski, B.W., Bean, D.W. and Petroski, R.J. 2005. The aggregation pheromone of Diorhabda elongata, a biological control agent of saltcedar (Tamarix spp.): identification of two behaviorally active components. Journal of chemical ecology, 31: 657-670. 45

Dahanukar, A., Hallem, E.A. and Carlson, J.R. 2005. Insect chemoreception. Current Opinion in Neurobiology, 15: 423-430.

Dahm, C.N., Cleverly, J.R., Allred Coonrod, J.E., Thibault, J.R., McDonnell, D.E. and Gilroy, D.J. 2002. Evapotranspiration at the land/water interface in a semi‐arid drainage basin. Freshwater Biology, 47: 831-843.

Dalin, P., O'Neal, M.J., Dudley, T. and Bean, D.W. 2009. Host plant quality of Tamarix ramosissima and T. parviflora for three sibling species of the biocontrol insect Diorhabda elongata (Coleoptera: Chrysomelidae).Environmental Entomology, 35: 1373-1378.

Deglow, E.K. and Borden, J.H. 1998. Green leaf volatiles disrupt and enhance response to aggregation pheromones by the ambrosia beetle, Gnathotrichus sulcatus (Coleoptera: Scolytidae). Canadian Journal of Forest Research, 28: 1697-1705.

DeLoach, C.J. 1991. Saltcedar, an exotic weed of western North American riparian areas: A review of its taxonomy, biology, harmful and beneficial values, and its potential for biological control. USDI Bureau of Reclamation, Lower Colorado Region, Final Report. Boulder City, Nevada, USA: 443

DeLoach, C.J. and Tracy, J.L. 1997. Effects of biological control of saltcedar (Tamarix ramosissima) on endangered species: Draft Biological Assessment, 17 October 1997. USDA/ARS, Temple, TX: 612

DeLoach, C.J., Carruthers, R.I., Dudley, T.L., Eberts, D., Kazmer, D.J., Knutson, A.E., Bean, D.W., Knight, J., Lewis, P.A., Milbrath, L.R. and Tracy, J.L. 2004, April. First results for control of saltcedar (Tamarix spp.) in the open field in the western United States. XI International Symposium on Biological Control of Weeds: 505- 519

DeLoach, C.J., Carruthers, R.I., Knutson, A.E., Moran, P.J., Ritzi, C.M., Dudley, T.L., Gaskin, J., Kazmer, D., Thompson, D.A., Bean, D. and Eberts, D. 2011. Twenty- five years of biological control of saltcedar (Tamarix: Tamaricaceae) in the Western US: Emphasis Texas-1986-2011. XIII International Symposium on Biological Control of Weeds. Kona, HI: 268-276

DeLoach, C.J., Carruthers, R.I., Lovich, J.E., Dudley, T.L. and Smith, S.D. 2000. Ecological interactions in the biological control of saltcedar (Tamarix spp.) in the United States: toward a new understanding. Proceedings of the X International Symposium on Biological Control of Weeds. Bozeman, Montana: 819-873.

DeLoach, C.J., Lewis, P.A., Herr, J.C., Carruthers, R.I., Tracy, J.L. and Johnson, J. 2003. Host specificity of the leaf beetle, Diorhabda elongata deserticola (Coleoptera: Chrysomelidae) from Asia, a biological control agent for saltcedars (Tamarix: Tamaricaceae) in the Western United States. Biological Control: 117-147.

DeLoach, C.J., Pitcairn, M.J. and Woods, D. 1996. Biological control of salt cedar in southern California. Proceedings of the saltcedar management workshop, Rancho Mirage, CA. California Exotic Pest Plants Council: 30-31.

Dethier, V.J. 1948. Chemical insect attractants and repellents. Soil Science, 65:1-307.

Di Tomaso, J.M. 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed Technology, 12: 326-336.

Dickens, J.C. 1989. Green leaf volatiles enhance aggregation pheromone of boll weevil, Anthonomus grandis. Entomologia Experimentalis et Applicata, 52: 191-203.

Dickens, J.C., Billings, R.F. and Payne, T.L. 1992. Green leaf volatiles interrupt aggregation pheromone response in bark beetles infesting southern pines. Experientia, 48: 523-524.

Dickens, J.C., Oliver, J.E., Hollister, B., Davis, J.C. and Klun, J.A. 2002. Breaking a paradigm: male-produced aggregation pheromone for the Colorado potato beetle. Journal of Experimental Biology, 205: 1925-1933.

Douglass, C.H., Nissen, S.J. and Hart, C.R. 2013. Tamarisk Management. Tamarix: a case study of ecological change in the American West. Oxford University Press, New York, NY:333-353.

Dreesen, D.R. and Wangen, L.E. 1981. Elemental composition of saltcedar (Tamarix chinensis) impacted by effluents from a coal-fired power plant. Journal of Environmental Quality, 10: 410-416.

Drus, G.M., Dudley, T.L., Brooks, M.L. and Matchett, J.R. 2013. The effect of leaf beetle herbivory on the fire behaviour of tamarisk (Tamarix ramosissima Lebed.). International Journal of Wildland Fire, 22: 446-458.

Dudley, T.L. and Bean, D.W., 2012. Tamarisk biocontrol, endangered species risk and resolution of conflict through riparian restoration. BioControl, 57: 331-347.

Dudley, T.L. and DeLoach, C.J. 2004. Saltcedar (Tamarix spp.), Endangered Species, and Biological Weed Control-Can They Mix? 1. Weed Technology, 18: 1542- 1551.

Dudley, T.L. and Kazmer, D.J. 2005. Field assessment of the risk posed by Diorhabda elongata, a biocontrol agent for control of saltcedar (Tamarix spp.), to a nontarget plant, Frankenia salina. Biological Control, 35: 265-275.

Dudley, T.L., Bean, D.W., Pattison, R.R. and Caires, A. 2012. Selectivity of a biological control agent, Diorhabda carinulata Desbrochers, 1870 (Coleoptera: Chrysomelidae) for host species within the genus Tamarix Linneaus, 1753. The Pan-Pacific Entomologist, 88: 319-341.

Dudley, T.L., Dalin, P., Bean, D.W. 2006. Status of the biological control of Tamarix spp. in California. Hoddle, M.S., Johnson, M.W. (Eds.), Proceedings of the V California Conference on Biological Control, 25–27 July 2006, Riverside, CA: 137–140.

Dudley, T.L., DeLoach, C.J., Levich, J.E. and Carruthers, R.I. 2000. Saltcedar invasion of western riparian areas: impacts and new prospects for control. McCabe, R.E., Loos, S.E. (Eds.), Transactions of the 65th North American Wildlife & Natural Resources Conference, 24–28 March, 2000, Chicago, IL. Wildlife Management Institute, Washington, DC: 345–381

Duncan, C.A., Jachetta, J.J., Brown, M.L., Carrithers, V.F., Clark, J.K., Di Tomaso, J.M., Lym, R.G., McDANIEL, K.C., Renz, M.J. and Rice, P.M. 2004. Assessing the economic, environmental, and societal losses from invasive plants on rangeland and wildlands. Weed Technology, 18: 1411-1416.

Duncan, K.W. and McDaniel, K.C. 1998. Saltcedar (Tamarix spp.) management with imazapyr. Weed Technology, 12: 337-344.

Elkinton, J.S. and Cardé, R.T. 1981. The use of pheromone traps to monitor distribution and population trends of the gypsy moth. Management of Insect Pests with semiochemicals. Springer, Boston, MA: 41-55.

Ellis, L.M. 2001. Short-term response of woody plants to fire in a Rio Grande riparian forest, Central New Mexico, USA. Biological Conservation, 97: 159-170.

Engel-Wilson, R.W. and Ohmart, R.D. 1978, December. Floral and attendant faunal changes on the lower Rio Grande between Fort Quitman and Presidio, Texas. 48

In Proceedings of the National Symposium for Protection and Management of Floodplain Wetlands: 139-147.

Everitt, B.L. 1980. Ecology of saltcedar—a plea for research. Environmental Geology, 3: 77-84.

Federal Register 2005. Designation of critical habitat for the Southwestern Willow Flycatcher (Empidonax traillii extimus). USDI-FWS, 50 CFR Part 17, RIN 1018– AT88. 60886 Federal Register: 201.

Fernandez, P. and Hilker, M. 2007. Host plant location by Chrysomelidae. Basic and Applied Ecology, 8: 97-116.

Fick, W.H. and Geyer, W.A. 2010. Cut-stump treatment of saltcedar (Tamarix ramosissima) on the Cimarron National Grasslands. Transactions of the Kansas Academy of Science, 113: 223-226.

Fontan, A., Audino, P.G., Martinez, A., Alzogaray, R.A., Zerba, E.N., Camps, F. and Cork, A. 2002. Attractant volatiles released by female and male Triatoma infestans (Hemiptera: Reduviidae), a vector of Chagas disease: chemical analysis and behavioral bioassay. Journal of Medical Entomology, 39: 191-197.

Frasier, G.W. and Johnsen Jr, T.N. 1991. Saltcedar (Tamarisk): classification, distribution, ecology and control. Noxious Range Weeds. Westview Press, Boulder CO: 377-393.

Friedman, J.M., Auble, G.T., Shafroth, P.B., Scott, M.L., Merigliano, M.F., Freehling, M.D. and Griffin, E.R. 2005. Dominance of non-native riparian trees in western USA. Biological Invasions, 7: 747-751.

Furniss, M.M., Daterman, G.E., Kline, L.N., McGregor, M.D., Trostle, G.C., Pettinger, L.F. and Rudinsky, J.A. 1974. Effectiveness of the Douglas-fir beetle antiaggregative pheromone methylcyclohexenone at three concentrations and spacings around felled host trees. The Canadian Entomologist, 106: 381-392.

Furniss, M.M., Young, J.W., McGregor, M.D., Livingston, R.L. and Hamel, D.R. 1977. Effectiveness of controlled-release formulations of MCH for preventing Douglas- fir beetle (Coleoptera: Scolytidae) infestation in felled trees. The Canadian Entomologist, 109: 1063-1069.

Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Grieco, P.A., Peterson, R.K.D., Weaver, D.K. 2018a. Semiochemicals to 49

enhance herbivory by Diorhabda carinulata aggregations in saltcedar (Tamarix spp.) infestations. Pests Management Science, doi: 10.1002/ps.4848

Gaskin, J.F. 2013. Genetics of Tamarix. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 21-29

Gaskin, J.F. and Schaal, B.A. 2002. Hybrid Tamarix widespread in US invasion and undetected in native Asian range. Proceedings of the National Academy of Sciences, 99: 11256- 11259.

Gaskin, J.F. and Shafroth, P.B. 2005. Hybridization of Tamarix ramosissima and T. chinensis (saltcedars) with T. aphylla (athel) (Tamaricaceae) in the southwestern USA determined from DNA sequence data. Madrono, 52: 1-10.

Gaskin, J.F., Birken, A.S. and Cooper, D.J. 2012. Levels of novel hybridization in the saltcedar invasion compared over seven decades. Biological Invasions, 14: 693- 699.

Gasry, H.L. 1963. Root distribution of five-stamen tamarisk, seepwillow and arrowweed. Forest Science, 9: 311-314.

Gaston, L.K, Shorey, H.H. and Saario, C.A. 1967. Insect population control by the use of sex pheromones to inhibit orientation between the sexes. Nature, 213: 1155-1155

Gerling, D. and Kugler, J. 1973. Evaluation of enemies of noxious plants in Israel as potential agents for the biological control of weeds. Final technical report to the U.S. department of Agriculture Research Service. Contract No. A10-ENT-36

Glenn, E.P. and Nagler, P.L. 2005. Comparative ecophysiology of Tamarix ramosissima and native trees in western US riparian zones. Journal of Arid Environments, 61: 419-446.

Glenn, E.P., Huete, A.R., Nagler, P.L., Hirschboeck, K.K. and Brown, P. 2007. Integrating remote sensing and ground methods to estimate evapotranspiration. Critical Reviews in Plant Sciences, 26: 139-168.

response to invasive Tamarix control in southwestern U.S. rivers: a collaborative study including 416 sites. Ecological Applications, doi:10.1002/eap.1566 González, E., Sher, A.A., Anderson, R.M., Bay, R.F., Bean, D.W., Bissonnete, G.J., Cooper, D.J., Dohrenwend, K., Eichhorst, K.D., El Waer, H. and Kennard, D.K. 2017b. Secondary invasions of noxious weeds associated with control of invasive Tamarix are frequent, idiosyncratic and persistent. Biological Conservation 213: 106-114.

Graf, W.L. 1978. Fluvial adjustments to the spread of tamarisk in the Colorado Plateau region. Geological Society of America Bulletin, 89: 1491-1501.

Grevstad, F.S. and Herzig, A.L. 1997. Quantifying the effects of distance and conspecifics on colonization: experiments and models using the loosestrife leaf beetle, Galerucella calmariensis. Oecologia, 110: 60-68.

Hardee, D.D., Mitchell, E.B. and Huddleston, P.M. 1967. Laboratory studies of sex attraction in the boll weevil. Journal of Economic Entomology, 60: 1221-1224.

Harris, P. and Zwölfer, H. 1968. Screening of phytophagous insects for biological control of weeds. The Canadian Entomologist, 100: 295-303.

Heath, R.R. and Manukian, A. 1992. Development and evaluation of systems to collect volatile semiochemicals from insects and plants using a charcoal-infused medium for air purification. Journal of Chemical Ecology, 18: 1209-1226.

Hem, J.D. 1967. Composition of Saline Residues on Leaves and Stems of Saltcedar (Tamarix pentandra Pallas). U.S. Government Printing Office, Washington D.C., USA: 9

Hokkanen, H.M. 1991. Trap cropping in pest management. Annual Review of Entomology, 36:119-138.

Horton, J.S. 1960. Use of a root plow in clearing tamarisk stands. United States Department of Agriculture, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 50

Horton, J.S. 1964. Notes on the Introduction of Deciduous Tamarisk. U.S. Department of Agriculture, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 16

Horton, J.S. and Campbell, C.J. 1974. Management of phreatophyte and riparian vegetation for maximum multiple use values. U.S. Department of Agriculture, 51

Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 17 Howse, P.E., Stevens, I.D.R. and Jones, C.O. 1998. Pest monitoring. Insect Pheromones andTheir Use in Pest Management. Springer Science & Business Media: 263-279

Hudgeons, J.L., Knutson, A.E., DeLoach, C.J., Heinz, K.M., McGinty, W.A. and Tracy, J.L. 2007a. Establishment and biological success of Diorhabda elongata elongata on invasive Tamarix in Texas. Southwestern Entomologist, 32: 157-168.

Hudgeons, J.L., Knutson, A.E., Heinz, K.M., DeLoach, C.J., Dudley, T.L., Pattison, R.R. and Kiniry, J.R. 2007b. Defoliation by introduced Diorhabda elongata leaf beetles (Coleoptera: Chrysomelidae) reduces carbohydrate reserves and regrowth of Tamarix (Tamaricaceae). Biological Control, 43: 213-221.

Hultine, K.R. and Bush, S.E. 2011. Ecohydrological consequences of non‐native riparian vegetation in the southwestern United States: A review from an ecophysiological perspective. Water Resources Research, 47: 1-13

Hultine, K.R., Dudley, T.L., Koepke, D.F., Bean, D.W., Glenn, E.P. and Lambert, A.M. 2015. Patterns of herbivory-induced mortality of a dominant non-native tree/shrub (Tamarix spp.) in a southwestern US watershed. Biological Invasions, 17: 1729- 1742.

Irvine, J.R. and West, N.E. 1979. Riparian tree species distribution and succession along the lower Escalante River, Utah. The Southwestern Naturalist, 24: 331-346.

Jacobs, J. and Sing, S 2007. Biology and management of saltcedar (Tamarix ramosissina, T. chinensis and T. ramosissima x T. chinensis hybrids). Invasive Species Technical Note MT-13. United States Department of Agriculture, Natural Resources Conservation Service.

Jarnevich, C.S., Evangelista, P. and Graham, J. 2013. Measuring Extent and Projections of Tamarix Distribution in North America. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 29-44.

Jermy, T. and Szentesi, A. 1978. The role of inhibitory stimuli in the choice of oviposition site by phytophagous insects. Entomologia Experimentalis et Applicata, 24: 458-471.

Johnson, S. 1987. Can saltcedar be controlled. Fremontia, 15: 19-20. 52

Kauffman, W. 2005. Program for Biological Control of Saltcedar (Tamarix spp) in Thirteen States. Environmental Assessment, June 2005. U.S. Department of Agriculture, Western Region, Fort Collins, CO, USA: 56

Kelsey, J.M. 1966. Costelytra zealandica (White) sex attraction. New Zealand Entomologist, 3: 57-59.

Kennard, D., Louden, N., Gemoets, D., Ortega, S., González, E., Bean, D., Cunningham, P., Johnson, T., Rosen, K. and Stahlke, A. 2016. Tamarix dieback and vegetation patterns following release of the northern tamarisk beetle (Diorhabda carinulata) in western Colorado. Biological Control, 101: 114-122.

Kennedy, T.A., Finlay, J.C. and Hobbie, S.E. 2005. Eradication of invasive Tamarix ramosissima along a desert stream increases native fish density. Ecological Applications, 15: 2072-2083.

Kerpez, T.A. and Smith, N.S. 1987. Saltcedar control for wildlife habitat improvement in the southwestern United States. U.S. Department of Agriculture, U.S. Fish and Wildlife Service Resource Publication, 169: 16.

Kleinkopf, G.E. and Wallace, A. 1974. Physiological basis for salt tolerance in Tamarix ramosissima. Plant Science Letters, 3: 157-163.

Knight, A.L. and Light, D.M. 2001. Attractants from Bartlett pear for codling moth, Cydia pomonella (L.), larvae. Naturwissenschaften, 88: 339-342.

Knutson, A. and Muegge, M. 2008. Biological Control of Saltcedar using a Leaf-Feeding Beetle in Texas; 2006–2008. USDA Natural Resources Conservation Innovation Grant Program. Texas AgriLife Extension Service, Dallas, Texas: 31.

Kovalev, O.V. 1995. Co-evolution of the tamarisks (Tamaricaceae) and pest arthropods (Insecta; Arachnida: Acarina), with special reference to biological control prospects. Proceedings of Zoological Institute, Russian Academy of Sciences, St. Petersburg, Russia, Vol. 259. Pensoft Publishers, Sofia, Bulgaria.

Krasnoff, S.B. and Dussourd, D.E. 1989. Dihydropyrrolizine attractants for arctiid moths that visit plants containing pyrrolizidine alkaloids. Journal of Chemical Ecology, 15: 47-60.

Ladenburger, C.G., Hild, A.L., Kazmer, D.J. and Munn, L.C. 2006. Soil salinity patterns in Tamarix invasions in the Bighorn Basin, Wyoming, USA. Journal of Arid Environments, 65: 111-128.

Lewis, P.A., DeLoach, C.J., Herr, J.C., Dudley, T.L. and Carruthers, R.I. 2003a. Assessment of risk to native Frankenia shrubs from an Asian leaf beetle, Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), introduced for biological control of saltcedars (Tamarix spp.) in the western United States. Biological Control, 27: 148-166.

Lewis, P.A., DeLoach, C.J., Knutson, A.E., Tracy, J.L. and Robbins, T.O. 2003b. Biology of Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), an Asian leaf beetle for biological control of saltcedars (Tamarix spp.) in the United States. Biological Control, 27: 101-116.

Liesner, D.R. 1971. Phytophagous insects of Tamarix spp. New Mexico. Master of Science thesis, New Mexico State University, Las Cruces, New Mexico.

Light, D.M., Flath, R.A., Buttery, R.G., Zalom, F.G., Rice, R.E., Dickens, J.C. and Jang, E.B. 1993. Host-plant green-leaf volatiles synergize the synthetic sex pheromones of the corn earworm and codling moth (Lepidoptera). Chemoecology, 4: 145-152.

Lindgren, C., Pearce, C. and Allison, K. 2010. The biology of invasive alien plants in Canada. 11. Tamarix ramosissima Ledeb., T. chinensis Lour. and hybrids. Canadian Journal of Plant Science, 90: 111-124. Lovich, J.E. and Bainbridge, D. 1999. Anthropogenic degradation of the southern California desert ecosystem and prospects for natural recovery and restoration. Environmental Management, 24: 309-326.

Lovich, J.E. and De Gouvenain, R.C. 1998. Saltcedar invasion in desert wetlands of the southwestern United States: ecological and political implications. Proceedings California Exotic Pest Council, 4: 45-55.

Martel, J.W., Alford, A.R. and Dickens, J.C. 2005. Synthetic host volatiles increase efficacy of trap cropping for management of Colorado potato beetle, Leptinotarsa decemlineata (Say). Agricultural and Forest Entomology, 7: 79-86.

McDaniel, K.C. and Taylor, J.P. 2003. Saltcedar recovery after herbicide-burn and mechanical clearing practices. Journal of Range Management, 56: 439-445.

McDonald, A.K. 2010. Hydrologic Impacts of Saltcedar Control Along a Regulated Dryland River. Doctoral dissertation, Texas A&M University. 54

McFadyen, R.E.C. 1998. Biological control of weeds. Annual Review of Entomology, 43: 369-393.

Merkel, D.L. and Hopkins, H.H. 1957. Life history of salt cedar (Tamarix gallica L.). Transactions of the Kansas Academy of Science, 60: 360-369.

Milbrath, L.R., DeLoach, C.J. and Tracy, J.L. 2007. Overwintering survival, phenology, voltinism, and reproduction among different populations of the leaf beetle Diorhabda elongata (Coleoptera: Chrysomelidae). Environmental Entomology, 36: 1356-1364.

Minks, A.K. and Voerman, S. 1973. Sex pheromones of the summer fruit tortrix moth, Adoxophyes orana: Trapping performance in the field. Entomologia Experimentalis et Applicata, 16: 541-549

Morgan, D.E. 2009. Trail pheromones of ants. Physiological Entomology, 34: 1-17.

Morris, C., Grossl, P.R. and Call, C.A. 2009. Elemental allelopathy: processes, progress, and pitfalls. Plant Ecology, 202: 1-11.

Murray, R.D.H., Martin, A. and Stride, G.O. 1972. Identification of the volatile phagostimulants in Solanum campylacanthum for Epilachna fulvosignata. Journal of Insect Physiology, 18: 2369-2373.

Nagler, P.L., Pearlstein, S., Glenn, E.P., Brown, T.B., Bateman, H.L., Bean, D.W. and Hultine, K.R. 2014. Rapid dispersal of saltcedar (Tamarix spp.) biocontrol beetles (Diorhabda carinulata) on a desert river detected by phenocams, MODIS imagery and ground observations. Remote Sensing of Environment, 104: 206- 219.

Nagler, P.L., Shafroth, P.B., LaBaugh, J.W., Snyder, K.A., Scott, R.L., Merritt, D.M. and Osterberg, J. 2009. The potential for water savings through the control of saltcedar and Russian olive. Saltcedar and Russian olive control demonstration act science assessment. Scientific Investigations Report, 5247: 35-47.

Natale, E., Zalba, S.M., Oggero, A. and Reinoso, H. 2010. Establishment of Tamarix ramosissima under different conditions of salinity and water availability: Implications for its management as an invasive species. Journal of Arid Environments, 74: 1399-1407.

Nault, L.R., Edwards, L.J. and Styer, W.E. 1973. Aphid alarm pheromones: secretion and reception. Environmental Entomology, 2:101-105.

Ohmart, R.D. and Anderson, B.W. 1982. North American desert riparian ecosystems. Reference handbook on the deserts of North America. Greenwood Press, Westport, CT: 433-479.

Owens, M.K. and Moore, G.W. 2007. Saltcedar water use: realistic and unrealistic expectations. Rangeland Ecology & Management, 60: 553-557.

Patten, D.T. 1998. Riparian ecosystems of semi-arid North America: Diversity and human impacts. Wetlands, 18: 498-512.

