IMPACT OF ALIEN ON NATIVE SEEDLINGS IN A DIVERSE MESIC FOREST, O‘AHU, HAWAI‘I, AND A STUDY OF FOOD PLANT PREFERENCES

A THESIS SUBITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAI‘I IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

IN

BOTANICAL SCIENCES

(BOTANY-ECOLOGY, EVOLUTION, AND CONSERVATION BIOLOGY)

MAY 2006

By Stephanie M. Joe

Thesis Committee: Curtis C. Daehler, Chairperson Donald R. Drake David C. Duffy Robert H. Cowie

We certify that we have read this thesis and that, in our opinion, it is satisfactory

in scope and quality as a thesis for degree of Master of Science in Botanical

Sciences (Botany-Ecology, Evolution, and Conservation Biology)

THESIS COMMITTEE

Chairperson ACKNOWLEDGEMENTS

Funding for this study was received from the Ecology, Evolution and

Conservation Biology Research Grant, the Beatrice Krauss Fellowship in Botany

and the Sigma Xi Grants in Aid of Research. My committee has been absolutely

vital throughout. I am grateful to my advisor Curtis C. Daehler for his enthusiasm

and diligence, Robert H. Cowie for sound advice and thorough editing, Donald R.

Drake for encouraging me to study slugs rather than worms and David C. Duffy for sticking by me through the whole ordeal. Additionally, I thank the U.S. Army

Garrison Natural Resource Staff, without whom this project would never have succeeded. They grew the , provided materials and staff and helped collect data. In particular, I am grateful to Naomi Arcand, Seth Cato, Vincent Costello,

Hoala Fraiola, Victor Göbel, Julia Gustine, Kapua Kawelo, Matthew Keir, David

K. Palumbo, Leanne Obra, Jobriath Rohrer, Dominic Souza, Robert Romualdo,

William Weaver and Lauren Weisenberger for their help. Students, researchers and staff from the University of Hawai‘i helped out as well. I would like to thank

Kyle K. Koza, Victor Cizankas, Will Haines, Mitsuko Yorkston, Charles Chimera and Karen Brimacombe. Above all, I thank my husband, Paul D. Krushelnycky.

iii ABSTRACT

Introduced have the potential to cause serious ecological disruption, particularly on oceanic islands. When introduced species invade natural areas, endemic species may be threatened, especially when the invasive species represent guilds or functional groups that were previously lacking. Hawai‘i has no native slugs, but over a dozen species are now established. Slugs are important seedling predators in their native habitats, and in introduced habitats they can cause major shifts in the abundance some plant species. In order to better investigate slug impacts on native plants in Hawai‘i, I carried out research which

1. identified differences in the acceptability of five native plant species to five alien slug species 2. assessed the effect of slug herbivory on the growth and survival of three native and two alien plant species, and 3. measured changes in seedling regeneration due to slug herbivory.

Results from feeding assays indicated a significant difference in palatability among plant species, but no statistical difference in overall feeding preference among slug species. Urera kaalae (Urticaceae) was found to be significantly more palatable than the other four plant species and, thus, is predicted to be more vulnerable to slug herbivory in the field.

I tracked the fate of planted seedlings and natural germinants from the seed bank in both slug-excluded and slug-accessible plots in diverse mesic forest in the

iv Wai‘anae Mountains on the island of O‘ahu. Among seedlings that survived to the end of the experiment, there was no significant difference between slug- herbivory treatments in growth index measurements. There was little germination from the seed bank, with no statistical difference in total number of seedlings between treatments. However, two of the three native species, Schidea obovata

(Caryophyllaceae) and superba () had significant reductions in survival of 49% and 53%, respectively, in the slug-exposed treatment. Survival of two invasive species, Clidemia hirta (Meslastomataceae) and Psidium cattleianum (Myrtaceae) was not significantly affected by slugs. This study demonstrates that slugs may pose a serious threat to native plant species by reducing their survival and thereby facilitate the success of certain invasive species.

v TABLE OF CONTENTS

Acknowledgements ...... iii Abstract ...... iv List of Tables ...... viii List of Figures ...... viii Chapter 1: Literature review and discussion of hypotheses ...... 1 Introduction ...... 1 Study organisms ...... 2 Impacts of slug herbivory ...... 3 Slugs in Hawai‘i ...... 5 Hypotheses ...... 7 Chapter 2: Palatability of native plant species to alien slug species ...... 16 Introduction ...... 16 Materials and methods ...... 18 Study organisms ...... 18 Feeding trial protocol ...... 21 Statistical Analysis ...... 24 Results ...... 24 Discussion ...... 25 Chapter 3: Impact of alien slugs on native plant seedlings ...... 37 Introduction ...... 37 Materials and methods ...... 40 Field site ...... 40 Study species ...... 42 Seedling preparation ...... 43 Experimental design ...... 44