Pattison, R.R., D’Antonio, C.M., Dudley, T.L., Allander, K.K. and Rice, B. 2011. Early impacts of biological control on canopy cover and water use of the invasive saltcedar tree (Tamarix spp.) in western Nevada, USA. Oecologia, 165: 605-616.

Paxton, E.H., Theimer, T.C. and Sogge, M.K. 2011. Tamarisk biocontrol using tamarisk beetles: potential consequences for riparian birds in the southwestern United States. The Condor, 113: 255-265.

Pearce, C.M. and Smith, D.G. 2003. Saltcedar: distribution, abundance, and dispersal mechanisms, northern Montana, USA. Wetlands, 23: 215-228.

Pearce, C.M. and Smith, D.G. 2007. Invasive saltcedar (Tamarix): its spread from the American Southwest to the northern Great Plains. Physical Geography, 28: 507- 530.

Pemberton, R.W. 2000. Predictable risk to native plants in weed biological control. Oecologia, 125: 489-494.

Peng, C. and Weiss, M.J. 1992. Evidence of an aggregation pheromone in the flea beetle, Phyllotreta cruciferae (Goeze) (Coleoptera: Chrysomelidae). Journal of Chemical Ecology, 18: 875-884.

Petranka, J.W. and Holland, R. 1980. A quantitative analysis of bottomland communities in south-central Oklahoma. The Southwestern Naturalist, 25: 207-213.

Pickett, J.A. 1990. Gas chromatography-mass spectrometry in insect pheromone identification: three extreme case histories. Chromatography and Isolation of Insect Hormones and Pheromones. Plenum Press, New York, NY: 299-309 56

Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D., Sparks, R.E. and Stromberg, J.C. 1997. The natural flow regime. BioScience, 47: 769- 784.

Quigley, M.F. 2013. Botany and horticulture of tamarisk. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 319-329.

Quiroz, A., Pettersson, J., Pickett, J.A., Wadhams, L.J. and Niemeyer, H.M. 1997. Semiochemicals mediating spacing behavior of bird cherry-oat aphid, Rhopalosiphum padi feeding on cereals. Journal of Chemical Ecology, 23: 2599- 2607.

Raffa, K.F., Phillips, T.W. and Salom, S.M. 1993. Strategies and mechanisms of host colonization by bark beetles. Beetle-pathogen interactions in conifer forests. Academic Press, London: 102-28.

Reichenbacher, F.W. 1984. Ecology and evolution of southwestern riparian plant communities. Desert Plants, 6: 15-22.

Reynolds, A.M. and Rhodes, C.J. 2009. The Lévy flight paradigm: random search patterns and mechanisms. Ecology, 90: 877-887.

Robinson, T.W. 1952. Symposium on phreatophytes and their relation to water in western United States. Eos, Transactions American Geophysical Union, 33: 57-61.

Robinson, T.W. 1965. Introduction, spread, and areal extent of saltcedar (Tamarix) in the western states. Geological Survey Professional Paper, Contact No. 491-A.

Rohlfs, M. and Hoffmeister, T.S. 2004. Spatial aggregation across ephemeral resource patches in insect communities: an adaptive response to natural enemies? Oecologia, 140: 654-661.

Rood, S.B., Goater, L.A., Mahoney, J.M., Pearce, C.M. and Smith, D.G. 2007. Floods, fire, and ice: disturbance ecology of riparian cottonwoods. Botany, 85: 1019- 1032.

Ross, D.W. and Daterman, G.E. 1995. Efficacy of an antiaggregation pheromone for reducing Douglas-fir beetle, Dendroctonus pseudotsugae Hopkins (Coleoptera: Scolytidae), infestation in high risk stands. The Canadian Entomologist, 127: 805- 811. 57

Rostás, M. and Hilker, M. 2002. Feeding damage by larvae of the mustard leaf beetle deters conspecific females from oviposition and feeding. Entomologia Experimentalis et Applicata, 103: 267-277.

Sala, A., Woodruff, D.R. and Meinzer, F.C. 2012. Carbon dynamics in trees: feast or famine? Tree Physiology, 32: 764-775.

Salinas, M.J., Blanca, G. and Romero, A.T. 2000. Riparian vegetation and water chemistry in a basin under semiarid Mediterranean climate, Andarax River, Spain. Environmental Management, 26: 539-552.

Schneider, D. 1957. Electrophysiological investigation on the antennal receptors of the silk moth during chemical and mechanical stimulation. Experientia, 13: 89-91.

Shafroth, P.B., Friedman, J.M. and Ischinger, L.S. 1995. Effects of salinity on establishment of Populus fremontii (cottonwood) and Tamarix ramosissima (saltcedar) in southwestern United States. The Great Basin Naturalist, 55: 58-65.

Shorey, H.H. and Gaston, L.K. 1965. Sex pheromones of noctuid moths. VII. Quantitative aspects of the production and release of pheromone by females of Trichoplusia ni (Lepidoptera: Noctuidae). Annals of the Entomological Society of America, 58: 604-608.

Smith, L.M., Sprenger, M.D. and Taylor, J.P. 2002. Effects of discing saltcedar seedlings during riparian restoration efforts. The Southwestern Naturalist, 47: 598-601.

Smith, S.D., Devitt, D.A., Sala, A., Cleverly, J.R. and Busch, D.E. 1998. Water relations of riparian plants from warm desert regions. Wetlands, 18: 687-696.

Snyder, K.A., Uselman, S.M., Jones, T.J. and Duke, S. 2010. Ecophysiological responses of salt cedar (Tamarix spp. L.) to the northern tamarisk beetle (Diorhabda carinulata Desbrochers) in a controlled environment. Biological Invasions, 12: 3795-3808.

Sogge, M.K., Paxton, E.H. and Tudor, A. 2005. Saltcedar and Southwestern Willow Flycatchers: lessons from long-term studies in central Arizona. In Aguirre-Bravo, Celedonio, et. al. eds, Monitoring Science and Technology Symposium: Unifying Knowledge for Sustainability in the Western Hemisphere; 2004 September 20-24; Denver, CO. Proceedings RMRS-P-37-CD. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. CD-ROM.

Sogge, M.K., Paxton, E.H. and Van Riper III, C. 2013. Tamarisk in riparian woodlands: a bird’s eye view. Tamarix: passenger vs. driver of ecological change. Oxford University Press, UK: 189-206.

Sogge, M.K., Sferra, S.J. and Paxton, E.H. 2008. Tamarix as habitat for birds: implications for riparian restoration in the southwestern United States. Restoration Ecology, 16: 146-154.

Stebbins, G.L. 1969. The significance of hybridization for plant taxonomy and evolution. Taxon, 18: 26-35.

Stevens, L.E. 1989. The status of ecological research on tamarisk (Tamaricaceae: Tamarix ramosissima) in Arizona. In Kunzmann, M.R., R.R. Johnson, and P.S. Bennet, P. Tamarisk control in southwestern United States: Proceedings of the Tamarisk Conference, University of Arizona, 2-3 September 1987. Special Report, 9: 99-105.

Strudley, S. and Dalin, P. 2013. Tamarix as invertebrate habitat. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 201-224.

Suckling, K., Hogan, D., Silver, R.D. and Guardians, F. 1992. Petition to list the southwest willow flycatcher Empidonax traillii extimus as a federally endangered species. Letter to the Secretary of the Interior. January 25, 1992. Luna, New Mexico.

Thompson, A.C. and Davis, F.M. 1981. The effect of temperature on the rate of metabolism of lipids and glycogen in diapausing southwestern corn borer, Diatraea grandiosella. Comparative Biochemistry and Physiology Part A: Physiology, 70: 555-558.

Thomson, W.W., Berry, W.L. and Liu, L.L. 1969. Localization and secretion of salt by the salt glands of Tamarix aphylla. Proceedings of the National Academy of Sciences, 63: 310-317.

Tracy, J.L. and Robbins, T.O. 2009. Taxonomic revision and biogeography of the Tamarix-feeding Diorhabda elongata (Brullé, 1832) species group (Coleoptera: Chrysomelidae: Galerucinae: Galerucini) and analysis of their potential in biological control of Tamarisk. Zootaxa 2101: 1-152.

Traniello, J.F. and Robson, S.K. 1995. Trail and territorial communication in social insects. Chemical Ecology of Insects. Springer US: 241-286

Tumlinson, J.H., Hardee, D.D., Gueldner, R.C., Thompson, A.C., Hedin, P.A. and Minyard, J.P. 1969. Sex pheromones produced by male boll weevil: isolation, identification, and synthesis. Science, 166: 1010-1012.

US Fish and Wildlife Service, 1995. Final rule determining endangered status for the southwestern willow flycatcher. Federal register, 60: 10694-10715.

U.S. Department of Agriculture Animal and Plant Health Inspection Service 1999. Field Release of a Nonindigenous Leaf Beetle, Diorhabda elongata (Coleoptera: Chrysomelidae), for Biological Control of Deciduous Saltcedar, Tamarix ramosissima and T. parviflora (Tamaraceae). Environmental Assessment, 1-31.

Verheggen, F.J., Haubruge, E. and Mescher, M.C. 2010. Chapter nine-alarm pheromones— Chemical signaling in response to danger. Vitamins & Hormones, 83: 215-239.

Visser, J.H. 1988. Host-plant finding by insects: orientation, sensory input and search patterns. Journal of Insect Physiology, 34: 259-268.

Visser, J.H. and Avé, D.A. 1978. General green leaf volatiles in the olfactory orientation of the Colorado beetle, Leptinotarsa decemlineata. Entomologia Experimentalis et Applicata, 24: 738-749.

Vité, J.P. and Baader, E. 1990. Present and future use of semiochemicals in pest management of bark beetles. Journal of Chemical Ecology, 16:3031-3041

Wadhams, L.J. 1990. The use of coupled gas chromatography: electrophysiological techniques in the identification of insect pheromones. Chromatography and Isolation of Insect Hormones and Pheromones. Plenum Press, New York, NY: 289-298.

Waisel, Y. 1960. Ecological studies on Tamarix aphylla (L.) Karst. Plant and Soil, 13: 356-364.

Wall, C. 1989. Monitoring and spray timing. Insect Pheromones in Plant Protection. John Wiley & Sons Limited: 39-66.

Wall, C. and Perry, J.N. 1987. Range of action of moth sex‐attractant sources. Entomologia Experimentalis et Applicata, 44: 5-14. 60

Wapshere, A.J. 1974. A strategy for evaluating the safety of organisms for biological weed control. Annals of Applied Biology, 77: 201-211.

Ward, J.P. 1981. A comparison of the behavioural responses of the haematophagous bug, Triatoma infestans, to synthetic homologues of two naturally occurring chemicals (n‐and iso‐butyric acid). Physiological Entomology, 6: 325-329.

Warren, D.K. and Turner, R.M. 1975. Saltcedar (Tamarix chinensis) seed production, seedling establishment, and response to inundation. Journal of the Arizona Academy of Science, 10: 135-144.

Watanabe, T. 1958. Substances in mulberry leaves which attract silkworm larvae (Bombyx mori). Nature, 182: 325 – 326.

Weeks, E.P., Weaver, H.L., Campbell, G.S. and Tanner, B.D. 1987. Water use by saltcedar and by replacement vegetation in the Pecos River floodplain between Acme and Artesia, New Mexico. US Geological Survey Professional Paper 491- G: 33.

Whitcraft, C.R., Talley, D.M., Crooks, J.A., Boland, J. and Gaskin, J. 2007. Invasion of tamarisk (Tamarix spp.) in a southern California salt marsh. Biological Invasions, 9: 875-879.

Whitman, D.W. and Eller, F.J. 1992. Orientation of Microplitis croceipes (Hymenoptera: Braconidae) to green leaf volatiles: Dose-response curves. Journal of Chemical Ecology, 18: 1743-1753.

Wilkinson, R.E. 1966. Seasonal development of anatomical structures of saltcedar foliage. Botanical Gazette, 127: 231-234.

Williams, W.I., Friedman, J.M., Gaskin, J.F. and Norton, A.P. 2014. Hybridization of an invasive shrub affects tolerance and resistance to defoliation by a biological control agent. Evolutionary Applications, 7: 381-393.

Wood, D.L. 1982. The role of pheromones, kairomones, and allomones in the host selection and colonization behavior of bark beetles. Annual Review of Entomology, 27: 411-446.

Yoccoz, N.G., Nichols, J.D. and Boulinier, T. 2001. Monitoring of biological diversity in space and time. Trends in Ecology & Evolution, 16:446-453.

Zainal, Z. and Ismail, I. 2015. Green leaf volatiles. Plant Biotechnology Journal, 13: 727- 739.

Zavaleta, E. 2000. The economic value of controlling an invasive shrub. AMBIO: a Journal of the Human Environment, 29: 462-467.

CHAPTER TWO

SEMIOCHEMICALS TO ENHANCE HERBIVORY BY DIORHABDA

CARINULATA AGGREGATIONS IN SALTCEDAR (TAMARIX SPP.)

INFESTATIONS

Contribution of Authors and Co-Authors

Manuscript in Chapter 2

Author: Alexander M. Gaffke

Contributions: Designed and implemented experiments, collected and processed and analyzed data, and prepared manuscript for publication

Co-Author: Sharlene E. Sing

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Tom L. Dudley

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Daniel W. Bean

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Justin A. Russak

Contributions: Developed methodology for chemical synthesis

Co-Author: Agenor Mafra-Neto

Contribution of Authors and Co-Authors Continued

Contributions: Provided critical input on the formulation of the semiochemcials for field deployement

Co-Author: Paul A. Grieco

Contributions: Developed methodology for chemical synthesis

Co-Author: Robert K. D. Peterson

Contributions: Provided input on result interpretation and manuscript preparation

Co-Author: David K. Weaver

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Manuscript Information Page

Alexander M. Gaffke, Sharlene E. Sing, Tom L. Dudley, Daniel W. Bean, Justin A. Russak, Agenor Mafra-Neto, Paul A. Grieco, Robert K. D. Peterson, David K. Weaver

Pest Management Science

Status of Manuscript: ____ Prepared for submission to a peer-reviewed journal ____ Officially submitted to a peer-review journal ____ Accepted by a peer-reviewed journal __x_ Published in a peer-reviewed journal

Published by John Wiley & Sons, Inc. Submitted September 2017 Accepted December 2017 Published January 2017

CHAPTER THREE

FIELD DEMONSTRATION OF A SEMIOCHEMICAL TREATMENT

THAT ENHANCES DIORHABDA CARINULATA BIOLOGICAL

CONTROL OF TAMARIX SPP.

Contribution of Authors and Co-Authors

Manuscript in Chapter 3

Author: Alexander M. Gaffke

Contributions: Designed and implemented experiments, collected, processed and analyzed data, and prepared manuscript for publication

Co-Author: Sharlene E. Sing

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Tom L. Dudley

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Daniel W. Bean

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Justin A. Russak

Contributions: Developed methodology for chemical synthesis

Co-Author: Agenor Mafra-Neto

Contribution of Authors and Co-Authors Continued

Contributions: Provided critical input on the formulation of the semiochemcials for field deployment

Co-Author: Robert K. D. Peterson

Contributions: Provided input on result interpretation and manuscript preparation

Co-Author: David K. Weaver

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Manuscript Information Page

Alexander M. Gaffke, Sharlene E. Sing, Tom L. Dudley, Daniel W. Bean, Justin A. Russak, Agenor Mafra-Neto, David K. Weaver

BioControl

Status of Manuscript: __X_ Prepared for submission to a peer-reviewed journal ____ Officially submitted to a peer-review journal ____ Accepted by a peer-reviewed journal ____ Published in a peer-reviewed journal

FIELD DEMONSTRATION OF A SEMIOCHEMICAL TREATMENT THAT

ENHANCES DIORHABDA CARINULATA BIOLOGICAL CONTROL OF TAMARIX

SPP.

Alexander M. Gaffkea, Sharlene E. Singb, Tom L. Dudleyc, Daniel W. Beand, Justin A. Russake, Agenor Mafra-Netof, Robert K. D. Petersona, David K. Weavera*

aDepartment of Land Resources and Environmental Sciences, Montana State University,

Bozeman MT, 59717, USA bUSDA Forest Service, Rocky Mountain Research Station, Bozeman MT, 59717, USA cMarine Science Institute, University of California, Santa Barbara CA 93106, USA dColorado Department of Agriculture, Palisade Insectary, Palisade CO, 81526, USA eDepartment of Chemistry and Biochemistry, University of California Santa Barbara,

Santa Barbara CA, 93106, USA fISCA Technologies, Incorporated, Riverside CA 92507, USA

*Correspondence to: David K. Weaver, Montana State University, P.O. Box 173120,

Bozeman, MT, 59717. Email: [email protected] 79

Abstract

The northern tamarisk leaf beetle Diorhabda carinulata (Desbrochers) was approved for release in the United States for classical biological control of a complex of invasive saltcedar species and their hybrids (Tamarix spp.). An aggregation pheromone that is used by D. carinulata to locate conspecifics is fundamental to colonization and reproductive success. A specialized matrix formulated for controlled release of this aggregation pheromone was developed as a lure to manipulate adult densities in the field. One application of the lure at onset of adult emergence for each generation provided long term attraction and retention of D. carinulata adults on treated Tamarix spp. plants. Treated plants exhibited greater levels of defoliation, dieback and canopy reduction. Application of a single, well-timed aggregation pheromone treatment per generation increased the efficacy of this classical weed biological control agent.

Key words: Aggregation pheromone, herbivory, attractant, invasive weed, Coleoptera,

Chrysomelidae

Introduction

Tamarix species and their hybrids (hereafter referred to as Tamarix) are invasive

Eurasian woody trees or shrubs increasingly present and dominant in riparian areas of the western and southwestern United States (Kovalev 1995, Robinson 1965, Gaskin et al.

2002, Friedman et al. 2005). The significant ecological and economic consequences of

Tamarix invasion are reflected in its legal status as a state or federal level noxious weed in 11 U.S. states (USDA, NRCS 2018). Outside of the United States, Tamarix is listed as one of the world’s 100 worst invasive alien species by the Global Invasive species

Database (Global Invasive Species Database 2018).

In the western United States, Tamarix outcompetes native riparian vegetation and can become the most locally abundant woody plant species (Friedman et al. 2005).

Tamarix dominance in western U.S. riparian ecosystems is correlated with habitat alteration, water resource depletion, flood damage, and increased soil salinity and fire potential (Horton and Campbell 1974, Hultine and Bush 2011, Shafroth et al. 1995,

Busch 1995). The disruption of ecological processes caused by Tamarix dominance in ecosystems has provided the impetus for massive control programs for Tamarix in many western U.S. states (Robinson 1952, Douglass et al. 2013). Along with management strategies such as flooding, prescribed burns, mechanical removal, and herbicide treatments, a classical biological control program was developed (Kauffman 2005). This classical biological control program resulted in the approval and release of the northern tamarisk beetle Diorhabda carinulata Desbrochers (formerly D. elongata deserticola

Brullé) (Coleoptera: Chrysomelidae) (USDA APHIS 1999). 81

In North America, D. carinulata populations are multivoltine, with 1-3 generations per year (Bean et al. 2007a, Lewis et al. 2003b). Diorhabda carinulata overwinters in the adult stage and emerges in the spring, coinciding with Tamarix bud break (Lewis et al. 2003b). After spring emergence, adults feed and mate several times before the females begin depositing clutches of eggs onto the Tamarix foliage (Bean et al.

2007a). Once the eggs hatch, the larvae feed within the Tamarix canopy and drop into the litter to form pupal cells (Lewis et al. 2003b). After metamorphosis, the adults emerge from the litter and re-enter the canopy of the Tamarix plants. When the adults emerge from the litter, they readily disperse both short and long distances (Bean et al. 2013).

Along with the dispersal behavior, the adults aggregate on plants (Cossé et al. 2005). In response to shorter day length at the approach of fall, D. carinulata adults enter a reproductive diapause. Once the adults enter reproductive diapause dispersal behavior ceases, metabolic reserves are accumulated, and they burrow into the litter to overwinter

(Bean et al. 2007b).

All life stages of D. carinulata feed exclusively on the foliage of Tamarix.

Feeding by D. carinulata abrades the Tamarix cuticle; widespread disruption of the cuticle by high densities of aggregated beetles results in significant desiccation of the foliage and may lead to partial or complete defoliation of Tamarix plants (Dudley and

Kazmer 2005). Defoliation by D. carinulata can dramatically increase the level of physiological stress experienced by affected plants, reducing their ability to store carbon, and diminishing leaf production and root mass in the years following a defoliation event

(Snyder et al. 2010, Hudgeons et al. 2007b, Hultine et al. 2010). Multiple years of intense 82 herbivory by D. carinulata can result in Tamarix mortality (Hudgeons et al. 2007b,

Dudley and Bean 2012).

Aggregations of D. carinulata occur in response to emission of an aggregation pheromone by reproductive males (Cossé 2005). This pheromone is not produced by D. carinulata in reproductive diapause (Bean et al. 2007b). Biosynthesis and conspecific response to female-emitted sex pheromones and male-emitted aggregation pheromones has been broadly confirmed within the Chrysomelidae (Sutton et al. 2017), including other weed biological control agents such as the galerucine Galerucella leaf beetles and alticine Longitarsus and Aphthona flea beetles (Bartelt et al. 2006; Zhang and McEvoy

1994; Bartelt et al. 2001). Observational studies similarly suggest a pheromonal basis for

Longitarsus jacobaeae (Zhang and McEvoy 1994) and Aphthona nigriscutis (Tansey et al. 2005) behavior and distribution. In addition to A. nigriscutis, male produced chemicals have been identified for A. flava, A. czwalinae, and A. cyparissiae (Bartelt et al. 2001).

Pheromone identification and function has been fully confirmed for Galerucella calmariensis and G. pusilla (Bartelt et al. 2006; Hambäck 2010; Fors et al. 2015) and

Diorhabda carinulata (Cossé et al. 2005, 2006, Gaffke et al. 2018a). Manipulation of aggregation pheromones has been hypothesized as a strategy for enhancing the efficacy of chrysomelid biological control agents (Landholt 1997; Dickens et al. 2002; Beran et al.

2016; Wood et al. 2007).

Techniques to manipulate densities of D. carinulata in the field have recently been investigated, specifically to increase the biocontrol efficacy of sparse beetle populations (Gaffke et al. 2018a). These manipulations involved the use of the well- 83 documented D. carinulata aggregation pheromone (2E,4Z)-2,4-heptadien-1-ol (Cossé et al. 2005, 2006). Treatments incorporating semiochemicals applied through a specialized pheromone lure application technology (SPLAT®: ISCA Technologies, Riverside, CA,

USA) made it possible to create or enhance aggregations of adults and larvae. Weekly application of 1-g doses of SPLAT impregnated with aggregation-causing semiochemicals resulted in persistent and higher densities of D. carinulata on treated plants. These increased densities resulted in intensified herbivory with increased rates of defoliation, plant dieback, and reduction in live canopy volume. These are desirable outcomes for a classical weed biological control program. Along with the intensified impact on Tamarix, the ability to aggregate populations of a biological control agent in the field is also useful to facilitate monitoring by biological control practitioners (Cossé et al. 2005, 2006, Yoccoz et al. 2001).

In a recent study, the release of the aggregation pheromone attained from 1-g doses of formulated SPLAT maintained aggregations of D. carinulata for only 7-10 days

(Gaffke et al. 2018a). The limited duration of lure activity was due to rapid volatilization of the aggregation pheromone from the small volume of the dollops used in the 1-g application. The limited duration of bioactivity of the 1-g treatment limits the practical utility of this technology. Treatments based on a 1-g lure required weekly applications throughout the summer to maintain consistent aggregation of D. carinulata (Gaffke et al.

2018a). The activity of SPLAT-based pheromone treatments can be extended by adjusting the volume applied (Mafra-Neto et al. 2013). An increased application rate of pheromone impregnated SPLAT was therefore investigated using larger dollops. The 84 objective of this study was to determine if three applications of 4-g doses of SPLAT targeting discrete adult emergence in this multivoltine beetle could result in season-long manipulation of D. carinulata densities. The rationale for evaluating targeted application of larger volumes of SPLAT-based semiochemical lures was to determine if longer term aggregations could be maintained, therefore reducing the time commitment required for practical deployment of this technique by land managers.

Material and Methods

Field Experiment Location and Insect Source

Tamarix stands along the Bighorn River near Lovell, WY were used as the study area during the summer of 2014, 2015, and 2016. Locations were selected based on existing infestations of saltcedar and the presence of D. carinulata populations. The D. carinulata populations present in the study area are bivoltine and are the descendants of individuals originally collected near the town of Fukang, Xinjiang China (DeLoach et al. 2003,

Lewis et al. 2003b).

Lure and Treatments

Two types of lures were prepared for field testing using ISCA® Technologies wax based formula SPLAT, a commercially available semiochemical release product.

The first type of lure, hereafter referred to as PH, consisted of SPLAT impregnated with

(2E,4Z)-2,4-heptadien-1-ol, an aggregation pheromone produced by D. carinulata (Cossé et. al 2005). The pheromone was synthesized according to Petroski (2003). The second type of lure, hereafter referred to as BL, consisted of the base formulation of SPLAT with 85 no active ingredient and functioned as a treatment control. Active ingredient of the semiochemical in each lure was 2.17% (2E,4Z)-2,4-heptadien-1-ol in PH and 0% in BL.

To allow for a single application of the SPLAT lures per generation, 4-g doses of the

SPLAT, referred to as dollops, were used. Treatments were applied to coincide with the emergence of 1) overwintered reproductive adults (P generation); 2) the first generation of adults (F1 generation), which were reproductive; and 3) the second generation of adults (F2 generation), which are non-reproductive and enter overwintering diapause. In

2014, the first application of SPLAT was June 16 (week 1), the second application of

SPLAT was July 16 (week 5), and the final application of SPLAT was August 18 (week

10). In 2015, the first application of SPLAT was May 26 (week 1), the second application of SPLAT was July 13 (week 8), and the final application of SPLAT was August 18

(week 13).

Release Rate Analysis

To determine the release rate of the pheromone from 4-g dollops, individual dollops were applied to cattle ear tags and exposed to outdoor conditions at the Bozeman

(Montana) Forestry Sciences Laboratory (USDA Forest Service, Rocky Mountain

Research Station). Cattle ear tags were used as the carrier for dollops because of their durability, portability, and general ease of handling. Volatile collections from these field aged dollops were made at 1, 3, 5, 10, 15, 25, and 31 days to characterize pheromone release rates. Dollops were temporarily transported from the outdoors to the laboratory for volatile collection, but were otherwise exposed to outdoor conditions. Collections were made in glass volatile collection chambers at varying durations depending on the 86 length of field exposure (15 min for 1, 3, 5, and 10 d; 30 min at 15 d; and 1 hr at 25 and

35 d). Volatile collection traps containing 30 mg of super-Q (Alltech Associates, Inc.,

Deerfield, IL, USA) adsorbent were fixed in place at the apical opening of the volatile collection chamber where purified air was delivered at a rate of 100 mlmin-1. Collected volatiles were eluted from the traps into vials using methylene chloride (200 µl) and the samples were spiked with 10 µl of a 0.84 ngµl-1 solution of 1-octanol in methylene chloride as an internal standard. Volatiles were analyzed using a gas chromatograph

(Agilent 6890; Agilent Technologies, Santa Clara, CA) coupled to a mass selective detector (MSD, Agilent 5973 instrument). Quantification was made relative to the internal standard.

Field Study

The existing arrays of Tamarix plants were randomly assigned treatments at two study sites located in a greasewood floodplain of the Bighorn River; characterized as having average Tamarix densities of 0.65 plantsm-2 (Landenburge et al. 2006). The study locations were approximately 1,000-m apart.

Twelve replicates of each treatment (PH and BL) were deployed on the study sites.

Dollops of SPLAT were applied to cattle ear tags using an 80-ml syringe. Application of

5 ml of formulated and control SPLAT yielded 4-g dollops. A wire mesh cage was fixed around each treated cattle ear tag to minimize accessibility and exposure of dollops to animals (Gaffke et al. 2018a). Treated ear tags were attached with a loop of wire to the branches of treatment plants. Treated plants were approximately 20 m apart. This distance minimized possible overlap of the volatile treatment (Cossé et al. 2005, 2006). 87

Sampling of Tamarix plants receiving 4-g dollops of SPLAT, with the addition of the aggregation pheromone (PH), or without added semiochemicals (BL), began in June with the emergence of the overwintered adults, and ended in September at the onset of overwintering diapause in the second generation adults. Year-to-year climatic variability influencing D. carinulata phenology resulted in sampling seasons that spanned 15 weeks in 2014 and 19 weeks in 2015. Treated Tamarix plants were sampled weekly using a heavy duty canvas sweep net that had an opening diameter of 30.5 cm and a depth of 61 cm. Treated plants were swept three times with a 1.25 m upward arc (Jamison and Lanci

2012). After sweeping, the captured D. carinulata adults and larvae were counted and returned to the tree where they originated. Care was taken to minimize capture-related damage to insects or Tamarix. Data on bivoltine D. carinulata adults were segregated for analysis according to either reproductive or non-reproductive status. Data on reproductive adults were therefore comprised of counts for P and F1 adults, while data on non- reproductive adults consisted only of sweep counts for F2 adults. Determination of reproductive status was based on the condition of the ovaries, ovarioles, and accessory glands from dissected adults (Bean et al. 2007b). Adults used for dissections were collected adjacent to the research locations. In 2014 and 2015, the emergence of non- reproductive F2 adults corresponded with week 12 and 14 of monitoring, respectively.

Damage due to D. carinulata was visually rated as the percentage of Tamarix foliar tissue damaged, and was measured in increments of 5% (Tamarisk Coalition 2017).

Feeding by adults or larvae results in foliage turning brown, withering, and dying. Plants that were completely brown are considered to be 100% damaged, even if the withered 88 foliage is still attached to the plant. Tamarix can recover from defoliation by D. carinulata and can even regrow foliage within the same season as the defoliation event. If regrowth of the Tamarix plants was observed, the percentage regrowth was visually estimated and the percentage was added as non-damaged green foliage.

Percentage dieback and canopy volume were determined once during the field seasons of 2015, and 2016 for each treated plant. Dieback was recorded as the percentage of wood that contained live foliage the previous year that did not produce foliage in the current year. Evaluation of dieback was made well after Tamarix bud burst to allow enough time for new growth to occur. Branch color, flexibility and the presence or absence of foliage were used to determine if the branch was dead. Canopy volume was recorded in the fall, to account for possible regrowth and recovery of the Tamarix plants.

Two perpendicular measurements were taken at the plant’s widest point to estimate canopy width, and one measurement was taken at the tallest point of live foliage to estimate canopy height.

Statistical Analysis

Data from the two research areas was pooled and for analysis. A repeated measure

ANOVA, using within-year sampling date as the repeated measure, and SPLAT application as the treatment factor, was used to detect differences in adult D. carinulata capture densities. Data were ln(x+1) transformed to better meet the assumptions of normality. Damage rating was also assessed using a repeated measures ANOVA, with sampling day as the repeated measure and SPLAT application as the treatment factor.

Potential changes in canopy volume were evaluated using a repeated measure ANOVA 89 with year as the repeated measure and SPLAT application as the treatment factor. Canopy volume was ln(x+1) transformed to better match the assumptions of normality. Repeated measure ANOVAs used type III error and Satterthwaite approximation for degrees of freedom. Percentage dieback was analyzed using a one-way ANOVA. ANOVAs were conducted using the lmer function in the lmerTest package of R software version 3.1.2

(Kuznetsova et al. 2015).

Results

Release Rate Analysis

Release rates of the D. carinulata aggregation pheromone (2E,4Z)-2,4-heptadien-1-ol from 4-g dollops of field-aged SPLAT over a 31-day period are displayed in Figure 1.

The release rate model y=19384x-2.34, yielded an R2 of 0.93 and a decay constant of -2.34.

On day one, mean D. carinulata aggregation pheromone emitted from treated dollops was 17514 nghr-1. Pheromone emissions from dollops rapidly decreased between the first and second day of volatile collection, but stabilized after day 15. The mean pheromone release rate between 15 and 31 days was 228 nghr-1for each dollop.

Field Study

Reproductive adults. In 2014, weekly sweeps of overwintered P adults yielded at a mean of 1.3 on plants receiving the BL treatment, compared to 3.5 on Tamarix receiving the PH treatment (P<0.001 on 1, 185.22 d.f.) (Fig 2a). The 2014 weekly sweeps of F1 adults resulted in much greater mean densities, 7.3 adults from BL vs. 14.31 adults from PH treatment plants (P<0.001 on 1, 232.1 d.f.) (Fig 2b). There was a significant site 90 effect for F1 adult numbers in 2014 (P=0.001 on 1, 232.1 d.f.); however, interaction of the numbers by treatment and site was not significant (P=0.58 on 1, 232.1 d.f.).

Mean weekly sweep captures of overwintered P adults in 2015 were less than recorded in 2014, averaging 1.07 adults on BL vs. 2.24 adults on PH treated Tamarix

(P<0.001 on 1, 186.27 d.f.) (Fig 1c). Counts of F1 adults averaged 1.16 and 2.65 adults per three sweeps, respectively, on BL vs. PH plants (P<0.001 on 1, 208.34 d.f.) (Fig 2d).

There was no significant site effect for the overwintered P or F1 adults in 2015 (P=0.21 on 1, 186.27 d.f.; P=0.09 on 1, 211.40 d.f.).

Non-reproductive Adults. No pheromone-mediated difference in weekly sweep captures of F2 adults occurred in 2014 (P=0.28 on 1, 234.38 d.f.) (Fig 3a). There was a significant site effect (P=0.008 on 1, 234.38 d.f.), but the interaction between site and treatment for F2 adults was not significant (P=0.32 on 1, 234.38 d.f.). In 2015, plants receiving the PH treatment had significantly higher densities of F2 adults, averaging 6.58 adults compared to 2.74 adults on the BL treated plants (P<0.001 on 1, 348.04 d.f.) (Fig

3b). There was a significant site effect in 2015 (P=0.02 on 1, 350.27 d.f.), and an ordinal interaction between site and treatment (P=0.01 on 1, 348.04 d.f.).

Larval Densities. The increased densities of adult D. carinulata captured that are reported above, also had increased mean densities of larvae in weekly sweeps of plants receiving the PH treatment (Fig 4). In 2014, larvae captured per three sweeps each week averaged 5.14 on plants treated with PH, compared to 3.66 on plants treated with BL

(P=0.007 on 1, 466.95 d.f.) (Fig 4a). There was no site effect for numbers of larvae in 91

2014 (P=0.23 on 1, 466.95 d.f.); however, there was a significant ordinal interaction between treatment and site (P=0.05 on 1, 466.95 d.f.). Densities of larvae captured in weekly sweeps during 2015 were lower overall, but the numbers were still greater on PH compared to BL plants, averaging 3.18 vs. 1.54, respectively (P<0.001 on 1, 514.11 d.f.)

(Fig 4b).

Damage, Dieback and Canopy Volume. In 2014, feeding damage ratings for PH- treated plants differed significantly from damage ratings reported for BL plants (P<0.001 on 1, 654.98 d.f.) (Fig 5a). Although all monitored plants (i.e., PH and BL) experienced near complete defoliation during summer 2014, in-season regrowth after defoliation reported for BL plants was significantly greater compared with regrowth after defoliation reported for PH plants (Fig. 5a). In 2015, PH plants had greater feeding damage than BL treated plants (P<0.001 on 1, 799.10 d.f.) (Fig 5b). No significant in-season regrowth was observed in 2015.

The increased feeding damage on PH plants was reflected in increased dieback that was observed the following year in both 2015 (P=0.003 on 1, 44 d.f.) and 2016

(P<0.001 on 1,43 d.f.) (Fig 6). Dieback was greater for both the PH and BL treated plants in 2015 compared to 2016, most likely due to the greater level of defoliation recorded across the study site in 2015 compared to 2016. Canopy volume did not significantly differ one year after defoliation. After a second season, PH treated plants had smaller canopy volumes compared with study plants receiving a second season of BL treatment

(P=0.01 on 1, 140 d.f) (Fig 7).

Discussion

The use of synthetic aggregation pheromones to manipulate the spatial distribution of biological control agents has been suggested as a way to intensify damage on under- exploited weed infestations (Wood et al. 2007). The results of this investigation confirm that the deployment of an aggregation pheromone is a feasible approach for manipulating the density and spatial distribution of D. carinulata to enhance biological control of

Tamarix. Semiochemical attraction and retention for intensification of biological control has been reported; however, existing methodologies were typically time-consuming and had limited activity often requiring frequent re-application of treatments. The methods described here yielded similar results to Gaffke et al. (2018), but required less frequent applications of the pheromone-based lure. Remarkably, effective spatial manipulations and enhanced herbivory were achieved with only three targeted applications of the aggregation pheromone per year. The ability of biological control practitioners to alter densities of a classical biological control agent weeks after an application of an attractive lure could have broad impacts in the field of biological control.

Emissions of synthetic (2E,4Z)-2,4-heptadien-1-ol, the D. carinulata aggregation pheromone, were highly variable for the first five days after field deployment and exposure of 4-g dollops of SPLAT to environmental conditions. Emission rates stabilized at 10 days and remained relatively constant over the following 21 days, until the study was terminated on day 31. On the first day, the hourly amount of synthetic aggregation pheromone emitted from 4-g dollops approached the amount estimated to be produced by

3,300 adult D. carinulata males in one hour. The average pheromone release rate from 93 the 4-g dollops between 10 and 31 days was equivalent to estimated hourly pheromone emissions by 60 adult male D. carinulata (Cossé et al. 2005). The specific pheromone emission rate required to initiate and maintain an aggregation of D. carinulata is unknown, but considering that aggregations typically would contain at least 50 individuals, the emissions reported from the 4-g treatment dollops would likely facilitate and maintain aggregations of D. carinulata for several weeks under field conditions. This estimate of treatment efficacy is supported by the sweep capture densities recorded in this study, which reported that increased densities of adults were evident weeks after the application of pheromone-treated dollops.

Differences in captures of reproductive adult beetles on Tamarix receiving the control vs. pheromone treatment indicates that a single application of a 4-g dollop of SPLAT impregnated with (2E,4Z)-2,4-heptadien-1-ol can maintain local aggregations of adult D. carinulata over the course of their generational life span (Fig 2). Increased densities of reproductive adults reported on PH vs. BL-treated Tamarix was generally correlated with more eggs being laid on the PH plants, and in turn resulted in greater numbers of larvae in sweep net counts. The response of the non-reproductive F2 adults was not consistent across the two years of this study. No pheromone mediated changes in density were observed in 2014, but increased F2 adult densities were recorded in 2015 (Fig 3).

The peak larval densities recorded in 2014 on control plants during week eight and nine of monitoring were composed of 1st and 2nd instars with very few 3rd instars observed (Fig 4a). Although these larvae are mobile and able to move from defoliated plants to seek new host plants, the extensive level of defoliation observed in 2014 likely 94 resulted in the starvation of many F2 larvae (Moran et al. 2009). The second generation of larvae likely did not complete development in 2014. This resulted in a localized collapse of the beetle population, this conclusion is supported by the reduced capture rates of F2 vs. F1 adults in 2014 (Fig 2b, Fig 3a). The lack of a second generation of adults in 2014 is likely the reason for the reduction in total numbers of adults observed in

2015 (Fig 3c, Fig 3d). The F1 reproductive adults failed to reproduce at the research locations in 2014, probably because the trees were already approaching complete defoliation while these adults were searching for suitable hosts. The F2 adults observed were likely transient adults from outside the research area that were moving through study plots in search of host plants. The lack of successful F1 reproduction in 2014 confirms that while this lure promotes aggregation of D. carinulata, reproduction will not succeed if an area is rendered unsuitable to support progeny due to the lack of foliage.

Increased rates of damage, dieback, and greater changes in canopy volume were correlated with greater D. carinulata densities. Damage ratings for monitored trees varied greatly across the years. In 2014, all of the monitored plants, regardless of the treatment received, were nearly completely defoliated by week 10, with observable damage occurring primarily after the week 5. After the week 10, the BL plants had substantially higher rates of regrowth compared to the PH plants, as indicated by a decrease in mean damage rating for BL plants (Fig 5a). This suggests that the increased damage rating for

PH plants earlier in the season is indicative of the influence of the pheromone treatment on the timing or intensity of herbivory, which impairs the ability of PH-treated plants to regrow after defoliation. 95

Higher in-season feeding damage ratings reported for all study plants in 2014 were correlated with a greater rate of dieback on all study plants in 2015. Differences in canopy volume were, in contrast to the more immediate influence of damage on dieback, only detected after plants experienced a second year of defoliation. However, the limited but targeted application of the D. carinulata aggregation pheromone resulted in a 73% reduction in canopy volume in only two years. This result was unexpected because

Tamarix exposed to two years of defoliation by Diorhabda spp. typically experiences a

25-33% reduction in canopy volume (Dudley and Deloach 2004).

Application of D. carinulata aggregation pheromone effectively allows for land managers to dictate where D. carinulata feeding damage will occur. This would make pheromone formulations a very useful tool in managing Tamarix stands that have not been defoliated enough times to cause canopy die back or mortality. Studies in which impact has been measured show a high degree of variation in die-back and mortality from site to site (Kennard et al 2016). In one study mortality was measured at nearly 80% at a site that had been defoliated multiple times, while at a neighboring site two kilometers away, which had been more sporadically defoliated, mortality was only 20% (Bean et al

2013). In such instances pheromone formulations could be used to attract and concentrate beetles to areas with low mortality or die back. There are areas in western Colorado where D. carinulata have been present for ten years, and where impact has been minimal

(Kennard et al 2016; Bean, personal observation). Land owners would welcome the use of pheromone as a tool to increase impact of the beetles on tamarisk. 96

By pulling beetles away from some areas where Tamarix is thought to provide transient ecosystem services, the use of pheromones could have broad implications for conservation, as areas sensitive to the removal or damage of Tamarix can be avoided

(Dudley and Deloach 2004), while the impacts can be intensified in other areas (Gaffke et al. 2018a). The application of this technology could be beneficial to the conservation of native habitat. For example, Tamarix commonly dominates landscapes, but there is usually a remnant population of native vegetation within the Tamarix stand (Friedman et al. 2005; Shafroth et al. 2013). Applying the aggregation pheromone to Tamarix plants surrounding remnant native plants can remove the competition exerted by Tamarix with minimal risk or disturbance to the native plants. Conditions facilitating recruitment of cottonwoods (Populus spp.) and willows (Salix spp.) are also commonly associated with the successful recruitment of Tamarix, resulting in seedlings nurseries that are composed of both the native plants and Tamarix (Horton et al. 1960; Fenner et al. 1984; Sher et al.

2002). While native seedlings can outcompete Tamarix seedlings under natural flow regimes in riparian habitats, altered flow regimes favor the dominance of Tamarix (Sher et al. 2002; Li et al. 2012; Merrit and Poff 2010). Selective removal of Tamarix seedlings from these mixed species nurseries, whether through herbicide application or mechanical removal, is time-consuming and all but impossible to accomplish, considering that

Tamarix seedlings can reach densities as high as 170,000 individuals/m2 on mud flats

(Warren and Turner 1975). If land managers used pheromone lures to increase D. carinulata aggregations on mixed native/Tamarix nurseries, the host specificity of the agent would confer a competitive advantage on non-Tamarix seedlings that could 97 compensate for altered flow growing conditions (Crawley 1989; Maron 1996). Increased mortality of Tamarix during seedling establishment has the potential to reduce long term negative environmental impacts of Tamarix invasion, such as salt accumulation and altered fire regimes, on the viability and diversity of affected native plant communities

(Salinas et al. 2000; Busch 1995).

The ability to manipulate the spatial distribution of D. carinulata has broad applications outside of enhancing biological control of Tamarix. By applying the aggregation pheromone, and manipulating the densities of D. carinulata in targeted areas, dispersal and establishment can be readily assessed. Specifically, the aggregation pheromone can be used to implement a monitoring procedure where plants are treated and act as ‘sentinels’ for the presence of D. carinulata (Britton et al. 2010). Treated sentinel plants would allow for easy monitoring of D. carinulata, especially if combined with passive yellow sticky card or delta traps (Cossé et al. 2005, 2006). The distribution of D. carinulata in the United States is well documented through the monitoring efforts of the Tamarisk Coalition (http://www.tamariskcoalition.org) and multiple collaborators, although current distributions have inconsistencies owing to the difficulty of monitoring small sized populations (Jamison and Lanci 2012). The rapid expansion of D. carinulata southward in the United States, due to post-release evolutionary changes in critical photoperiod inducing diapause, has resulted in the urgent need for effective and accurate monitoring techniques (Dudley and Bean 2012, Sogge et al. 2008). A network of aggregation pheromone treated sentinel plants could yield an accurate and robust early detection system for population expansion of D. carinulata. 98

Pheromone applications used where D. carinulata populations are small or newly established would also facilitate mate finding and produce protective aggregations, increasing the odds of persistence for vulnerable populations. Aggregations are important to many insect species for protecting individuals, increasing mating frequency, and increasing feeding (Wertheim et al. 2005). Aggregation also offers protection from Allee effects, decreased fitness associated with low density or small populations that may result in extinction (Allee et al. 1949). Allee effects are considered a primary factor driving establishment failure of many biological control agents (Hopper and Roush 1993;

Grevstad 1999; Fagan et al. 2005). Mate-finding Allee effects are especially detrimental to newly released or small established populations of biological control agents (Hopper and Roush 1993; Gascoigne et al. 2009; Fauvergue 2012; Robinet et al. 2008). Successful location of a mate declines with decreasing density and population size (Gascoigne et al.

2009). The location of mates can also be confounded by emigration of individuals out of the system, and where there is insufficient compensatory immigration, it is likely that

Allee effects will drive such populations to extinction (Berec et al. 2001). Since aggregation pheromones greatly increase the likelihood of finding mates (Wertheim et al.

2005), the application of an aggregation pheromone could result in substantially increased likelihood of establishment and persistence of a biological control agent by retaining and directing insects into continued conspecific contact. When making a new release of D. carinulata, applying the aggregation pheromone in conjunction with the release would likely result in increased retention of the insects at the release site, increasing successful establishment while reducing or eliminating Allee effects. 99

A further strategy that could be integrated to enhance the ability to manipulate the spatial distribution of D. carinulata would be the deployment of a repellent compound in conjunction with the aggregation pheromone (Witzgall et al. 2010). This would develop a

‘push-pull’ strategy that would allow for further control potential over the distribution of

D. carinulata (Borden et al. 2006). Not only could land managers dictate where they want D. carinulata to aggregate, they could also dictate where they don’t want D. carinulata to aggregate. The application of a repellent compound could deter D. carinulata form colonizing areas that are at risk of flooding or have high predator densities, thus conserving individuals in the population (Bean et al. 2007a). The combined application of a repellent compound with the aggregation pheromone could further increase the operational feasibility of semiochemical applications to enhance D. carinulata biological control of Tamarix.

Acknowledgements

This work was supported by agreement 16-CA-11420004-047 of the Forest

Health Technology and Enterprise Team- Biological Control of Invasive Native and Non-

Native Plants (FHTET-BCIP). We thank Jon Rico, Rodrigo Silva and Carmen Bernardi for formulating the aggregation pheromone into SPLAT. We thank Megan Hofland and

Adam Cook for assistance. We also thank the Bighorn National Recreation Area (BICA-

2014-SCI-0002) for providing research locations. 100

100000

10000

1000

1 -

nghr 100

1 1 3 5 10 15 20 25 31 Days

Figure 1: Mean ± SE of (2E,4Z)-2,4-heptadien-1-ol emitted from 4-g dollops. 101

(a) 2014 (b) 2014

40 40 35 35 30 30 25 25 20 20 BL BL

15 15 MeanDensity MeanDensity PH PH 10 10 5 5 0 0 1 2 3 4 5 6 7 8 9 Weeks Weeks

10 10 9 9 8 8 7 7 6 6 5 5 BL BL

4 4 MeanDensity MeanDensity 3 PH 3 PH 2 2 1 1 0 0 1 2 3 4 5 8 9 10 11 12 Weeks Weeks

Figure 2: Weekly mean ± SE capture per three sweeps of reproductive adult Diorhabda carinulata on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL). Densities captured 2014-15 of P generation overwintered adults (a, c) and F1 first generation of current year adults (b, d). 102

(a) 2014 (b) 2015

25 25

20 20

15 15

BL BL

10 10 MeanDensity MeanDensity PH PH 5 5

0 0 10 11 12 13 14 13 14 15 16 17 18 19 Weeks Weeks

Figure 3: Weekly mean ± SE capture per three sweeps of non-reproductive adult Diorhabda carinulata on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL). Densities of the non-reproductive overwintering F2 generation of adults captured in 2014 (a), and 2015 (b).

103

(a) 2014 (b) 2015

25 25

20 20

15 15

BL BL

10 10 MeanDensity MeanDensity PH PH 5 5

0 0 1 2 3 4 5 6 7 8 9 1011 4 6 8 10 12 14 16 Week Week

Figure 4: Weekly mean ± SE larval Diorhabda carinulata capture per three sweeps on plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or 4g dollops of blank SPLAT (BL), in 2014 (a), and 2015 (b).

104

2014 2015 100 100 90 90 80 80 70 70 60 60 50 50 BL BL 40 40 30 PH 30 PH

20 20 Mean Damage Rating(%) Mean Damage Rating(%) 10 10 0 0 1 3 5 7 9 11 13 1 3 5 7 9 11 13 15 17 19 Weeks Weeks

Figure 5: Weekly mean ± SE damage rating for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL), in 2014 (a) and 2015 (b).

105

100 b 90

80 a 70

60 b

50 BL 40 a PH

30 MeanDieback rating(%) 20

0 2015 2016 Year

Figure 6: Mean ± SE dieback rating for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL) in 2015 and 2016.

106

20 BL

15 PH

Canopy Volume (m3) 10

0 2014 2015 2016 Year

Figure 7: Change in canopy volume 2014-2016 for plants receiving 4-g dollops of pheromone impregnated SPLAT (PH) or blank SPLAT (BL).

107

References

Allee, W.C., Park, O., Emerson, A.E., Park, T. and Schmidt, K.P. 1949. Principles of animal ecology. Saunders Co., Philadelphia, PA, USA.

Bartelt, R.J., Cossé, A.A., Zilkowski, B.W., Weisleder, D. and Momany, F.A. 2001. Male-specific sesquiterpenes from Phyllotreta and Aphthona flea beetles. Journal of Chemical Ecology, 27: 2397-2423.

Bartelt, R.J., Cossé, A.A., Zilkowski, B.W., Weisleder, D., Grode, S.H., Wiedenmann, R.N. and Post, S.L. 2006. Dimethylfuran-lactone pheromone from males of Galerucella calmariensis and Galerucella pusilla. Journal of Chemical Ecology, 32: 693-712.

Bean, D.W., Dudley, T.L. and Keller, J.C. 2007a. Seasonal timing of diapause induction limits the effective range of Diorhabda elongata deserticola (Coleoptera: Chrysomelidae) as a biological control agent for tamarisk (Tamarix spp.). Environmental Entomology, 36: 15-25.

Beran, F., Jiménez-Alemán, G.H., Lin, M.Y., Hsu, Y.C., Mewis, I., Srinivasan, R., Ulrichs, C., Boland, W., Hansson, B.S. and Reinecke, A. 2016. The aggregation pheromone of Phyllotreta striolata (Coleoptera: Chrysomelidae) revisited. Journal of Chemical Ecology, 42: 748-755.

Berec, L., Boukal, D.S. and Berec, M. 2001. Linking the Allee effect, sexual reproduction, and temperature-dependent sex determination via spatial dynamics. The American Naturalist, 157: 217-230.

Britton, K.O., White, P., Kramer, A. and Hudler, G. 2010. A new approach to stopping the spread of invasive insects and pathogens: early detection and rapid response via a global network of sentinel plantings. New Zealand Journal of Forestry Science, 40: 109-114.

108

Borden, J.H., Birmingham, A.L. and Burleigh, J.S. 2006. Evaluation of the push-pull tactic against the mountain pine beetle using verbenone and non-host volatiles in combination with pheromone-baited trees. The Forestry Chronicle, 82: 579-590.

Busch, D.E., 1995. Effects of fire on southwestern riparian plant community structure. The Southwestern Naturalist, 40: 259-267.

Crawley, M.J. 1989. Insect herbivores and plant population dynamics. Annual Review of Entomology, 34: 531-562.

Douglass, C.H., Nissen, S.J. and Hart, C.R. 2013. Tamarisk Management. Tamarix: a case study of ecological change in the American West. Oxford University Press, New York, NY:333-353.

Dudley, T.L. and Bean, D.W., 2012. Tamarisk biocontrol, endangered species risk and resolution of conflict through riparian restoration. BioControl, 57: 331-347.

Dudley, T.L. and DeLoach, C.J. 2004. Saltcedar (Tamarix spp.), Endangered Species, and Biological Weed Control-Can They Mix? 1. Weed Technology, 18: 1542- 1551.

109

Fagan, W.F., Lewis, M.A., Neubert, M.G. and Van Den Driessche, P. 2002. Invasion theory and biological control. Ecology Letters, 5: 148-157.

Fauvergue, X., Vercken, E., Malausa, T. and Hufbauer, R.A. 2012. The biology of small, introduced populations, with special reference to biological control. Evolutionary applications, 5: 424-443.

Fenner, P., Brady, W.W. and Patton, D.R. 1984. Observations on seeds and seedlings of Fremont cottonwood. Desert Plants, 6: 55-58.

Fors, L., Liblikas, I., Andersson, P., Borg-Karlson, A.K., Cabezas, N., Mozuraitis, R. and Hambäck, P.A. 2015. Chemical communication and host search in Galerucella leaf beetles. Chemoecology, 25: 33-45.

Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Grieco, P.A., Peterson, R.K.D., Weaver, D.K. 2018a. Semiochemicals to enhance herbivory by Diorhabda carinulata aggregations in saltcedar (Tamarix spp.) infestations. Pests Management Science, doi: 10.1002/ps.4848

Gascoigne, J., Berec, L., Gregory, S. and Courchamp, F. 2009. Dangerously few liaisons: a review of mate-finding Allee effects. Population Ecology, 51: 355-372.

Gaskin, J.F. and Schaal, B.A. 2002. Hybrid Tamarix widespread in US invasion and undetected in native Asian range. Proceedings of the National Academy of Sciences, 99: 11256- 11259.

Global Invasive Species Database 2018. Available via DIALOG. http://www.iucngisd. org/gisd/100_worst.php on 14-12-2018.Cited March 9, 2018

Grevstad, F.S. 1999. Experimental invasions using biological control introductions: the influence of release size on the chance of population establishment. Biological Invasions, 1: 313-323.

Hambäck, P.A., 2010. Density-dependent processes in leaf beetles feeding on purple loosestrife: aggregative behaviour affecting individual growth rates. Bulletin of Entomological Research, 100: 605-611.

Hopper, K.R. and Roush, R.T. 1993. Mate finding, dispersal, number released, and the success of biological control introductions. Ecological Entomology, 18: 321-331.

110

Horton, J.S. 1960. Use of a root plow in clearing tamarisk stands. United States Department of Agriculture, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 50.

Horton, J.S. and Campbell, C.J. 1974. Management of phreatophyte and riparian vegetation for maximum multiple use values. U.S. Department of Agriculture, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 17.

Hultine, K.R., Nagler, P.L., Morino, K., Bush, S.E., Burtch, K.G., Dennison, P.E., Glenn, E.P. and Ehleringer, J.R. 2010. Sap flux-scaled transpiration by tamarisk (Tamarix spp.) before, during and after episodic defoliation by the saltcedar leaf beetle (Diorhabda carinulata). Agricultural and Forest Meteorology, 150: 1467- 1475.

Jamison L and J Lanci. 2012. Tamarisk Leaf Beetle Monitoring Protocol Colorado Department of Agriculture, Tamarisk Coalition, and University of California- Santa Barbara. Located at:http://tamariskcoalition.org/sites/default/files /files/2012_Tamarisk_Leaf_Beetle_Monitoring_Protocol_07_01_2014.pdf.

111

Kuznetsova, A., Brockhoff, P.B. and Christensen, R.H.B. 2015. Package ‘lmerTest’. R package version, 2.

Ladenburger, C.G., Hild, A.L., Kazmer, D.J. and Munn, L.C. 2006. Soil salinity patterns in Tamarix invasions in the Bighorn Basin, Wyoming, USA. Journal of Arid Environments, 65: 111-128.

Landolt, P.J. 1997. Sex attractant and aggregation pheromones of male phytophagous insects. American Entomologist, 43:12-22.

Lewis, P.A., DeLoach, C.J., Knutson, A.E., Tracy, J.L. and Robbins, T.O. 2003. Biology of Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), an Asian leaf beetle for biological control of saltcedars (Tamarix spp.) in the United States. Biological Control, 27: 101-116.

Li, J., Yu, B., Zhao, C., Nowak, R.S., Zhao, Z., Sheng, Y. and Li, J., 2012. Physiological and morphological responses of Tamarix ramosissima and Populus euphratica to altered groundwater availability. Tree Physiology, 33: 57-68.

Mafra-Neto, A., de Lame, F.M., Fettig, C.J., Munson, A.S., Perring, T.M., Stelinski, L.L., Stoltman, L.L., Mafra, L.E., Borges, R. and Vargas, R.I. 2013. Manipulation of insect behavior with specialized pheromone and lure application technology (SPLAT®). Pest management with natural products. American Chemical Society, Washington DC, USA: 31-58.

Maron, J.L. 1997. Interspecific competition and insect herbivory reduce bush lupine (Lupinus arboreus) seedling survival. Oecologia, 110: 284-290.

Merritt, D.M. and Poff, N. 2010. Shifting dominance of riparian Populus and Tamarix along gradients of flow alteration in western North American rivers. Ecological Applications, 20: 135-152.

Moran, P.J., DeLoach, C.J., Dudley, T.L. and Sanabria, J., 2009. Open field host selection and behavior by tamarisk beetles (Diorihabda spp.)(Coleoptera: Chrysomelidae) in biological control of exotic saltcedars (Tamarix spp.) and risks to non-target athel (T. aphylla) andnative Frankenia spp. Biological Control, 50: 243-261.

Petroski, R.J., 2003. Straightforward preparation of (2 E, 4 Z)-2, 4-heptadien-1-ol and (2 E, 4 Z)-2, 4-heptadienal. Synthetic Communications, 33: 3233-3241.

Robinet, C., Lance, D.R., Thorpe, K.W., Onufrieva, K.S., Tobin, P.C. and Liebhold, A.M. 2008. Dispersion in time and space affect mating success and Allee effects in invading gypsy moth populations. Journal of Animal Ecology, 77: 966-973. 112

Robinson, T.W. 1952. Symposium on phreatophytes and their relation to water in western United States. Eos, Transactions American Geophysical Union, 33: 57-61.

Robinson, T.W. 1965. Introduction, spread, and areal extent of saltcedar (Tamarix) in the western states. Geological Survey Professional Paper, Contact No. 491-A.

Salinas, M.J., Blanca, G. and Romero, A.T. 2000. Riparian vegetation and water chemistry in a basin under semiarid Mediterranean climate, Andarax River, Spain. Environmental Management, 26: 539-552.

Shafroth P, Merritt D, Briggs M, Beauchamp V, Lair K, Scott M, Sher A (2013). Riparian restoration in the context of Tamarix control. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 404-425.

Sher, A.A., Marshall, D.L. and Taylor, J.P. 2002. Establishment patterns of native Populus and Salix in the presence of invasive nonnative Tamarix. Ecological Applications, 12: 760- 772.

Sogge, M.K., Sferra, S.J. and Paxton, E.H. 2008. Tamarix as habitat for birds: implications for riparian restoration in the southwestern United States. Restoration Ecology, 16: 146-154.

Sutton, G.F., Paterson, I.D., Compton, S.G. and Paynter, Q. 2017. Predicting the risk of non-target damage to a close relative of a target weed using sequential no-choice tests, paired-choice tests and olfactory discrimination experiments. Biocontrol Science and Technology, 27: 364-377.

Tamarisk Coalition. 2017. Colorado Department of Agriculture, Palisade Insectary TamariskMonitoring Protocol. http://www.tamariskcoalition.org/sites/default/ files/files/Tamarisk_Impact_Monitoring_Protocol_2013_Palisade_Insectary.pdf. Accessed on 21-July-2017.

113

Tansey, J.A., McClay, A.S., Cole, D.E. and Keddie, B.A. 2005. Evidence for the influence of conspecific chemical cues on Aphthona nigriscutis (Coleoptera: Chrysomelidae) behaviour and distribution. BioControl, 50: 343-358.

USDA, NRCS (2018). The PLANTS Database (http://plants.usda.gov, February 15, 2018). National Plant Data Team, Greensboro, NC 27401-4901 USA.

Warren, D.K. and Turner, R.M. 1975. Saltcedar (Tamarix chinensis) seed production, seedling establishment, and response to inundation. Journal of the Arizona Academy of Science, 10: 135-144.

Wertheim, B., van Baalen, E.J.A., Dicke, M. and Vet, L.E., 2005. Pheromone-mediated aggregation in nonsocial arthropods: an evolutionary ecological perspective. Annual Review of Entomology, 50: 321-346.

Wood, M., Flores, A., McGinnis, L. and Peabody, E. 2007. Lassoing wicked weeds of the West. Agricultural Research, 55: 16.

Yoccoz, N.G., Nichols, J.D. and Boulinier, T., 2001. Monitoring of biological diversity in space and time. Trends in Ecology & Evolution, 16: 446-453.

Zhang, Z.Q. and McEvoy, P.B., 1994. Attraction of Longitarsus jacobaeae males to cues associated with conspecific females (Coleoptera: Chrysomelidae). Environmental entomology, 23: 732-737.

114

CHAPTER FOUR

ALLEE EFFECTS AND AGGREGATION PHEROMONES: NEW RELEASES OF

DIORHABDA CARINULATA REMAIN LONGER IN THE PRESENCE OF

PHEROMONE FORMULATIONS

Contribution of Authors and Co-Authors

Manuscript in Chapter 4

Author: Alexander M. Gaffke

Contributions: Designed and implemented experiments, collected, processed and analyzed data, and prepared manuscript for publication

Co-Author: Sharlene E. Sing

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Tom L. Dudley

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Daniel W. Bean

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Justin A. Russak

Contributions: Developed methodology for chemical synthesis

Co-Author: Agenor Mafra-Neto

115

Contribution of Authors and Co-Authors Continued

Contributions: Provided critical input on the formulation of the semiochemcials for field deployment

Co-Author: Robert K. D. Peterson

Contributions: Provided input on result interpretation and manuscript preparation

Co-Author: David K. Weaver

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

116

Manuscript Information Page

Alexander M. Gaffke, Sharlene E. Sing, Tom L. Dudley, Daniel W. Bean, Justin A. Russak, Agenor Mafra-Neto, Robert K. D. Peterson, David K. Weaver

Biological Invasion

Status of Manuscript: [Put an x in one of the options below, delete this] __x_ Prepared for submission to a peer-reviewed journal ____ Officially submitted to a peer-review journal ____ Accepted by a peer-reviewed journal ____ Published in a peer-reviewed journal

117

ALLEE EFFECTS AND AGGREGATION PHEROMONES: NEW RELEASES OF

DIORHABDA CARINULATA REMAIN LONGER IN THE PRESENCE OF

PHEROMONE FORMULATIONS

Alexander M. Gaffkea, Sharlene E. Singb, Tom L. Dudleyc, Daniel W. Beand, Justin A. Russake, Agenor Mafra-Netof, Robert K. D. Petersona, David K. Weavera* aDepartment of Land Resources and Environmental Sciences, Montana State University,

Santa Barbara CA, 93106, USA fISCA Technologies, Inc., Riverside CA 92507, USA

*Correspondence to: David K. Weaver, Montana State University, P.O. Box 173120, Bozeman, MT, 59717. Email: [email protected]

118

Abstract

The northern tamarisk beetle Diorhabda carinulata (Desbrochesrs) was introduced to the United States as a biological control agent for the exotic, invasive plant saltcedar (Tamarix spp.). In many areas, populations of D. carinulata have not established, likely due to Allee effects associated with small founding populations at the time of initial release. We therefore assessed whether species-specific aggregation pheromone lures could enable newly released adults to better persist at target locations. This was evaluated by comparing numbers of adult beetles at release sites with a pheromone lure with numbers at paired nearby release sites without a pheromone lure 10-14 days after the release of 300 or 500 individuals. A greater number of adult D. carinulata were captured where the pheromone lures had been strategically deployed. In contrast, the limited numbers of adult beetles at sites without a pheromone lure 10-14 days after the release was made confirms the propensity of D. carinulata to rapidly disperse from release sites. The application of aggregation pheromone resulted in larger populations after the initial release and for several generations. Application of aggregation pheromone when a new release of biological control agents is planned may allow biological control practitioners to limit negative outcomes due to Allee effects.

Key words: Biological control, colonization, release protocol, biological invasion, weed control

119

Introduction

The probability of populations establishing in a location following unplanned biological invasions by pests, or after the intentional introduction of biological control agents, typically increases with the number of individuals introduced (Hopper and Roush

1993; Memmott et al. 1998; Bergren 2001; Grevstad 1999). Factors that heavily influence biological invasions and successful colonization are Allee effects (Dennis 2002; Berec et al. 2008), environmental suitability (Duncan 2016), and life history (Roecchi et al. 2001).

The two factors with the greatest effect on successful colonization are environment

(Peterson 2003) and Allee effects (Fauvergue et al. 2012; Gascoigne et al. 2009; Duncan et al. 2014). Environmental conditions are critically important, resulting in intensive research to match biological control agents with suitable habitat to ensure that small cohorts are introduced under favorable conditions (Robertson et al 2008; Hoelmer et al.

2005; Winston et al. 2014). This knowledge allows scientists and biocontrol practitioners to release agents in suitable habitats, minimizing negative impacts on establishment

(Fauvergue et al. 2012). Because this knowledge counters negative environmental effects on the establishment of a biological control agent, Allee effects become the factor most likely to impede the successful establishment of biological control agents (Hopper and

Roush 1993; Simberloff 2009; Lockwood et al. 2005).

Many cases where agents failed to establish are undoubtedly due to attrition from biological and climatic or other environmental challenges (Grevstad et al. 2011). Apart from those significant challenges, most introductions are inherently vulnerable to Allee effects due to small population sizes (Hopper and Roush 1993). The loss of mean 120 individual fitness correlated with sustained low population density is manifested in Allee effects. Allee effects can arise when population density drops below a threshold resulting in difficulty or failure in mate location (Gascoigne et al. 2009), when the ability to avoid predation through cooperation cannot be achieved (Fauvergue 2012), and because foraging efficacy and environmental conditioning can also decrease (Raffa et al. 1993).

These vulnerabilities can be magnified in the field of biological control when practitioners have a limited number of individuals available for making new releases. In classical biological control, a ‘release’ is the process of introducing an agent into a new area with the goal of establishing a self-sustaining population of the agent. Releases typical contain multiple individuals but can range from sizes of one gravid female

(Grevstad 1999), to tens of thousands of individuals (Grevstad et al. 2003). Limited access to agents represents the major and common challenge to getting populations of a new agent established in as many areas as possible, and as quickly as possible (Memmott et al 1996). A review of related literature suggests that a single large release at a single site would potentially result in the greatest likelihood of establishment, but this strategy could result in complete failure when unforeseen catastrophic disturbance leads to local extinction, or if environmental attributes present or missing from the site make it unfavorable or unsuitable for agent establishment. Although multiple releases of smaller sizes would protect against single, site-specific events that could prevent establishment, the likelihood of establishment is lower when smaller releases are used (Grevstad 1999;

Fauvergue 2012). If this strategy is adopted, the question then becomes how to minimize 121

Allee effects on smaller releases to maximize the establishment of multiple smaller releases (Freckelton 2000; Shea and Possingham 2000).

Organisms have evolved multiple strategies, ranging from mating calls and sex pheromones to hermaphroditism, to minimize deleterious Allee effects (Gascoigne et al.

2009). The formation of aggregations can also limit Allee effects, especially mate-finding

Allee effects, by directly increasing population densities (Landolt 1997). Many nonsocial arthropods utilize aggregation pheromones to eliminate Allee effects (Wertheim et al. 2005; Raffa et al. 1993; Wood 1982; Turchin and Kareiva 1989). Several insects used for weed biological control are known to aggregate in the field, especially those within the Chrysomelidae. Gallerucella calmarienis (L.) and G. pusilla (Duftschmidt) are two weed biological control agents known to facilitate aggregations through the production of aggregation pheromones (Bartelt et al. 2006; Hambäck 2010; Fors et al. 2015).

Diorhabda carinulalta (Desbrochers), another chrysomelid introduced to North America for the classical biological control of the invasive weed Tamarix spp. also aggregates utilizing a male-produced aggregation pheromone (Cossé et al. 2005, 2006; Gaffke et al.

2018a, 2018b). Chrysomelid biological control agents of leafy spurge, the flea beetles

Aphthona nigriscutis, A. flava, A. czwalinae, and A. cyparissiae also display aggregation behavior and males produce compounds likely to be aggregation pheromones (Tansey et al. 2005; Bartelt et al. 2001). The production of aggregation pheromones has been hypothesized as a critical component to the successful establishment, spread, and colonization of G. calmariensis, G. pusilla, D. carinulata, and D. elongata (Bartelt et al.

2001; Cossé et al. 2006; Hudgeons et al. 2007a). For these weed biological control 122 species, the males are considered to be the ‘pioneers’ leaving the aggregation and searching for other suitable hosts. Dispersing individuals alight on host plants and pheromone production begins, rapidly calling in nearby conspecifics to the area to form a new aggregation and thereby diminish the potential for detrimental Allee effects acting on the newly colonized area (Bartelt et al. 2001; Cossé et al. 2006). Deployment of species-specific aggregation pheromones with small release sizes of weed biological control agents could increase the likelihood of establishment by retaining an adequate number of individuals on site to sufficiently reduce or prevent Allee effects.

Diorhabda carinulata was approved for use as a biological control agent for the invasive weed Tamarix spp. (hereafter referred to as Tamarix) in the western United

States (DeLoach et al. 2003). Tamarix was introduced to the United States in the early

1800s for erosion control and as an ornamental species (Robinson 1965). Tamarix has become the third most prevalent woody riparian plant in the western United States

(Friedman et al. 2005). This rapid expansion has resulted in degradation of wildlife habitat and significant economic costs approaching $185 million annually (Dudley et al.

2000; Zavaleta 2000).

Diorhabda carinulata is multivoltine, typically having 2-3 generations per year

(Lewis et al. 2003b). The adults overwinter in the soil and emerge in the spring, coinciding with Tamarix budbreak (Bean et al. 2007b). The adults are reproductively active in spring and early summer, depositing clusters of eggs on Tamarix foliage.

Reproductive males produce two aggregation pheromone compounds, (2E,4Z)-2,4- heptadien-1-ol and (2E,4Z)-2,4-heptadienal. The alcohol component of the pheromone 123 blend is the primary behaviorally active compound that stimulates the formation of aggregations and facilitates colonization of new areas (Cossé 2005; Bean et al. 2013).

Production of an aggregation pheromone is critical to maintaining high densities when the insects are dispersing over short and long distances (Grevstad and Herzig 1997;

Fernandez and Hilker 2007; Wood 1982). The adults enter a reproductive diapause in the fall, initiated when day length drops below a critical threshold. Non-reproductive adults abandon all dispersal related activity, the production of the aggregation pheromone ceases, then all drop from their host tree into accumulated understory leaf litter to overwinter (Bean et al. 2007b).

Field deployment of a D. carinulata aggregation pheromone compound resulted in increased densities of the agent on host plants baited with aggregation pheromone lures when compared to untreated control plants (Gaffke et al. 2018a, 2018b). The deployment of a D. carinulata aggregation pheromone lure in conjunction with new releases could result in greater retention of the agent on release sites, reducing Allee effects (Hudgeons et al. 2007a). This could result in a strategy that would allow for scientists and biological control practitioners to maximize the number of releases made by making smaller releases while still maintaining a high probability of establishment. One hypothesized reason that pheromone deployment while making new releases will result in increased establishment is that successful colonization of new locations by field populations of D. carinulata is already governed by semiochemicals, chiefly the aggregation pheromones

(Cossé et. al 2005, 2006; Hudgeons et al. 2007a). If too little pheromone is being produced by males to maintain an aggregation during the initial release, it is possible that 124 the beetles will disperse and fail to establish (Carruthers et al. 2008). Not only could the strategy of applying the aggregation pheromone in conjunction with a new release result in greater persistence but it also has the potential for increased population growth by maintaining more of the individuals in the system. This study investigates the potential of an application of D. carinulata aggregation pheromone to increase establishment of populations after weed biological control agents are released.

Materials and Methods

Lures and Chemicals

Diorhabda carinulata aggregation pheromone, (2E,4Z)-2,4-heptadien-1-ol, was synthesized for the purposes of this study according to Petroski (2003). The single alcohol component of the pheromone was used for this study based on the reported greater capture rates of the alcohol component compared to the aldehyde component or a blend of the two (Cossé et al. 2005, Gaffke et al. 2018a, 2018b). Aggregation pheromone lures were prepared for field deployment by formulation into ISCA® technologies wax based formula SPLAT® (Specialized Pheromone & Lure Application Technology).

SPLAT is a technology commonly utilized for field deployment of pheromones, and has successfully delivered the same D. carinulata aggregation pheromone in the field (Gaffke et al. 2018a, 2018b). Two types of lures were prepared for field testing. Treatment lures consisted of SPLAT impregnated with the aggregation pheromone (2.17% active ingredient). Lures placed in control patches consisted of SPLAT without active ingredient. 125

The release rate of compounds from SPLAT has been well characterized and is controlled by the volume of the SPLAT applied (Mafra-Neto et al. 2013). Two doses of formulated aggregation pheromone have been studied for potential manipulation of D. carinulata in the field, 1-g and 4-g doses of SPLAT (Gaffke et al. 2018a, 2018b). A single 4-g application of SPLAT was used to investigate the potential for lures incorporating the D. carinulata aggregation pheromone to increase the likelihood of establishment of new releases of this agent. Release rates of the D. carinulata aggregation pheromone from lures based on 4-g dollops of SPLAT were reported in the same study (Gaffke et al. 2018b).

Insect Source

Lab colonies of D. carinulata were started from field-collected individuals obtained from a population located in southern Montana, USA. This population is believed to have originated from D. carinulata populations near the city of Lovell,

Wyoming, USA, that dispersed northward. Beetles in the original Lovell releases were initially sourced from Fukang, Xinjiang China (DeLoach et al 2003; Lewis et al. 2003b).

Beetle populations used in this study were propagated over the winter and spring. The greenhouse-based continuous culture of Diorhabda carinulata was sustained on potted

Tamarix plants enclosed in insect rearing sleeve cages (MegaView Science Co.,

LTD. No. 656-2, Fuya Rd., Taichung 40762 Taiwan). Tamarix plants used to mass rear

D. carinulata were sourced from the sites intended for future releases. Adult beetles from the greenhouse-based continuous culture that were unnecessary for propagation of the culture, were placed in a growth chamber with a photoperiod conducive for the initiation 126 of reproductive diapause. Diorhabda carinulata in reproductive diapause can be stored for several months, allowing for long-term storage of adult beetles until the occurrence of conditions appropriate to make field releases of founding individuals. Diapausing adults were removed from the growth chamber and placed in a greenhouse with a 16/8 hr day/night cycle for two weeks before field release. The adult beetles rapidly transitioned into the reproductive phase when exposed to 15 or more hours of daylight conditions

(Bean et al. 2007b). Beetles used in releases made in 2016 were sixth to eighth generation progeny of beetles field collected in 2015. Once releases were made in 2016, the lab colony was restarted with new field collected individuals. Beetle populations used in the

2017 releases similarly represent six to eight generations that were initiated from adults that were field collected in 2016.

Site Selection and Setup

Eighteen sites were selected across Montana, ten in 2016 and eight in 2017, for release of D. carinulata. Within each site, two patches of Tamarix were identified for releases. Patches were spaced a minimum of 250 m apart within the same study site.

Within each of the study patches, five plants were identified as permanent monitoring plants: a central release plant and four plants surrounding the central plant. Trees selected to receive the population of individuals released were located at the approximate center of each study patch and surrounded by the other monitoring plants. The four plants surrounding the central plant corresponded to cardinal directions and were located approximately 5-10 m from the central plant. Monitoring plants were also selected because they were adequately isolated from adjoining trees or other physical obstructions 127 to facilitate whole tree sweep sampling. At each of the study sites, one of the central release plants was treated with a 4-g dose of pheromone impregnated SPLAT, the other paired central release plant was treated with 4-g dose of SPLAT with no active ingredient.

Field-Release Protocol

Releases consisted of newly eclosed reproductive adults from the greenhouse- based continuous culture, and ‘stored’ adults manipulated by exposure to longer day length to break reproductive diapause and transition into an active reproductive state

(Bean et al. 2007a). All reproductive D. carinulata adults were pooled before field release to randomize against potential effects of progeny age from the continuous lab cultures on the overall success of individual releases. Mass rearing efforts in 2016 produced 3,000 adult D. carinulata for field release. Field releases, each consisting of

300 individuals, were made at ten locations between 19-21 July 2016. In 2017, mass rearing yielded 4,000 individuals for field releases. Five hundred individuals were released on 19 or 21 June 2017 at each of eight locations. All releases were made with slightly chilled adults on dense foliage that was easily accessed. A 240-ml paper container holding adults was supported in the branches of the release plant and the lid was removed to allow the adults to exit and crawl into the canopy of the plant. This protocol was used to minimize immediate disturbance and rapid dispersal of the beetles from the plant targeted to receive the insects. Adults were released only on the central plant and were not distributed throughout the patch.

128

Insect Sampling and Population Survey

Sweep net and direct counts were used to assess D. carinulata abundance. The five monitoring plants in each study patch were swept using five 1-m passes through the foliage. After each plant was swept, the contents of the net were examined and D. carinulata present were tallied. Adults are commonly observed on foliage towards the interior of the canopy. Adults in these interior positions are difficult to sample with a sweep net, therefore in 2017, direct counts of D. carinulata were added to the protocol to minimize the possibility of false negatives. Direct counts were conducted for 3 min on each of the central release plants.

One week after the releases were made, sites were assessed for the presence of D. carinulata. This post-release survey was conducted to determine if failure to establish resulted from rapid dispersal from the release site or a more gradual decline in abundance over time in resident beetle populations. Subsequent surveys following the initial post- release survey were made at the time of emergence of subsequent generations of D. carinulata and were conducted to assess the persistence of individuals from new releases.

Statistical Analysis

The effect of the aggregation pheromone on immediate dispersal of D. carinulata from the release patch was analyzed using the Wilcoxon rank-sum test. The rank-sum test was utilized due to the small sample size and highly variable data. The effect of the aggregation pheromone on persistence of a new release of D. carinulata was analyzed using a generalized linear regression with a Poisson distribution. The Poisson distribution 129 was used because it is a distribution commonly used for count data, and is the most appropriate for counts of rare events.

Results

Immediate Dispersal

In 2016, two weeks after initial releases of 300 adults were made at each of the ten eastern Montana study patches, D. carinulata was detected by sweep net at four of the five release sites with a pheromone lure. Adult and larval D. carinulata were found together on two of the five release sites with a pheromone lure. No D. carinulata were detected at sites where a release was made without a pheromone lure. Sites where releases with a pheromone lure were made averaged 1.8 individuals per 25 sweeps compared to 0.0 individuals at release sites without a pheromone lure (Wilcoxon rank sum test, P= 0.02) (Fig 1a).

In 2017, sweep sampling yielded no difference in D. carinulata density between release sites with a pheromone lure and releases without a pheromone lure (Wilcoxon rank sum test, P=0.5) (Fig 1c). Release sites without a pheromone lure averaged 0.25 individuals per 25 sweeps while release sites with a pheromone lure averaged two individuals. Additional monitoring was incorporated in 2017 to better account for possible discrepancies between observed and sweep based densities of D. carinulata at the release sites. Ten days after the release of 500 adult D. carinulata, an average of eight individuals was detected during 3-min counts on the central release plant at sites with a pheromone lure (Fig 1b). This average was significantly larger than the average detection 130 of 0.5 individuals on the central release plant at release sites without a pheromone lure

(Wilcoxon rank sum test, P=0.05).

Persistence of Individuals at a Release

Following the release of 300 individuals on each of ten sites in 2016, establishment beyond the first generation was confirmed at only one site. Across the four generations that were monitored, D. carinulata was found only immediately after the release (designated by the 0 in Figure 2), and during the 3rd and 4th generation after release (Fig 2). There was no detectable effect of the pheromone on the establishment and persistence of D. carinulata populations originating from releases of 300 individuals

(P=0.99). No D. carinulata were detected while monitoring the releases without a pheromone lure. Three of the releases receiving 300 adults (two releases with a pheromone lure and one release without a pheromone lure) were lost to herbicide application after monitoring the 1st generation.

The release of 500 individuals in 2017 did result in differences in detection at release sites with a pheromone lure and those without a pheromone lure (Table 1).

Specifically, when the plants at release sites with a pheromone lure were swept there was an average of 1.86 more adults than at sites without a pheromone lure (P=0.002) (Fig 3a).

The number of generations passed after the initial release was made was also a significant factor, with an average reduction of 0.59 individuals for each additional generation

(P=0.03). When the central plant was visually inspected, the release sites with a pheromone lure had an average of 1.84 individuals more than the sites without a pheromone lure (P<0.001) (Fig 3b). The number of generations was also a significant 131 factor when the plants were visually inspected. There was an average reduction of 0.86 individuals for each additional generation (P<0.001).

Discussion

The purpose of this study was to elucidate the effects that D. carinulata aggregation pheromone have on potential negative outcomes when new releases are made. Diorhabda carinulata is known to rapidly disperse away from release areas, resulting in populations vulnerable to Allee effects (Carruthers et al. 2008). This study demonstrates that D. carinulata adults can be retained at release sites by application of aggregation pheromone. The effects of the application of the aggregation pheromone were apparent in future generations, with more individuals detected at release sites with the aggregation pheromone lure compared to those without. This suggests that a single application of the aggregation pheromone during the initial release can have ongoing beneficial effects and impact future population growth at the release site. Increased population growth will increase the likelihood that newly released individuals will establish (Grevstad 1999; Dennis 1989). The application of aggregation pheromone of a biological control agent at the time of a new release may provide a tool for scientists and biological control practitioners to limit dispersal of biological control agents from release locations, thereby reducing Allee effects, increasing population growth, and subsequently increasing the likelihood of long term population establishment.

Releases of 5,000 D. elongata are recommended to optimize the likelihood of establishment (Knutson and Muegge 2008). The release numbers used in this study are 132 significantly smaller than this recommendation. Because the releases made for this study represent 6-10% of the number recommended, it is surprising that any D. carinulata were detected after release (Carruthers et al. 2008). The results of this study support observations made by Hudgeons et al. (2007a), documenting an increased probability that

D. elongata would be detected not with larger release sizes but rather with the presence of conspecifics. Initial releases of 200 adults that were made with caged populations nearby resulted in subsequent detection of released individuals, presumably due to the presence of the aggregation pheromone from the caged individuals. Releases of as many as 1600 individuals that were made with no conspecifics in the vicinity resulted in an inability to confirm establishment of those releases. The negative relationship between D. carinulata density and generation number that we observed suggests that numbers of individuals used in releases for this study are not sufficient to result in persistent, long- term establishment of the agent. Factors affecting the establishment of D. carinulata when release sizes this small are used are unlikely to be overcome with the application of the aggregation pheromone. However, the combination of the aggregation pheromone and greater numbers of individuals in releases could result in significantly higher rates of establishment than without the pheromone. This research suggests that the aggregation pheromone can keep individuals in a targeted area (Gaffke et al. 2018a, 2018b).

Cages prevent initial Allee effects at release sites by restricting the dispersal capacity of the agent. Multiple releases in cages have been made with several thousand

D. carinulata in Montana without successful establishment (Deloach et al. 2011). Several caged release sites were used during this current study (see Supplementary material). The 133 results were similar to those achieved in previous attempts with no D. carinulata confirmed as established in the field cages (Table S1). Due to confirmation of the reported failures of D. carinulata in field cages in Montana, no further caged trials were conducted during the present study.

The observed dispersal pattern of D. carinulata follows a characteristic pattern.

The adults disperse short distances within a Tamarix stand until the resources are depleted (Bean et al. 2013). Once the stand is depleted, D. carinulata adults initiate long distance dispersal to find suitable hosts (DeLoach et al. 2004; Hudgeons et al. 2007a).

Adult males subsequently disperse from the stand, alight on a new patch of saltcedar and begin to feed before producing the aggregation pheromone (Cossé et al. 2005). Other D. carinulata moving through the area are attracted by the aggregation pheromone. These events culminate in the formation of an aggregation at the new location, resulting in reproduction and subsequent establishment of a new population.

Diorhabda carinulata males vary emission of the aggregation pheromone, which is produced in conjunction with feeding (Cossé et al. 2005). Insects are commonly kept in small, dark containers with foliage before a new release is made. The insects to be released are typically chilled to reduce activity. It is likely that males within these containers are not producing aggregation pheromones. Once released, the males need to first warm, become active, and initiate feeding before they begin to release aggregation pheromones (Cossé et al. 2005). If newly released males start to disperse rather than feed, then aggregation pheromone will not be produced to retain the adults at the release site, consequently making the population highly susceptible to Allee effects. These issues may 134 potentially be overcome through the strategic deployment of D. carinulata aggregation pheromone. Applications of 4-g dollops of pheromone impregnated SPLAT emit large amounts of the aggregation pheromone, which quickly saturates the area surrounding released individuals and thereby prevents dispersal.

It is also possible that the treatment of only a single plant with the aggregation pheromone may produce antagonistic interactions. Many chrysomelid beetles are known to be repelled by conspecific larvae (Schindek and Hilker 1996; Gross and Hilker 1994;

Hilker 1989). It is possible that the purposefully manipulated long term aggregation of D. carinulata on specific plants receiving released individuals could result in reduced population growth due to the antagonistic interaction between the presence of larvae and the aggregation pheromone emitted by a lure. This potential antagonistic interaction was not specifically tested for in the current study, but no antagonistic response was detected.

However, the efficacy of the pheromone treatment could be increased simply by treating more plants around release sites to minimize interactions between adults and newly hatched larvae.

Many biological and environmental parameters need to be assessed to identify optimal species-specific conditions consistently leading to successful and persistent establishment of biological control agents (Grevstad et al. 2011). Biological factors such as propagule pressure and life history traits can be significant drivers in the success of colonization (Simberloff 2009; Rosecchi et al. 2001). Other studies have demonstrated that factors such as species-specific site characteristics are the primary divers of colonization success (Duncan 2016). The results of this study illustrate the importance of 135 considering whether pheromones influencing the presence and intensity of Allee effects during the release of biological control agents are likely to play a significant role in the ability of the agent to establish. If individuals are not producing pheromones at the time of release as they typically would otherwise, detrimental Allee effects could increase and result in extinction of the population. This could be readily overcome through the application of aggregation pheromone. This study demonstrated that newly released individuals will persist at release locations via the application of aggregation pheromone.

Higher densities of insects that are achieved through retention of individual agents at the release site can result in larger populations in subsequent generations. This will ultimately increase the likelihood of establishment.

Acknowledgements

Many people provided critical support for this project. A special thanks must be provided to Patricia Gilbert, Amy Adler, and Jennifer Kramer for identifying release locations. Thanks also to the Montana Saltcedar Team for assisting in coordination of the project and helping to find additional collaborators. Megan Hofland and Norma Irish offered helpful advice on the mass rearing process. Funding for this project was provided through the USDA Forest Service’s Forest Health Technology and Enterprise Team -

Biological Control of Invasive Native and Non-Native Plants (FHTET-BCIP) competitive grant program.

136

a) 2016 b) 2017

14 14 b

12 12

10 10

8 Control 8 Control PH PH

6 6

Mean Density Mean MeanDensity 4 4 b 2 2 a a 0 0

c) 2017

8 Control PH

6 MeanDesnity 4 a

2 a 0

Figure 1: Sampling for D. carinulata adults for immediate dispersal 10-14 days after the initial release. Mean adult density ± SE per 25 sweeps (a, c) and mean adult density ± SE per 3 min count (b) of D. carinulata at release sites. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control).

137

4 Control 3

mean mean density PH 2

0 0 1 2 3 4 generation

Figure 2: Sampling for mean adult density ± SE per 25 sweeps for release persistence after the initial release. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control). Post-release monitoring at 10-14 days after the initial release is designated as generation 0 on x-axis.

138

a) b)

7 16

6 14 12 5 10 4 8 3 Control Control

6 MeanDensity 2 PH MeanDensitty PH 4 1 2 0 0 0 1 2 0 1 2 generation generation

Figure 3: Sampling for D. carinulata adults for release persistence after the initial release. Mean adult density ± SE per 25 sweeps (a) and mean adult density ± SE per 3 min count (b) of D. carinulata at release sites. Releases were made with a pheromone lure (PH) or without a pheromone lure (Control). Post-release monitoring at 10-14 days after the initial release designated as generation 0 on x-axis.

139

Table 1: Poisson regression analysis for mean densities per 20 sweeps, and mean counts per 3 mins

Release method parameter Coefficient Standard Z value P size error

300 sweep treatment 19.32 1923.4 0.010 0.99 generation 0.038 0.12 0.25 0.80 500 sweep treatment 1.87 0.61 3.01 0.002 generation -0.59 0.27 -2.17 0.03

count treatment 1.85 0.36 5.15 <0.001 generation -0.87 0.18 -4.83 <0.001

140

Supplemental Material

Material and methods

Caged release

Six field cages were set up in 2016 to act as positive controls and reduce potential

Allee effects on new releases. Five of the field cages received releases of 200 adults, while the last field cage received 500 adults. Release sizes were made based on the number of beetles available. Caged releases were made on June 17, June 23, and July 19, and July 21.

Results

Caged releases

Caged releases of reproductive adult D. carinulata were made in 2016 at three locations: Fort Peck, Forsyth, and Hysham MT, USA. Fort Peck received one caged release of 500 adults, while Forsyth received two caged rereleases of 200 adults and

Hysham received three caged releases of 200 adults (Table 1). The caged release at Fort

Peck resulted in extensive defoliation of the Tamarix within the cage. Not only did the larvae defoliate the Tamarix plants inside the cage, but they were able to escape from the cage and completely defoliated adjacent Tamarix plants. Approximately 40% of the plants were defoliated within the two cages located at Forsyth, with D. carinulata adults and larvae present during the September survey. Larvae were present in Hysham cage 3 and had defoliated approximately 40% of the plants within the cage. Plants contained 141 within Hysham cages 1 and 2 were defoliated by the tamarisk leafhopper, Opsius stactigalus (Hemiptera: Cicadellidae), and did not show any evidence of D. carinulata in the September survey.

In 2017, all of the caged releases set up in 2016 were surveyed for establishment of D. carinulata using aggregation pheromone as a lure on sentinel plants. Locations were surveyed three times throughout 2017. During the initial surveys in June, five of the six cages were observed to be submerged under several cm of water from the adjacent

Yellowstone River. No evidence of flooding was observed at the Fort Peck caged release.

At no point throughout the summer of 2017 was evidence of D. carinulata detected at any of the caged release locations.

142

Table S1: Number of adult D. carinulata released and detected within caged releases in 2016. Caged releases lost to O. stactogalus damage are designated with (---).

Location Date June 17,23 July 19, 20,21 August 3,4 September 14,15 Adults Larva Adult Larva Adult Larva Adult Larva e Hysham 200 0 2 0 ------Cage 1 Hysham 200 0 3 3 0 5 ------Cage 2 Hysham 200 0 50 0 0 15 Cage 3 Forsyth 200 0 10 7 0 8 3 2 Cage 1 Forsyth 200 0 25 9 0 6 1 1 Cage 2 Fort Peck 500 0 100 15 0 24 Cage

143

References:

Bartelt, R.J., Cossé, A.A., Zilkowski, B.W., Weisleder, D. and Momany, F.A., 2001. Male-specific sesquiterpenes from Phyllotreta and Aphthona flea beetles. Journal of Chemical Ecology, 27: 2397-2423.

Berec, L., Angulo, E. and Courchamp, F. 2007. Multiple Allee effects and population management. Trends in Ecology & Evolution, 22: 185-191.

Berggren, Å. 2001. Colonization success in Roesel's bush cricket Matrioptera roeseli: the effects of propagule size. Ecology 82:274‐280.

Western US: Emphasis Texas-1986-2011. XIII International Symposium on Biological Control of Weeds. Kona, HI: 268-276

Dennis, B. 2002. Allee effects in stochastic populations. Oikos, 96: 389-401.

Duncan, R.P. 2016. How propagule size and environmental suitability jointly determine establishment success: a test using dung beetle introductions. Biological invasions, 18: 985-996.

Duncan, R.P., Blackburn, T.M., Rossinelli, S. and Bacher, S. 2014. Quantifying invasion risk: the relationship between establishment probability and founding population size. Methods in Ecology and Evolution, 5: 1255-1263.

Fauvergue, X., 2013. A review of mate‐finding Allee effects in insects: from individual behavior to population management. Entomologia Experimentalis et Applicata, 146: 79-92.

Fauvergue, X., Vercken, E., Malausa, T. and Hufbauer, R.A. 2012. The biology of small, introduced populations, with special reference to biological control. Evolutionary applications 5: 424-443.

Fernandez, P. and Hilker, M. 2007. Host plant location by Chrysomelidae. Basic and Applied Ecology, 8: 97-116.

Freckleton, R.P. 2000. Biological control as a learning process. Trends in Ecology & Evolution, 15: 263-264.

145

Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Weaver, D.K. 2018b. Field demonstration of a semiochemical treatment that enhances Diorhabda carinulata biological control of Tamarix spp.. In Review

Gascoigne, J., Berec, L., Gregory, S. and Courchamp, F. 2009. Dangerously few liaisons: a review of mate-finding Allee effects. Population Ecology, 51: 355-372.

Grevstad F.S., Coombs, E.M. and McEvoy, P.B., 2011. Revisiting release strategies in biological control of weeds: are we using enough releases?. In Proceedings of the XIII International Symposium on Biological Control of Weeds, Waikoloa, United States, 11-16 September 2011, pp 368-376).

Grevstad, F. S. 1999. Experimental invasions using biological control introductions: the influence of release size on the chance of population establishment. Biological Invasions, 1: 313-323.

Grevstad, F.S. and A.L. Herzig. 1997. Quantifying the effects of distance and conspecifics on colonization: experiments and models using the loosestrife leaf beetle, Galerucella calmariensis. Oecologia, 110: 60-68.

Grevstad, F.S., Strong, D.R., Garcia-Rossi, D., Switzer, R.W. and Wecker, M.S. 2003. Biological control of Spartina alterniflora in Willapa Bay, Washington using the planthopper Prokelisia marginata: agent specificity and early results. Biological Control, 27: 32-42.

Gross, J. and Hilker, M. 1994. Chemoecological studies of the exocrine glandular larval secretions of two chrysomelid species (Coleoptera): Phaedon cochleariae and Chrysomela lapponica. Chemoecology, 5: 185-189.

Hambäck, P.A. 2010. Density-dependent processes in leaf beetles feeding on purple loosestrife: aggregative behaviour affecting individual growth rates. Bulletin of Entomological Research 100: 605-611.

Hilker, M. 1989. Intra‐and interspecific effects of larval secretions in some chrysomelids (Coleoptera). Entomologia Experimentalis et Applicata, 53: 237-245. 146

Hoelmer, K.A. and Kirk, A.A. 2005. Selecting arthropod biological control agents against arthropod pests: Can the science be improved to decrease the risk of releasing ineffective agents?. Biological control, 34: 255-264

Hopper, K.R. and Roush, R.T. 1993. Mate finding, dispersal, number released, and the success of biological control introductions. Ecological entomology, 18: .321-331.

Hudgeons, J.L., Knutson, A.E., DeLoach, C.J., Heinz, K.M., McGinty, W.A. and Tracy, J.L. 2007a. Establishment and Biological Success of Diorhabda elongata elongata on Invasive Tamarix in Texas. Southwestern Entomologist, 32: 157-168.

Landolt, P.J., 1997. Sex attractant and aggregation pheromones of male phytophagous insects. American Entomologist, 43:12-22.

Memmott, J., Fowler, S.V. and Hill, R.L. 1998. The Effect of Release Size on the Probability of Establishment of Biological Control Agents: Gorse Thrips (Sericothrips staphylinus) Released Against Gorse (Ulex europaeus) in New Zealand. Biocontrol Science and Technology, 8: 103-115.

Memmott, J., Fowler, S.V., Harman, H.M., Hayes, L.M., Moran, V.C. and Hoffmann, J.H. 1996. How best to release a biological control agent. Proceedings of the 9th international symposium on biological control of weeds, Stellenbosch, South Africa, 19-26 January 1996. University of Cape Town: 291-296

Peterson, A.T. 2003. Predicting the geography of species’ invasions via ecological niche modeling. The Quarterly Review of Biology, 78: 419-433.

Petroski, R.J. 2003. Straightforward preparation of (2 E, 4 Z)-2, 4-heptadien-1-ol and (2 E, 4 Z)-2, 4-heptadienal. Synthetic Communications 33: 3233-3241. 147

Raffa, K.F., Phillips, T.W. and Salom, S.M. 1993. Strategies and mechanisms of host colonization by bark beetles. Beetle-pathogen interactions in conifer forests. Academic Press, London: 102-28.

Robertson, M.P., Kriticos, D.J. and Zachariades, C. 2008. Climate matching techniques to narrow the search for biological control agents. Biological control, 46: 442- 452.

Robinson, T.W. 1965. Introduction, spread, and areal extent of saltcedar (Tamarix) in the western states. Geological Survey Professional Paper, Contact No. 491-A.

Rosecchi, E., Thomas, F. and Crivelli, A.J., 2001. Can life‐history traits predict the fate of introduced species? A case study on two cyprinid fish in southern France. Freshwater Biology, 46: 845-853.

Schindek, R. and Hilker, M. 1996. Influence of larvae of Gastrophysa viridula on the distribution of conspecific adults in the field. Ecological Entomology, 21:370- 376.

Shea, K. and Possingham, H.P. 2000. Optimal release strategies for biological control agents: an application of stochastic dynamic programming to population management. Journal of Applied Ecology, 37: 77-86.

Simberloff, D. 2009. The role of propagule pressure in biological invasions. Annual Review of Ecology, Evolution, and Systematics, 40: 81-102.

Syrett, P., Briese, D.T. and Hoffmann, J.H. 2000. Success in biological control of terrestrial weeds by arthropods. Biological control: measures of success. Springer Science and Business Media Dordrecht, Netherlands: 189-230.

Tansey, J.A., McClay, A.S., Cole, D.E. and Keddie, B.A., 2005. Evidence for the influence of conspecific chemical cues on Aphthona nigriscutis (Coleoptera: Chrysomelidae) behaviour and distribution. BioControl 50: 343-358.

Turchin, P. and Kareiva, P., 1989. Aggregation in Aphis varians: an effective strategy for reducing predation risk. Ecology, 70: 1008-1016.

Winston, R., Randall, C.B., De Clerck-Floate, R., McClay, A., Andreas, J., Schwarzländer, M., 2014. Field guide for the biological control of weeds in the 148

Northwest. FHTET-2014-08. Moscow, ID: USDA Forest Service and University of Idaho Extension.

Wood, D.L., 1982. The role of pheromones, kairomones, and allomones in the host selection and colonization behavior of bark beetles. Annual Review of Entomology, 27:411-446.

Zavaleta, E. 2000. The economic value of controlling an invasive shrub. AMBIO: a Journal of the Human Environment, 29: 462-467.

149

CHAPTER FIVE

BEHAVIORAL RESPONSES OF DIORHABDA CARINULATA ADULTS TO

REPELLENT COMPOUNDS ISOLATED AND IDENTIFIED IN VOLATILES

COLLECTED FROM CONSPECIFIC LARVAE

Contribution of Authors and Co-Authors

Manuscript in Chapter 5

Author: Alexander M. Gaffke

Contributions: Designed and implemented experiments, collected, processed and analyzed data, and prepared manuscript for publication

Co-Author: Sharlene E. Sing

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Jocelyn Millar

Contributions: Developed methodology for chemical synthesis

Co-Author: Tom L. Dudley

Contributions: Provided input on experimental design, and manuscript preparation

Co-Author: Daniel W. Bean

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

Co-Author: Robert K. D. Peterson

Contributions: Provided input on result interpretation and manuscript preparation

150

Contribution of Authors and Co-Authors Continued

Co-Author: David K. Weaver

Contributions: Developed experimental objectives, and provided input on experimental design, implementation, data analysis and manuscript preparation

151

Manuscript Information Page

Alexander M. Gaffke, Sharlene E. Sing, Jocelyn Millar, Tom L. Dudley, Daniel W. Bean, Robert K. D. Peterson, David K. Weaver

Journal of Chemical Ecology

152

BEHAVIORAL RESPONSES OF DIORHABDA CARINULATA ADULTS TO

REPELLENT COMPOUNDS ISOLATED AND IDENTIFIED IN VOLATILES

COLLECTED FROM CONSPECIFIC LARVAE

Alexander M. Gaffkea, Sharlene E. Singb, Jocelyn G. Millarc, Tom L. Dudleyd, Daniel W. Beane, Robert K. D. Petersona, David K. Weavera* aDepartment of Land Resources and Environmental Sciences, Montana State University,

Bozeman MT, 59717, USA bUSDA Forest Service, Rocky Mountain Research Station, Bozeman MT, 59717, USA cDepartment of Entomology, University of California, Riverside, CA 92521, USA dMarine Science Institute, University of California, Santa Barbara CA 93106, USA eColorado Department of Agriculture, Palisade Insectary, Palisade CO, 81526, USA

*Correspondence to: David K. Weaver, Montana State University, P.O. Box 173120,

Bozeman, MT, 59717. Email: [email protected]

153

Abstract

The leaf beetle Diorhabda carinulata Desbrochers (Coleoptera: Chrysomelidae) was first introduced to the United States in 1999 as a classical biological control agent for the invasive weed saltcedar (Tamarix spp.). Multiple aggregation-causing semiochemicals have been identified for this species. We report on a new volatile compound produced by conspecific larvae that has repellent activity against adult D. carinulata. Collection of volatile aerations, identification of candidate compounds and subsequent electroantennography suggested a behaviorally active compound. Behavioral assays were conducted to test for repellency in adult beetles. Bioassays indicated unknown compound(s) emanating from larval hemolymph and in aerations of groups of larvae that were repellent to reproductive adults, but adults in reproductive diapause did not respond. The compound(s) present in the larval hemolymph which were repellent to adults could not be isolated or identified. However, individual larvae produced an average of 3.89 ng of the repellent compound 4-oxo-(E)-2-hexenal over a three-day volatile collection period. 4-oxo-(E)-2-hexenal elicited electrophysiological and behavioral responses due to repellency across multiple doses in reproductive D. carinulata adults. Reproductive females were repelled at lower doses of 4-oxo-(E)-2-hexenal than reproductive males.

Key words: EAG, Olfactometer, Chrysomelidae, biological control, Tamarix, southwestern willow flycatcher 154

Introduction

Negative ecological impacts of Tamarix spp. (hereafter referred to as Tamarix) invasions of North America have been numerous and widespread (Deloach 1991).

Tamarix has invaded riparian ecosystems and now occupies up to 650,000 hectares since its intentional introduction to the United States in the early 1800s (Jarnevich et al. 2013).

Tamarix is considered to be the third most prevalent woody riparian plant in the western

United States (Friedman et al. 2005, Mcshane et al. 2015).

Tamarix is well adapted to thrive under a wide range of environmental conditions.

It can withstand long term drought, inundation, saline environments, and frequent burning (Deloach 1991). The establishment and expansion of Tamarix is also promoted by many anthropogenic habitat alterations. Such broad ecological amplitude allows

Tamarix to outcompete most native woody plants, especially in the presence of altered fire and flooding regimes (Yapp 1922; Busch 1995; Schlatter et al. 2017). Although the displacement of native plant species is promoted by anthropogenic habitat alterations,

Tamarix is also rapidly spreading into pristine, unaltered habitat in Montana and North

Dakota (Pearce and Smith 2003; Pearce and Smith 2007). Tamarix dominance of plant communities greatly degrades wildlife habitat. Tamarix monocultures are correlated with reduced assemblages of insects, fish, rodents, and birds (DeLoach et al. 2003; Kennedy et al. 2005; Hildebrandt and Ohmart 1982; Dudley and DeLoach 2004).

A biological control program was initiated against Tamarix due to its widespread distribution and attributed role as a source or facilitator of extensive environmental damage (DeLoach 2003). The leaf beetle Diorhabda carinulata Desbrochers 155

(Coleoptera: Chrysomelidae), formerly Diorhabda elongata deserticola Chen (Tracy and

Robbins 2009), was screened, approved, and released in the United States as a biological control agent for Tamarix (DeLoach 2003, USDA-APHIS 1999). Both adult and larval D. carinulata feed on Tamarix, rasping and scraping the leaves and photosynthetic stems

(Dudley and Kazmer 2005). These rasping and scraping feeding activities compromise the water regulating properties of the epidermis (Dudley and Kazmer 2005). Significant feeding damage causes affected tissues to desiccate and die (Snyder et al. 2010). Tamarix foliar desiccation and death is considered a defoliation event, despite limited actual biomass removal through phytophagy. Defoliation causes extensive stress to the plants, reducing non-structural root crown carbohydrate reserves, decreasing leaf production, and decreasing root mass (Hultine et al. 2011; Hudgeons et al. 2007b). While a modest level of defoliation (30%) can reduce annual stem and root growth, large defoliations

(>80%) can result in significant dieback and even mortality (Gaffke et al. 2018a, 2018b).

The high reproductive capacity of D. carinulata, and its ability to form natural aggregations in the field through the use of an aggregation pheromone has resulted in one of the most effective weed biological control programs in the United States (Lewis et al.

2003b; Cossé et al. 2005, 2006; Dudley and Bean 2012). Diorhabda carinulata has spread widely in the United States and caused continuous years of defoliation which have resulted in a reduction of plant volume by 93%, and a 40% morality rate in some areas

(Pattison et al. 2011; Dudley et al. 2006). Thousands of hectares have been defoliated multiple times over the course of a single season with this multivoltine beetle (Pattison et al. 2011). 156

The distribution of D. carinulata in the United States was initially expected to be more northerly. The beetles were believed to be incapable of establishing south of the

38th North parallel due to physiological constraints arising from a day length controlled diapause (Bean et al. 2007a). Specifically, when the photoperiod drops below 14 hr and

39 min, the insects cease to reproduce, develop fat stores, and drop into the soil to overwinter as adults (Bean et al. 2007a; Bean et al. 2012). South of the 38th North parallel, day length never exceeds the critical photoperiod that allows for reproduction. Adult D. carinulata enter diapause prematurely and are killed from winter conditions (Bean et al. 2007a). However, since being introduced, D. carinulata has rapidly evolved and the critical photoperiod inducing diapause is no longer limiting (Bean et al. 2012). Diorhabda carinulata now occupies vast areas south of the 38th

North parallel and is successfully defoliating and impacting Tamarix in these areas.

Although southward expansion of the range of D. carinulata may have been welcomed for Tamarix management, these burgeoning, defoliating populations have now brought D. carinulata into conflict with habitat needed by the endangered southwestern willow flycatcher (Empidonax traillii extimus). The southwestern willow flycatcher was listed as an endangered species in 1995 and in 2005 Tamarix was listed as a critical habitat component for the species (US Fish and Wildlife Service 1995; Federal Register

2005). The southwestern willow flycatcher occupies riparian areas south of the 38th North parallel, so interaction between D. carinulata and this flycatcher was not anticipated due to initial non-overlapping distributions within the United States. Post-release evolution and subsequent expansion of D. carinulata into the range of the southwestern willow flycatcher raised concerns that D. carinulata could negatively impact the survival of this 157 avian species. Rapid defoliation of Tamarix would result in the loss of adequate canopy cover for southwestern willow fly catcher nesting sites. In addition to concerns over the specific loss of Tamarix canopy cover for nesting habitat, the rapid and repeated nature of

D. carinulata defoliation could kill all or most Tamarix plants before native plant species recovery was advanced enough to provide suitable habitat to replace large barren stands of dead Tamarix. These detrimental interactions with the beetle could worsen the situation for the endangered flycatcher (Sogge et al. 2008).

It has recently been demonstrated that populations of D. carinulata can be directed to and retained in targeted areas through the field deployment of lures emitting the aggregation pheromone produced primarily by adult males (Gaffke et al. 2018a,

2018b). The application of the aggregation pheromone within the shared range of D. carinulata and the southwestern willow flycatcher could be used to temporarily minimize interaction between these two species. Using a series of strategically placed pheromone lures to concentrate aggregations in specific locations could potentially divert beetles from forming aggregations on Tamarix with existing nests. However, as beetle populations locally increase, all Tamarix in an area are likely be defoliated regardless of semiochemical treatment (Gaffke et al. 2018a, 2018b). Even minimal levels of defoliation could prove detrimental to the southwestern willow flycatcher. Identification and strategic deployment of a repellent compound that could provide a long term deterrent from the formation of aggregations by D. carinulata could significantly minimize harmful interactions between the biological control agent and the endangered southwestern willow flycatcher. 158

Multiple studies report that defensive compounds produced by larval conspecifics are repellent to adults (Hilker and Weitzel 1991; Rostas and Hilker 2002; Nufio and

Papaj 2001). Larval defensive secretions from the chrysomelid species Gastrophysa viridula, Phaedon cochleariae, Plagiodera versicolora, and Phratora vitellinae were shown to act as oviposition and feeding deterrents against adult conspecifics (Hilker

1989). While larvae of D. carinulata do not appear to possess exocrine glands for defensive secretions, reflex bleeding is a prominent characteristic in the Galerucinae

(Deroe and Pasteels 1977, 1982; Wallace and Blum 1971), including the genus

Diorhabda, and both the adults and larvae reflex bleed when disturbed. Reflex bleeding, or autohemorrhaging, is the exposure of hemolymph and any toxic compound contained within, to the outside environment through integumental ruptures (Stocks 2008). The chemistry of D. carinulata hemolymph has not been investigated but the Galerucinae are known to accumulate deterrent substances in the hemolymph for release during reflex bleeding (Pasteels et al. 1988). These deterrent substances provide an anti-predation mechanism but could also provide important behavioral cues to conspecifics. The conspecific activity of defensive secretions has been reported in the related subfamily

Chrysomelinae (Schindek and Hilker 1996; Gross and Hilker 1994; Hilker 1989).

The objective of this study is to identify potential repellent compounds produced by D. carinulata larvae. Any biologically active repellent chemicals produced constitutively or in the hemolymph while reflex bleeding are of interest and could play a critical role in protecting avian species that could be negatively affected by defoliation due to D. carinulata herbivory. Any “push” effect that repels D. carinulata adults could, 159 whether applied alone or in conjunction with the aggregation pheromone to produce a

“push-pull” system, provide a critical management tool that could help conserve habitat for the southwestern willow flycatcher and minimize detrimental interaction between D. carinulata and this avian species.

Methods and material

Insects

Diorhabda carinulata were sourced from wild populations established in Tamarix stands outside of Lovell, Wyoming and from a continuous laboratory culture initiated from adult beetles field-collected in south-central Montana. Individuals from both sources are progeny from established, intentional releases of beetles introduced from

Fukang, Xinjiang China (DeLoach et al. 2003; Lewis et al. 2003b). A small cohort of adults were dissected before experimentation to determine reproductive status. The condition of the ovaries, ovarioles, and accessory glands indicated whether individuals were in a reproductive state, or were non-reproductive and entering diapause (Bean et al.

2007b). Lab reared individuals used for behavioral trials were progeny of 12-16 generations of continuous laboratory culture.

Hemolymph Collection

Bulk collection of larval hemolymph from D. carinulata was obtained from individuals in the third stadium. Larvae were induced to reflex bleed by agitating with wide-point featherweight forceps. Larvae were grasped along the length of their abdomen with the tips of the forceps extending to just posterior of the head. Pressure was applied 160 until the onset of reflex bleeding (Fig S1). The larval hemolymph was then collected by touching the head of the larva and the subsequent drop of hemolymph to the side of a 1.5 ml GC-MS vial insert. Vials were then capped and stored at -30 °C until sample preparation. Hemolymph produced by third instars while reflex bleeding in response to physical agitation is hereafter referred to as ‘larval hemolymph’.

Volatile Collection and Analysis

Volatiles were collected from both the larval hemolymph and from larvae.

Volatiles were collected from 10 µl (approximately five third instar equivalents) aliquots of the bulk collected larval hemolymph that had been applied to filter paper. Volatile collections from larval hemolymph were made in glass volatile collection chambers

(inner diameter 35 mm, length 625 mm) over a two-hour period. Volatile collections from intact feeding larvae were conducted in a greenhouse using glass volatile collection chambers (inner diameter 95 mm, length 625 mm) that contained 20 third instars and live

Tamarix. Volatiles were also collected from live Tamarix plants and from mechanically wounded Tamarix plants. The undamaged and damaged Tamarix plants were used as a reference source for plant produced volatiles. Mechanical wounding consisted of removing approximately 25% of the apical growing points from the plants with scissors.

All Tamarix plants had approximate volumes of 0.33 m3. Volatile collections involving live larvae or live Tamarix plants were conducted over a 3-day period.

Volatile collection traps contained 30mg of super-Q (Alltech Associates, Inc.,

Deerfield, IL, USA) adsorbent and were fixed in place at the apical opening of collection chambers. Purified air was supplied to chambers at a rate of 100 mlmin-1. Collected 161 volatiles were eluted from the traps into vials using methylene chloride (200 µl) and the samples were spiked with 10 µl of a 0.84 ngµl-1 solution of 1-octanol in methylene chloride as internal standard. Volatiles were analyzed using a GC (Agilent

6890instrument; Agilent Technologies, Santa Clara, California) coupled to a mass selective detector (MSD, Agilent 5973 instrument; Agilent Technologies, Santa Clara,

California).

Preparation of 4-oxo-(E)-2-hexenal for Bioassays

4-oxo-(E)-2-hexenal, a common component of the defensive sections of true bugs, and the primary component of the volatiles emitted from larval D. carinulata feeding on Tamarix, was evaluated for repellency to adult D. carinulata. 4-oxo-(E)-2- hexenal was synthesized according to Moreira and Millar (2005) and prepared for bioassays and electrophysiological experiments. Aliquots of the 4-oxo-(E)-2-hexenal solution were diluted with dichloromethane to concentrations of 1ngμl-1, 10 ngμl-1, 50 ngμl-1, and 100 ngμl-1. These four doses were used to assess the behavior of D. carinulata adults exposed to 4-oxo-(E)-2-hexenal in an olfactometer.

Electrophysiology

Laboratory reared individuals were used to assess antennal activity of male and female D. carinulata in response to exposure to 4-oxo-(E)-2-hexenal and only the highest concentration of 100 ngμl-1 was tested. A negative treatment control consisting of dichloromethane and a positive treatment control of 100 ngμl-1 of the host plant attractant

(E)-2-hexenal in dichloromethane were also tested (Cossé et al. 2006). The negative 162 control of pure dichloromethane was included to account for any response to the solvent and to quantify mechanical stimulation by moving air. D. carinulata is known to antennally and behaviorally respond to (E)-2-hexenal (Cossé et al. 2006). This positive control was therefore included to compare the uncharacterized D. carinulata antennal response to 4-oxo-(E)-2-hexenal, to the known antennal response to (E)-2-hexenal.

Treatments were applied in the following sequence: solvent control, (E)-2-hexenal, solvent control, 4-oxo-(E)-2-hexenal, and solvent control.

Electroantennography (EAG) was performed by inserting a drawn glass capillary filled with a 0.9% NaCl solution over a silver electrode into the back of an excised D. carinulata head. The distal end of one antenna was connected to a second recording electrode identical to the one inserted into the back of the head (Fig S2). The head was then positioned via micromanipulator into a glass odor delivery tube containing a continual stream of purified air. The micromanipulator was connected to a serial data- acquisition interface amplifier, type IDAC-232 (Syntech, Hilversum, Netherlands). The serial data-acquisition interface amplifier stabilized the input signal from the connected

D. carinulata head and amplified the signal. The signal was processed and analyzed using EAG software (Syntech NL 2001, Hilversum, Netherlands).

Assays were conducted by applying 10 μl of solution to filter paper located within a glass pipette connected to a controlled, purified air source. Solvents were allowed to evaporate from the filter paper for two minutes before experimentation. The tip of the glass pipette was inserted into the odor delivery tube after solvent had evaporated from the filter paper. Once a stable base line was evident in traces displayed by the acquisition 163 software, a 0.4 second puff of air was initiated, delivering the compound from the glass pipette into the continual airstream and over the antenna. The EAG software recorded the maximum deflection in volts that occurred from the depolarization of neurons in the antenna immediately after exposure to the puff in the continuous airstream.

Two-choice Olfactometer Bioassay

A 4-chambered 30 cm x 30 cm x 2.5 cm olfactometer (Sigma Scientific LLC,

Micanopy, FL, USA) with a 2-chamber adaptor was used to determine the response of adult D. carinulata to different odor sources. The olfactometer was used to assess for a potential locomotory response in test subjects, as a consistent flight response is difficult to obtain with D. carinulata in a small indoor enclosure. Purified air was supplied to the olfactometer at a rate of 200 mlmin-1. The olfactometer consisted of a central area with two lateral arms leading into an odor source chamber. The bioassays were conducted in a room designed with balanced lighting and muted background colors to minimize confounding visual cues, and located at the Rocky Mountain Research Station Forestry

Sciences Laboratory (USDA Forest Service, Bozeman MT, USA). The bioassay room was maintained at 25°C during experimentation.

Adult beetles were starved and stored individually in 1.5-ml microcentrifuge vials for 12 hours before testing. For each bioassay, a 1.5 cm2 wedge of filter paper was treated with the test solution and placed in one of the odor source chambers of the olfactometer, while the other arm received 10 μl of dichloromethane on the filter paper wedge to act as a control. Solvents were allowed to evaporate from the filter paper for two minutes before experimentation. After the paired treatments were placed in their respective odor source 164 chambers, the top of the olfactometer was removed and an individual D. carinulata was removed from its vial and placed in the vertical entry tube at the center of the olfactometer. The olfactometer cover was immediately secured and sealed. Each individual was allowed three minutes to respond to odor sources. When an individual entered one arm of the olfactometer, it was removed and the results were recorded. If the individual did not enter an arm of the olfactometer in three minutes, it was recorded as not responding and was not included in subsequent analysis. After each individual was tested, the Teflon olfactometer was cleaned with hot, soapy water and dried. A new treatment paper was added before each new test subject was introduced. After half the bioassay cohort was tested, placement of the odor treatments in the chambers was reversed. Two substances were evaluated for behavioral activity: larval hemolymph, and

4-oxo-(E)-2-hexenal. Dosage of the larval hemolymph consisted of 10 µl, which was approximately five third instar equivalents. The 4-oxo-(E)-2-hexenal treatments consisted of four doses: 10 ng, 100 ng, 500 ng, and 1000 ng in 10 μl of dichloromethane.

Statistical Analyses

Paired t-tests were used to analyze the EAG data. The maximum deflection in volts from the depolarization of the antenna to control puffs of solvent delivered immediately prior to, and also after, puffs of (E)-2-hexenal or 4-oxo-(E)-2-hexenal were averaged and acted as the control for the t-tests. The average of the maximum deflection from the control puffs before and after the treatment stimulus accounted for desensitization of the receptor cells in the antennae over the duration of the EAG protocol. The differences in response ratio between (E)-2-hexenal: control and 4-oxo-(E)- 165

2-hexenal: control were also analyzed using paired t-tests to determine if EAG responses differed. The response of D. carinulata adults in the olfactometer was analyzed using χ2 tests. All analysis was conducted using R software version 3.1.2.

Results

Volatile Collections

Although the larval hemolymph of D. carinulata is aromatically detected by the human nose, volatiles collected from 10 μl aliquots of the larval hemolymph produced very few distinctively identifiable compounds when analyzed on the GC-MS. Due to limitations in the availability of adequate volumes of the larval hemolymph and rapid oxidation of this limited supply, investigation of volatiles from larval hemolymph was discontinued.

GC-MS comparisons among volatiles collected from undamaged Tamarix plants, mechanically damaged Tamarix plants, and Tamarix plants with feeding D. carinulata larvae present identified a single compound unique to the volatile collections from chambers containing larvae and host plants (Fig 1). Mass spectra of the larva-produced compound indicated a molecular weight of 112, with the primary ion fragmentation molecular weight of 83 and 55. A search of the mass spectral library NIST MS Search v.2.0 indicated a close match with the compound 4-oxo-(E)-2-hexenal. The mass spectra and retention time matched a sample of 4-oxo-(E)-2-hexenal synthesized according to

Moreira and Millar (2005). Larval D. carinulata released 4-oxo-(E)-2-hexenal at an 166 average rate of 3.89 ng over the 3 day collection period. At no point was 4-oxo-(E)-2- hexenal detected in the volatile collections of undamaged or damaged Tamarix plants.

EAG Responses

Electrophysiological responses obtained from experiments using excised D. carinulata heads and 1-µg puffs of 4-oxo-(E)-2-hexenal (Fig 2) or (E)-2-hexenal (Fig S3) confirmed that significant antennal responses could be obtained at the concentrations tested. For the males, a greater depolarization resulted from exposure to 4-oxo-(E)-2- hexenal and (E)-2-hexenal than from exposure to the dichloromethane control (paired t- test P=0.002, n=10; paired t-test P=0.052, n=10). Responses of male antennae assessed using the 4-oxo-(E)-2-hexenal: control ratio and the (E)-2-hexenal: control ratio did not differ (paired t-test P=0.241, n=10) (Fig S4). Similar to males, female antennal receptors displayed a greater depolarization when exposed to 4-oxo-(E)-2-hexenal and (E)-2- hexenal compared to the control (paired t-test P<0.001, n=10; paired t-test P<0.001, n=10). The response of the female antennal receptors evident in the 4-oxo-(E)-2-hexenal: control ratio did not differ from response evident in the (E)-2-hexenal: control ratio

(paired t-test P=0.584, n=10) (Fig S4).

Repellency Trials

Based on the olfactory cues produced from 10 μl aliquots of larval hemolymph, female D. carinulata, whether reproductive or non-reproductive, consistently demonstrated no selective preference for the larval hemolymph treatment (χ2=0.18,

P=0.67, n= 48; χ2=0, P=1, n= 39) (Fig 3). Reproductive males conversely exhibited an 167 aversion to volatiles from the larval hemolymph, entering the control arm of the olfactometer more often than the arm with the larval hemolymph treatment (χ2=16.7,

P<0.001, n= 69) (Fig 3). Non-reproductive males did not exhibit any orientation response to odors from larval hemolymph and entered both arms of the olfactometer equally (χ2=0,

P=1, n= 34).

Reproductive male D. carinulata exhibited a preference for the control arm at the

500 ng and 1000 ng dose of 4-oxo-(E)-2-hexenal which indicated the compound was repellent (χ2=4.01, P=0.045, n=60; χ2=8.9, P=0.003, n=60) (Fig 4). The reproductive males did not respond to the 10 ng and 100 ng dose of 4-oxo-(E)-2-hexenal (χ2=0.11,

P=0.74, n=30; χ2=0, P=1, n=60) (Fig 4). Only one dose of 4-oxo-(E)-2-hexenal was tested for the non-reproductive adult males, 1000 ng, and the males displayed no preference (χ2=0.11, P=0.74, n=30) (Fig 5).

Reproductive female D. carinulata also exhibited a preference for the control arm of the olfactometer at the 100 ng, 500 ng, and 1000 ng dose of 4-oxo-(E)-2-hexenal, indicating the compound was repellent to both sexes (χ2=14.2, P<0.001, n=60; χ2=10.6,

P<0.001, n=60; χ2=10.6, P<0.001, n=60) (Fig 4). The reproductive females exhibited no preference at the 10-ng dose of 4-oxo-(E)-2-hexenal (χ2=0.11, P=0.74, n=30). Only one dose (1000 ng) of the 4-oxo-(E)-2-hexenal was tested for the non-reproductive adult females, and the females displayed no preference at this dose (χ2=0.64, P=0.42, n=30)

(Fig 5).

168

Discussion

The objective of this study was to identify semiochemicals repellent to adult D. carinulata. The hemolymph expelled by third instars during reflex bleeding was confirmed to be a source of volatile semiochemicals that were repellent to reproductive adult males. Reproductive females did not display any behavioral preference when exposed to larval hemolymph volatiles in the olfactometer. The compound 4-oxo-(E)-2- hexenal produced by the larvae was also identified as a repellent. This compound was repellent across multiple doses for reproductive males and females. These results suggest future deployment of 4-oxo-(E)-2-hexenal in the field could deter D. carinulata adults from colonizing or defoliating areas of critical habitat for avian species such as the southwestern willow flycatcher (Sogge et al. 2008; Paxton 2011).

Larvae in the current study emitted an average of 3.8 ng of 4-oxo-(E)-2-hexenal over 3 days of volatile collection. The emission rate from the larvae was variable, with

11.9 ng being the largest amount collected and with a minimum collection of 1.03 ng over 3 day aerations. The absence of 4-oxo-(E)-2-hexenal from the plant volatile profiles confirms that it is insect sourced. The production of 4-oxo-(E)-2-hexenal has been reported in the order Coleoptera, but has not been widely reported in the leaf beetle family Chrysomelidae (Dettner and Reissenweber 1991).

An unexpected result of the study was the difference in behavioral responses between the reproductive males and reproductive females to tested doses of 4-oxo-(E)-2- hexenal. The reproductive females were repelled by all but the lowest tested dose (10 ng), while the reproductive males were only repelled by the two highest doses (500 ng and 169

1000 ng). The increased sensitivity of the females to this compound might be expected.

Since D. carinulata is gregarious and aggregates in the field, there is a constant trade-off between increasing food competition, especially among offspring, and the benefits of being a member of an aggregation (Gaffke et al. 2018c; Stephan et al. 2015).

Reproductive females should select the habitat and hosts most advantageous for successfully rearing their offspring (Gripenberg et al. 2010). When depositing egg masses, the females benefit from the ability to assess cues on conspecific density to minimize intraspecific competition, especially when older immatures could potentially outcompete the newest offspring (Mitchell 1975). A greater ability of females to perceive the presence and density of conspecific larvae on specific host plants would allow for better oviposition choices (Raitanen et al. 2013), including the number of eggs to deposit

(Godfray 1991), and the evaluation of resource quality (Bergstrom et al. 2006). While reproductive males do not experience the same pressure to assure survival of offspring, the reproductive success of the males indirectly depends on resource quality and the amount of conspecific competition. Male mating success is dependent on the sensory responses and behaviors of females, which selects for males that respond to cues similarly to females (Sakaluk 2000; Ryan and Keddy-Hector 1992).

Non-reproductive males and females were not repelled by the larval hemolymph or the 4-oxo-(E)-2-hexenal, even at the highest tested dose. Considering that the non- reproductive adults are seeking hosts for development of metabolic reserves to overwinter and do not seek potential mates, it is not surprising that the non-reproductive adults were not repelled by odors collected from conspecific larvae (Bean et al. 2007a; 170

Rohr et al. 2002). In this situation, conspecific larvae function as direct competition for the non-reproductive adults, competing for limited resources as winter approaches. The non-reproductive adults could experience a compromised ability to overwinter if they were deterred from food resources that are occupied by conspecific larvae (Bean et al

2007a).

The 1000 ng dose of 4-oxo-(E)-2-hexenal elicited different maximum antennal depolarizations according to sex of adult D. carinulata. Depolarizations in response to the

1000 ng dose were greater for males than females. The difference in the magnitude of the depolarization was surprising, especially given the greater sensitivity of the females to low doses of 4-oxo-(E)-2-hexenal in the olfactometer bioassays. Males had the greatest depolarization of antennal receptors but females had a much larger depolarization ratio when compared to the solvent control. These differences in the magnitude of the depolarization and the greater ratio of the response to 4-oxo-(E)-2-hexenal to the solvent control suggests a potential difference in the number or the distribution of olfactory sensillae that respond to this compound on the antennae of male and female D. carinulata.

The difference in responses of reproductive males and females to the larval hemolymph vs the 4-oxo-(E)-2-hexenal in the olfactometer bioassays suggest that the 4- oxo-(E)-2-hexenal is not the compound responsible for the repellent response of the reproductive males to volatiles from hemolymph collected while larvae were reflex bleeding. Due to logistical constraints, further investigations into the volatiles from the larval hemolymph could not be completed. The rapid oxidation and breakdown of the 171 larval hemolymph could have made the hemolymph refractory to volatile collections.

Further experiments should be conducted with the larval hemolymph to isolate and identify the compounds that resulted in the observed repellency of reproductive adult males exposed to odors from larval hemolymph.

Many chrysomelid beetles are reported to be repelled by the presence of larval conspecifics and the defensive secretions of larvae (Hilker 1989; Hilker and Weitzel

1991; Rostas and Hilker 2002, Nufio and Papaj 2001). Our study demonstrates that compounds from larval hemolymph and a volatile compound sourced from intact larvae are repellent to adult D. carinulata. While the results of this study do not demonstrate field activity of the compounds, the present identification and bioassay results will allow for definitive field experiments to be performed. The repellent semiochemical should also be tested in the field in conjunction with the male produced aggregation pheromone and also, aggregation causing host plant volatiles (Cossé et al. 2005, 2006; Gaffke et al.

2018a, 2018b, 2018c). It would be of particular interest to determine if the repellent compounds could effectively negate the attractiveness of aggregation causing semiochemicals or if the compounds are only repellent in the absence of attractants.

The field application of semiochemicals repellent to adults could have broad impacts for the management of D. carinulata in the United States and could be used to decrease conflicts arising from the use of D. carinulata as a biological control agent for

Tamarix; more specifically conflict between the southwestern willow flycatcher habitat requirements and D. carinulata defoliation of this habitat. The potential interaction of D. carinulata and the southwestern willow flycatcher repeatedly delayed the broad 172 implementation of the Tamarix biological control program, and eventually resulted in the suspension of the program (Deloach et al. 2004; Dudley et al. 2000; APHIS 2010).

Although the suspension of the Tamarix biological control program may reduce the potential risk D. carinulata poses to the southwestern willow flycatcher, it also reduces the protection of natural ecosystems and native biodiversity that Tamarix biological control could have provided (Dudley and Deloach 2004; Dudley and Bean 2012). The purposeful deployment of these compounds around the critical habitat for the southwestern willow flycatcher could effectively ‘push’ D. carinulata out of specific locations, minimizing their damage on Tamarix nesting habitat. Furthermore, integration of these repellent compounds with known aggregation causing compounds could produce a ‘push-pull’ system that allows for even more effective manipulation of D. carinulata’s spatial distribution (Cossé et al. 2005, 2006; Gaffke et al. 2018a, 2018b, 2018c; Cook et al. 2007).

Acknowledgements

This work was supported by agreement 16-CA-11420004-047 of the Forest

Health Technology and Enterprise Team- Biological Control of Invasive Native and Non-

Native Plants (FHTET-BCIP). We thank Megan Hofland and Dr. Justin Runyon for critical method development and technical assistance. 173

A 500000 * 400000 300000 200000 100000 0 6 6.5 7 7.5 8 8.5 9

500000 400000 300000 200000 100000 0 6 6.5 7 7.5 8 8.5 9

Figure 1: Gas chromatography profiles of volatiles collected from A) larval D. carinulata feeding on Tamarix; B) undamaged Tamarix; C) mechanically damaged Tamarix. *Peak identified as 4-oxo-(E)-2-hexenal.

174

-0.0005 a

-0.001

-0.0015 Negative control -0.002

b 4-oxo-(E)-2-hexenal EAG response response EAG (V) -0.0025 a

-0.003 b

-0.0035 Female Male

Figure 2: Electroantennogram (EAG) response of D. carinulata to a negative control (dichloromethane) and 1,000 ng of 4-oxo-(E)-2-hexenal. Different letters below error bars denote statistical differences between the treatments.

175

Reproductive Adults Non-reproductive Adults 100 100 90 90 b 80 80 70 70 60 a Control 60 Control a a a a 50 a 50

40 10 μl 40 10 μl Response Response (%) Response Response (%) hemolymph hemolymph 30 a 30 20 20 10 10 0 0 FEMALE MALE FEMALE MALE

Figure 3: Response (%) of adult D. carinulata in a two-choice olfactometer to 10 μl of larval hemolymph vs control (purified air). Different letters above the error bars denote statistical differences between the treatments. 176

Reproductive Males 100 90

80 a 70 a 60 a a a 50 a Control 40 b 4-oxo-(E)-2-hexenal

Response Response (%) b 30 20 10 0 10 ng 100 ng 500 ng 1000 ng

Reproductive Females 100 90 a 80 a a 70 60 a 50 a Control 40 4-oxo-(E)-2-hexenal

Response Response (%) b b 30 b 20 10 0 10 ng 100 ng 500 ng 1000 ng

Figure 4: Response (%) of reproductive adult D. carinulata in a two-choice olfactometer to four doses of 4-oxo-(E)-2-hexenal vs the control (dichloromethane). Different letters above the bars denote statistical differences between the treatments. 177

Non-reproductive Adults 100 90 80 70 a 60 a Control b 50 b 40 4-Oxo-(E)-2-

Response Response (%) hexenal 30 20 10 0 females males

Figure 5: Response (%) of non-reproductive adult D. carinulata in a two-choice olfactometer to 1,000 ng of 4-oxo-(E)-2-hexenal vs the control (dichloromethane). Different letters above the bars denote statistical differences between the treatments.

178

Supplemental Material

Figure S1: Third instar induced to reflex bleed through pressure applied from forceps 179

Figure S2: Excised D. carinulata head attached to recording electrodes of the electroanntenogram

180

-0.0005 a -0.001

-0.0015 Negative control -0.002 Positive control

EAG response response EAG (V) -0.0025 a b -0.003 b -0.0035 Female Male

Figure S3: Electroantennogram (EAG) response of D. carinulata to a negative control (dichloromethane) and a positive control (1,000 ng of (E)-2-hexenal). Different letters below error bars denote statistical differences between the treatments.

181

10 a 9 a 8

6 postive control: control ratio 5

4 4-oxo-(E)-2-hexenal: control

Response Response ratio ratio 3 a 2 a 1

0 Female Male

Figure S4: Electroantennogram response ratio of D. carinulata adults to the ratio of positive control: negative control vs 4-oxo-(E)-2-hexenal: negative control. Different letters above error bars denote statistical differences between the ratios.

182

References

Bergström, A., Janz, N. and Nylin, S. 2006. Putting more eggs in the best basket: clutch‐ size regulation in the comma butterfly. Ecological Entomology, 31: 255-260.

Busch, D.E. 1995. Effects of fire on southwestern riparian plant community structure. The Southwestern Naturalist, 40: 259-267.

Cook, S.M., Khan, Z.R. and Pickett, J.A. 2006. The use of push-pull strategies in integrated pest management. Annual Review of Entomology, 52: 375-400

183

Deroe, C. and Pasteels, J.M. 1977. Defensive mechanisms against predation in the Colorado beetle (Leptinotarsa decemlineata Say). Archives of Biological Sciences, 88: 289-304.

Deroe, C. and Pasteels, J.M. 1982. Distribution of adult defense glands in chrysomelids (Coleoptera: Chrysomelidae) and its significance in the evolution of defense mechanisms within the family. Journal of Chemical Ecology, 8: 67-82.

Dettner, K., and Reissenweber, F. 1991. The defensive secretion of Omaliinae and Proteininae (Coleoptera: Staphylinidae): its chemistry, biological and taxonomic significance. Biochemical Systematics and Ecology, 19: 291-303.

Dudley, T.L. and DeLoach, C.J. 2004. Saltcedar (Tamarix spp.), endangered species, and biological weed control - can they mix? Weed Technology, 18: 1542-1551.

Dudley, T.L. and Bean, D.W. 2012. Tamarisk biocontrol, endangered species risk and resolution of conflict through riparian restoration. BioControl, 57: 331-347.

Dudley, T.L., Dalin, P., Bean, D.W. 2006. Status of the biological control of Tamarix spp. in California. In: Hoddle, M.S., Johnson, M.W. (Eds.), Proceedings of the V California Conference on Biological Control, 25–27 July 2006, Riverside, CA: 137–140.

Federal Register 2005. Designation of critical habitat for the Southwestern Willow Flycatcher (Empidonax traillii extimus). USDI-FWS, 50 CFR Part 17, RIN 1018– AT88. 60886 Federal Register: 201.

Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Grieco, P.A., Peterson, R.K.D., Weaver, D.K. 2018a. Semiochemicals to enhance 184

herbivory by Diorhabda carinulata aggregations in saltcedar (Tamarix spp.) infestations. Pests Management Science, doi: 10.1002/ps.4848 Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Weaver, D.K. 2018b. Field demonstration of a semiochemical treatment that enhances Diorhabda carinulata biological control of Tamarix spp. In review

Gaffke, A.M., Sing, S.E., Dudley, T.L., Bean, D.W., Russak, J.A., Mafra-Neto, A., Weaver, D.K. 2018c. Allee effects and aggregation pheromones: the influence of Diorhabda carinulata’s aggregation pheromone on the retention of new releases. In review

Godfray, H.C.J., Partridge L., and Harvey P.H. 1991. Clutch size. Annual Review of Ecology and Systematics, 22: 409-429.

Gripenberg, S., Mayhew, P.J., Parnell, M. and Roslin, T. 2010. A meta‐analysis of preference–performance relationships in phytophagous insects. Ecology Letters, 13: 383-393.

Hildebrandt, T.D. and Ohmart, R.D. 1982. Biological Resource Inventory (Vegetation and Wildlife) Pecos River Basin, New Mexico and Texas. Final Report to the Bureau of Reclamation, Contract No. 9-07-57-V0567.

Hilker, M. 1989. Intra‐and interspecific effects of larval secretions in some chrysomelids (Coleoptera). Entomologia Experimentalis et Applicata, 53: 237-245.

Hilker, M. and Weitzel, C. 1991. Oviposition deterrence by chemical signals of conspecific larvae in Diprion pini (Hymenoptera: Diprionidae) and Phyllodecta vulgatissima (Coleoptera: Chrysomelidae). Entomologia Generalis, 15: 293-302.

185

Kennedy, T.A., Finlay, J.C. and Hobbie, S.E. 2005. Eradication of invasive Tamarix ramosissima along a desert stream increases native fish density. Ecological Applications, 15: 2072-2083.

McShane, R.R., Auerbach, D.A., Friedman, J.M., Auble, G.T., Shafroth, P.B., Merigliano, M.F., Scott, M.L. and Poff, N.L. 2015. Distribution of invasive and native riparian woody plants across the western USA in relation to climate, river flow, floodplain geometry and patterns of introduction. Ecography, 38: 1254- 1265.

Mitchell, R. 1975. The evolution of oviposition tactics in the bean weevil, Callosobruchus maculatus (F.). Ecology, 56: 696-702.

Moreira, J.A. and Millar, J.G. 2005. Short and simple syntheses of 4-oxo-(E)-2-hexenal and homologs: pheromone components and defensive compounds of Hemiptera. Journal of Chemical Ecology, 31: 965-968.

Nufio, C.R. and Papaj, D.R. 2001. Host marking behavior in phytophagous insects and parasitoids. Entomologia Experimentalis et Applicata, 99: 273-293.

Ohmart, R.D. and Anderson, B.W. 1982. North American desert riparian ecosystems. Reference handbook on the deserts of North America. Greenwood Press, Westport, CT, pp.433-479.

Pasteels, J.M., Rowell‐Rahier, M., Randoux, T., Braekman, J.C. and Daloze, D. 1988. Pyrrolizidine alkaloids of probable host‐plant origin in the pronotal and elytral secretion of the leaf beetle Oreina cacaliae. Entomologia Experimentalis et Applicata, 49: 55-58.

Pearce, C.M. and Smith, D.G. 2003. Saltcedar: distribution, abundance, and dispersal mechanisms, northern Montana, USA. Wetlands, 23: 215-228.

Pearce, C.M. and Smith, D.G. 2007. Invasive saltcedar (Tamarix): its spread from the American Southwest to the northern Great Plains. Physical Geography, 28: 507- 530.

Raitanen, J., Forsman, J.T., Kivelä, S.M., Mäenpää, M.I. and Välimäki, P. 2013. Attraction to conspecific eggs may guide oviposition site selection in a solitary insect. Behavioral Ecology, 25: 110-116.

Rohr, J.R., Madison, D.M. and Sullivan, A.M. 2002. Sex differences and seasonal trade- offs in response to injured and non-injured conspecifics in red-spotted newts, Notophthalmus viridescens. Behavioral Ecology and Sociobiology, 52: 385-393.

Rostás, M. and Hilker, M. 2002. Feeding damage by larvae of the mustard leaf beetle deters conspecific females from oviposition and feeding. Entomologia Experimentalis et Applicata, 103: 267-277.

Ryan, M. J. and A. Keddy-Hector, 1992. Directional patterns of female mate choice and the role of sensory biases. The American Naturalist, 139: 4–35.

Sakaluk, S. K. 2000. Sensory exploitation as an evolutionary origin to nuptial food gifts in insects. Proceedings of the Royal Society of London Series B - Biological Sciences, 267: 339–343.

Schindek, R. and Hilker, M. 1996. Influence of larvae of Gastrophysa viridula on the distribution of conspecific adults in the field. Ecological Entomology, 21:370- 376.

Schlatter, K.J., Grabau, M.R., Shafroth, P.B. and Zamora-Arroyo, F. 2017. Integrating active restoration with environmental flows to improve native riparian tree establishment in the Colorado River Delta. Ecological Engineering, 106: 661-674.

187

Sogge, M.K., Sferra, S.J. and Paxton, E.H. 2008. Tamarix as habitat for birds: implications for riparian restoration in the southwestern United States. Restoration Ecology, 16:146-154.

Stephan, J.G., Stenberg, J.A. and Björkman, C. 2015. How far away is the next basket of eggs? Spatial memory and perceived cues shape aggregation patterns in a leaf beetle. Ecology, 96: 908-914.

Stocks, I. 2008. Reflex bleeding (autohemorrage). Encyclopedia of Entomology, second ed. Springer: 3132-3139. Tracy, J.L. and Robbins, T.O. 2009. Taxonomic revision and biogeography of the Tamarix-feeding Diorhabda elongata (Brullé, 1832) species group (Coleoptera: Chrysomelidae: Galerucinae: Galerucini) and analysis of their potential in biological control of Tamarisk. Zootaxa 2101: 1-152.

US Fish and Wildlife Service, 1995. Final rule determining endangered status for the southwestern willow flycatcher. Federal register, 60: 10694-10715.

Wallace, J.B. and Blum, M.S. 1971. Reflex bleeding: a highly refined defensive mechanism in Diabrotica larvae (Coleoptera: Chrysomelidae). Annals of the Entomological Society of America, 64: 1021-1024.

Yapp, R.H. 1922. The concept of habitat. Journal of Ecology, 10: 1-17.

188

CHAPTER SIX

CONCLUSIONS

Over the nearly two decades since the first release of the northern tamarisk leaf beetle Diorhabda carinulata in North America, five significant challenges have become readily apparent: (1) this agent often rapidly disperses soon after being released, (2) Allee effects likely constrain successful establishment of new releases, (3) low density populations of D. carinulata have little impact on Tamarix, (4) D. carinulata is difficult to detect and accurately monitor at low densities, and (5) distributions for D. carinulata and endangered native fauna nesting in Tamarix now overlap. The research presented in this dissertation was intended to provide actionable information on the chemical ecology of D. carinulata, and how it could be applied to resolve the challenges facing Tamarix biological control. The proposed resolution of these issues universally rests on applied research which demonstrated how densities and spatial distributions of D. carinulata can be advantageously manipulated in the field through the application of lures emitting a species-specific aggregation pheromone and behaviorally active host plant volatiles.

Operational adoption and deployment of semiochemically based treatments directly addresses each of these challenges. Our results illustrate that the use of lures to manipulate aggregation can enhance Tamarix management by slowing the dispersal of agents away from release sites, minimizing detrimental Allee effects associated with small populations, potentiating low density agent populations, increasing agent monitoring efficiency and accuracy, and ‘pulling’ adult beetles into aggregations away 189 from Tamarix nesting sites. Repellent compounds have also been identified; preliminary results indicate these could facilitate dispersal, and with further (tech transfer) development could aid in ‘pushing’ D. carinulata from areas in conflict with avian species.

Through the application of the aggregation pheromone and behaviorally active host plant volatiles, ineffectively small populations of D. carinulata were potentiated, resulting in fitness compromising levels of defoliation on semiochemically treated plants.

Applications of semiochemical treatments for the purpose of manipulating D. carinulata spatial distributions were equally successful, resulting in aggregations of adult beetles on specific, selected plants. Aggregations of reproductive adult beetles resulted in higher densities of larvae on treated plants, higher levels of damage, increased rates of plant dieback, and in increased reduction in Tamarix canopy volume. These results provide land managers with an informed, effective strategy to dictate where D. carinulata does its damage. Field trials of 1-g and 4-g doses of SPLAT impregnated with the aggregation pheromone demonstrate the robust, field-ready potential of a semiochemical exploiting strategy to enhance the control of Tamarix. Deployment of 4-g doses of SPLAT alleviated practical constraints associated with the time-limited treatment activity, necessitating repeated applications, of the smaller volume (1 g) lure initially tested in proof of concept field experiements. Enhanced control of Tamarix was achieved with only three targeted applications of aggregation pheromone lures over D. carinulata’s period of annual reproductivity. 190

Applications of aggregation pheromone-based lures were found to be so effective at stimulating aggregations in free-living populations of D. carinulata that plants treated with the aggregation pheromone could serve as ‘sentinels’, increasing the ease and accuracy of monitoring this agent. Information recorded from monitoring activities would be more robust because the probability of recording false negatives would be minimized or eliminated where pheromones were used to manipulate aggregations of D. carinulata adults, larvae, and their impacts on selected locations. Increasing the accuracy of agent monitoring would provide land managers with essential information on which to base management decisions with potentially long term ramifications. With the continuing expansion of the distribution of D. carinulata in the United States, there remains an urgent need for effective and accurate monitoring tools. The application of the aggregation pheromone for monitoring purposes can address this need, but additional research should be conducted on techniques to maximize the effectiveness of the aggregation pheromone as a monitoring tool. Applied research should also be conducted to assess the potential application of the aggregation pheromone in conjunction with passive traps, such as yellow sticky card or delta traps.

The rapid dispersal of D. carinulata after release makes D. carinulata particularly vulnerable to Allee effects. These effects could be mitigated through the application of aggregation pheromone based lures. A single application of the aggregation pheromone during the initial release of the adults resulted in more individuals being retained around the release area, and in larger initial populations. By applying aggregation pheromone at the time of a new release, biological control practitioners may reduce detrimental Allee 191 effects; using the same strategy might also increase overall successful establishment and treatment efficacy with multiple, smaller releases. The current study did demonstrate the potential for reduction of Allee effects by increasing initial retention of newly released adults on release sites through the application of aggregation pheromone based lures.

However, for many additional reasons, establishment of these releases was in fact low.

Future research should be conducted to identify specific impacts aggregation pheromones have on Allee effects, and ultimately, the influence aggregation pheromones have on release size and the likelihood of establishment of biological control agents.

This dissertation also reports on the isolation, identification, and behavioral activity of repellent semiochemical compounds. The development of management tools exploiting repellent semiochemicals is critical, as being able to control interactions between D. carinulata and native fauna is now a regulatory concern for most land managers. Identification of repellent compounds could have broad implications for the spatial manipulation of D. carinulata. The purposeful deployment of repellent compounds within southwestern willow flycatcher critical habitat could effectively

‘push’ D. carinulata out of the area, minimizing interaction with this and other imperiled avian species. Repellent compounds could be used alone as a ‘push’ to deter D. carinulata from remaining or aggregating in a specific area, and in conjunction with the aggregation pheromone to develop a ‘push-pull’ system to further manipulate D. carinulata densities. The research on the repellent compounds is not complete, but the foundational science has been conducted. Field validation is needed to assess responses 192 to the larval produced compound, and further attempts at isolation and identification are needed on the hemolymph.

Through the use of chemical ecology, a ‘pull’ system has been developed to address many of the challenges facing the use of D. carinulata as a biological control agent for Tamarix. The foundational research has also been conducted on a ‘push’ system that can be deployed to manipulate D. carinulata. The integration of these two strategies can resolve the challenges facing D. carinulata and can be used to further enhance control strategies for Tamarix.

193

REFERENCES CITED

194

Allee, W.C., Park, O., Emerson, A.E., Park, T. and Schmidt, K.P. 1949. Principles of animal ecology. Saunders Co., Philadelphia, PA, USA.

Andersen, D.C. and Nelson, S.M. 2013. Floral ecology and insect visitation in riparian Tamarix sp. (saltcedar). Journal of Arid Environments 94: 105-112.

Anderson, B.W. and Ohmart, R.D. 1984. Vegetation management study for the enhancement of wildlife along the lower Colorado River. Final report to USA Bureau of Reclamation, Contract No. 7-07-30-V0009

APHIS 2010. USDA APHIS PPQ Moratorium for biological control of saltcedar. https://www.aphis.usda.gov/plant_health/ea/downloads/SaltcedarMemo-2010.pdf

Barrows, C. 1998. The case for wholesale removal. Restoration and management notes 16: 130-134

Baum, B. 1966. Monographic revision of the genus Tamarix. Final research report for the USDA. Department of Botany, Hebrew University, Jerusalem, Israel.

Baum, B. 1978. The genus Tamarix. Academy of Science and Humanities, Jerusalem, Israel.

195

Berec, L., Angulo, E. and Courchamp, F. 2007. Multiple Allee effects and population management. Trends in Ecology & Evolution, 22: 185-191.

Berec, L., Boukal, D.S. and Berec, M. 2001. Linking the Allee effect, sexual reproduction, and temperature-dependent sex determination via spatial dynamics. The American Naturalist, 157: 217-230.

Berggren, Å. 2001. Colonization success in Roesel's bush cricket Matrioptera roeseli: the effects of propagule size. Ecology 82:274‐280.

Bergström, A., Janz, N. and Nylin, S. 2006. Putting more eggs in the best basket: clutch‐ size regulation in the comma butterfly. Ecological Entomology, 31: 255-260.

Berry, W.L. 1970. Characteristics of salts secreted by Tamarix aphylla. American Journal of Botany, 57: 1226-1230.

Blum, M.S. and Brand, J.M. 1972. Social insect pheromones: their chemistry and function. American Zoologist, 12: 553-576.

Borden, J.H. 1989. Semiochemicals and bark beetle populations: exploitation of natural phenomena by pest management strategists. Ecography, 12: 501-510.

Brock, J.H. 1994. Tamarix spp. (salt cedar), an invasive exotic woody plant in arid and semi-arid riparian habitats of western USA. Ecology and Management of Invasive Riverside Plants. Wiley, New York, NY: 27-44.

Brotherson, J.D., and Field, D. (1987). Tamarix: impacts of a successful weed. Rangelands, 9: 110-112.

Bruce, T.J. and Pickett, J.A. 2011. Perception of plant volatile blends by herbivorous insects–finding the right mix. Phytochemistry, 72: 1605-1611.

Bruce, T.J., Wadhams, L.J. and Woodcock, C.M. 2005. Insect host location: a volatile situation. Trends in Plant Science, 10: 269-274.

Busch, D.E. 1995. Effects of fire on southwestern riparian plant community structure. The Southwestern Naturalist, 40: 259-267.

197

Busch, D.E. and Smith, S.D. 1993. Effects of fire on water and salinity relations of riparian woody taxa. Oecologia, 94: 186-194.

Byrne, K.J., Gore, W.E., Pearce, G.T. and Silverstein, R.M. 1975. Porapak-Q collection of airborne organic compounds serving as models for insect pheromones. Journal of Chemical Ecology, 1: 1-7.

Cardé, R.T. and Minks, A.K. 1995. Control of moth pests by mating disruption: successes and constraints. Annual Review of Entomology, 40: 559-585.

Carman, J.G. and Brotherson, J.D. 1982. Comparisons of sites infested and not infested with saltcedar (Tamarix pentandra) and Russian olive (Elaeagnus angustifolia). Weed Science, 30: 360-364.

Carreira, J.A. and Niell, F.X. 1992. Plant nutrient changes in a semi‐arid Mediterranean shrubland after fire. Journal of Vegetation Science, 3: 457-466.

Center for Biological Diversity 2009. Lawsuit filed to save endangered songbird; southwestern willow flycatcher threatened by release of imported beetle. http://www.biologicaldiversity.org/news/press_releases/2009/southwestern- willow-flycatcher-03-27-2009.html.

Chapman, R.F. 2003. Contact chemoreception in feeding by phytophagous insects. Annual Review of Entomology, 48: 455-484.

Chew, M.K. 2009. The monstering of tamarisk: how scientists made a plant into a problem. Journal of the History of Biology, 42: .231-266.

198

Cleverly, J. 2013. Water use by Tamarix. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West.Oxford University Press, New York, NY: 85-98.

Cleverly, J.R., Smith, S.D., Sala, A. and Devitt, D.A. 1997. Invasive capacity of Tamarix ramosissima in a Mojave Desert floodplain: the role of drought. Oecologia, 111:12-18.

Cook, S.M., Khan, Z.R. and Pickett, J.A. 2006. The use of push-pull strategies in integrated pest management. Annual Review of Entomology, 52: 375-400

Crawley, M.J. 1989. Insect herbivores and plant population dynamics. Annual Review of Entomology, 34: 531-562.

Dahanukar, A., Hallem, E.A. and Carlson, J.R. 2005. Insect chemoreception. Current Opinion in Neurobiology, 15: 423-430.

199

DeLoach, C.J. and Tracy, J.L. 1997. Effects of biological control of saltcedar (Tamarix ramosissima) on endangered species: Draft Biological Assessment, 17 October 1997. USDA/ARS, Temple, TX: 612

DeLoach, C.J., Lewis, P.A., Herr, J.C., Carruthers, R.I., Tracy, J.L. and Johnson, J. 2003. Host specificity of the leaf beetle, Diorhabda elongata deserticola (Coleoptera: 200

Chrysomelidae) from Asia, a biological control agent for saltcedars (Tamarix: Tamaricaceae) in the Western United States. Biological Control, 27: 117-147.

Dennis, B. 2002. Allee effects in stochastic populations. Oikos, 96: 389-401.

Deroe, C. and Pasteels, J.M. 1977. Defensive mechanisms against predation in the Colorado beetle (Leptinotarsa decemlineata Say). Archives of Biological Sciences, 88: 289-304.

Dethier, V.J. 1948. Chemical insect attractants and repellents. Soil Science, 65:1-307.

Di Tomaso, J.M. 1998. Impact, biology, and ecology of saltcedar (Tamarix spp.) in the southwestern United States. Weed Technology, 12: 326-336.

Dickens, J.C. 1989. Green leaf volatiles enhance aggregation pheromone of boll weevil, Anthonomus grandis. Entomologia Experimentalis et Applicata, 52: 191-203.

Dickens, J.C., Billings, R.F. and Payne, T.L. 1992. Green leaf volatiles interrupt aggregation pheromone response in bark beetles infesting southern pines. Experientia, 48: 523-524.

Douglass, C.H., Nissen, S.J. and Hart, C.R. 2013. Tamarisk Management. Tamarix: a case study of ecological change in the American West. Oxford University Press, New York, NY: 333-353.

201

Dreesen, D.R. and Wangen, L.E. 1981. Elemental composition of saltcedar (Tamarix chinensis) impacted by effluents from a coal-fired power plant. Journal of Environmental Quality, 10: 410-416.

Drus, G.M. 2013.Fire ecology of Tamarix. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 377-403

Dudley, T.L. and Bean, D.W. 2012. Tamarisk biocontrol, endangered species risk and resolution of conflict through riparian restoration. BioControl, 57: 331-347.

Dudley, T.L. and DeLoach, C.J. 2004. Saltcedar (Tamarix spp.), endangered species, and biological weed control - can they mix? Weed Technology, 18: 1542-1551.

Duncan, K.W. and McDaniel, K.C. 1998. Saltcedar (Tamarix spp.) management with imazapyr. Weed Technology, 12: 337-344.

Duncan, R.P. 2016. How propagule size and environmental suitability jointly determine establishment success: a test using dung beetle introductions. Biological invasions, 18: 985-996.

Ellis, L.M. 2001. Short-term response of woody plants to fire in a Rio Grande riparian forest, Central New Mexico, USA. Biological Conservation, 97: 159-170.

Engel-Wilson, R.W. and Ohmart, R.D. 1978, December. Floral and attendant faunal changes on the lower Rio Grande between Fort Quitman and Presidio, Texas. In Proceedings of the National Symposium for Protection and Management of Floodplain Wetlands: 139-147.

Everitt, B.L. 1980. Ecology of saltcedar—a plea for research. Environmental Geology, 3: 77-84.

Fagan, W.F., Lewis, M.A., Neubert, M.G. and Van Den Driessche, P. 2002. Invasion theory and biological control. Ecology Letters, 5: 148-157.

Fauvergue, X., 2013. A review of mate‐finding Allee effects in insects: from individual behavior to population management. Entomologia Experimentalis et Applicata, 146: 79-92.

Fauvergue, X., Vercken, E., Malausa, T. and Hufbauer, R.A. 2012. The biology of small, introduced populations, with special reference to biological control. Evolutionary applications 5: 424-443.

Federal Register 2005. Designation of critical habitat for the Southwestern Willow Flycatcher (Empidonax traillii extimus). USDI-FWS, 50 CFR Part 17, RIN 1018– AT88. 60886 Federal Register: 201.

Fenner, P., Brady, W.W. and Patton, D.R. 1984. Observations on seeds and seedlings of Fremont cottonwood. Desert Plants, 6: 55-58. 203

Fernandez, P. and Hilker, M. 2007. Host plant location by Chrysomelidae. Basic and Applied Ecology, 8: 97-116.

Fick, W.H. and Geyer, W.A. 2010. Cut-stump treatment of saltcedar (Tamarix ramosissima) on the Cimarron National Grasslands. Transactions of the Kansas Academy of Science, 113: 223-226.

Frasier, G.W. and Johnsen Jr, T.N. 1991. Saltcedar (Tamarisk): classification, distribution, ecology and control. Noxious Range Weeds. Westview Press, Boulder CO: 377-393.

Freckleton, R.P. 2000. Biological control as a learning process. Trends in Ecology & Evolution, 15: 263-264.

204

Gascoigne, J., Berec, L., Gregory, S. and Courchamp, F. 2009. Dangerously few liaisons: a review of mate-finding Allee effects. Population Ecology, 51: 355-372.

Gaskin, J.F. 2013. Genetics of Tamarix. In: Sher A., Quigley M.F., eds. Tamarix: A Case Study of Ecological Change in the American West. Oxford University Press, New York, NY: 21-29

Gaskin, J.F. and Schaal, B.A. 2002. Hybrid Tamarix widespread in US invasion and undetected in native Asian range. Proceedings of the National Academy of Sciences, 99: 11256-11259.

Gaskin, J.F., Birken, A.S. and Cooper, D.J. 2012. Levels of novel hybridization in the saltcedar invasion compared over seven decades. Biological Invasions, 14: 693- 699.

Gasry, H.L. 1963. Root distribution of five-stamen tamarisk, seepwillow and arrowweed. Forest Science, 9: 311-314.

Gaston, L.K, Shorey, H.H. and Saario, C.A. 1967. Insect population control by the use of sex pheromones to inhibit orientation between the sexes. Nature, 213: 1155-1155

Glenn, E.P. and Nagler, P.L. 2005. Comparative ecophysiology of Tamarix ramosissima and native trees in western US riparian zones. Journal of Arid Environments, 61: 419-446.

205

Global Invasive Species Database 2018. Available via DIALOG. http://www.iucngisd. org/gisd/100_worst.php on 14-12-2018.Cited March 9, 2018

Godfray, H.C.J., Partridge L., and Harvey P.H. 1991. Clutch size. Annual Review of Ecology and Systematics, 22: 409-429.

González, E., Sher, A. A., Anderson, R. M., Bay, R. F., Bean, D. W., Bissonnete, G. J., Bourgeois, B., Cooper, D. J., Dohrenwend, K., Eichhorst, K. D., El Waer, H., Kennard, D. K., Harms-Weissinger, R., Henry, A. L., Makarick, L. J., Ostoja, S. M., Reynolds, L. V., Robinson, W. W. and Shafroth, P. B. 2017a. Vegetation response to invasive Tamarix control in southwestern U.S. rivers: a collaborative study including 416 sites. Ecological Applications, doi:10.1002/eap.1566

González, E., Sher, A.A., Anderson, R.M., Bay, R.F., Bean, D.W., Bissonnete, G.J., Cooper, D.J., Dohrenwend, K., Eichhorst, K.D., El Waer, H. and Kennard, D.K. 2017b. Secondary invasions of noxious weeds associated with control of invasive Tamarix are frequent, idiosyncratic and persistent. Biological Conservation 213: 106-114.

Graf, W.L. 1978. Fluvial adjustments to the spread of tamarisk in the Colorado Plateau region. Geological Society of America Bulletin, 89: 1491-1501.

Grevstad, F. S. 1999. Experimental invasions using biological control introductions: the influence of release size on the chance of population establishment. Biological Invasions, 1: 313-323.

Gripenberg, S., Mayhew, P.J., Parnell, M. and Roslin, T. 2010. A meta‐analysis of preference–performance relationships in phytophagous insects. Ecology Letters, 13: 383-393.

Hardee, D.D., Mitchell, E.B. and Huddleston, P.M. 1967. Laboratory studies of sex attraction in the boll weevil. Journal of Economic Entomology, 60: 1221-1224.

Harris, P. and Zwölfer, H. 1968. Screening of phytophagous insects for biological control of weeds. The Canadian Entomologist, 100: 295-303.

Hem, J.D. 1967. Composition of Saline Residues on Leaves and Stems of Saltcedar (Tamarix pentandra Pallas). U.S. Government Printing Office, Washington D.C., USA: 9

Hilker, M. 1989. Intra‐and interspecific effects of larval secretions in some chrysomelids (Coleoptera). Entomologia Experimentalis et Applicata, 53: 237-245.

207

Hokkanen, H.M. 1991. Trap cropping in pest management. Annual Review of Entomology, 36:119-138.

Hopper, K.R. and Roush, R.T. 1993. Mate finding, dispersal, number released, and the success of biological control introductions. Ecological entomology, 18: .321-331.

Horton, J.S. 1964. Notes on the Introduction of Deciduous Tamarisk. U.S. Department of Agriculture, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. Research Note 16

Howse, P.E., Stevens, I.D.R. and Jones, C.O. 1998. Pest monitoring. Insect Pheromones andTheir Use in Pest Management. Springer Science & Business Media: 263-279

beetle (Diorhabda carinulata). Agricultural and Forest Meteorology, 150: 1467- 1475.

Irvine, J.R. and West, N.E. 1979. Riparian tree species distribution and succession along the lower Escalante River, Utah. The Southwestern Naturalist, 24: 331-346.

Jermy, T. and Szentesi, A. 1978. The role of inhibitory stimuli in the choice of oviposition site by phytophagous insects. Entomologia Experimentalis et Applicata, 24: 458-471.

Johnson, S. 1987. Can saltcedar be controlled. Fremontia, 15: 19-20.

Kelsey, J.M. 1966. Costelytra zealandica (White) sex attraction. New Zealand Entomologist, 3: 57-59.

Kennedy, T.A., Finlay, J.C. and Hobbie, S.E. 2005. Eradication of invasive Tamarix ramosissima along a desert stream increases native fish density. Ecological Applications, 15: 2072-2083. 209

Kleinkopf, G.E. and Wallace, A. 1974. Physiological basis for salt tolerance in Tamarix ramosissima. Plant Science Letters, 3: 157-163.

Knight, A.L. and Light, D.M. 2001. Attractants from Bartlett pear for codling moth, Cydia pomonella (L.), larvae. Naturwissenschaften, 88: 339-342.

Krasnoff, S.B. and Dussourd, D.E. 1989. Dihydropyrrolizine attractants for arctiid moths that visit plants containing pyrrolizidine alkaloids. Journal of Chemical Ecology, 15: 47-60.

Kuznetsova, A., Brockhoff, P.B. and Christensen, R.H.B. 2015. Package ‘lmerTest’. R package version, 2.

Ladenburger, C.G., Hild, A.L., Kazmer, D.J. and Munn, L.C. 2006. Soil salinity patterns in Tamarix invasions in the Bighorn Basin, Wyoming, USA. Journal of Arid Environments, 65: 111-128.

Landolt, P.J. 1997. Sex attractant and aggregation pheromones of male phytophagous insects. American Entomologist, 43:12-22.

Liesner, D.R. 1971. Phytophagous insects of Tamarix spp. New Mexico. Master of Science thesis, New Mexico State University, Las Cruces, New Mexico.

Lovich, J.E. and Bainbridge, D. 1999. Anthropogenic degradation of the southern California desert ecosystem and prospects for natural recovery and restoration. Environmental Management, 24: 309-326.

Maron, J.L. 1997. Interspecific competition and insect herbivory reduce bush lupine (Lupinus arboreus) seedling survival. Oecologia, 110: 284-290.

McDaniel, K.C. and Taylor, J.P. 2003. Saltcedar recovery after herbicide-burn and mechanical clearing practices. Journal of Range Management, 56: 439-445.

McDonald, A.K. 2010. Hydrologic Impacts of Saltcedar Control Along a Regulated Dryland River. Doctoral dissertation, Texas A&M University.

211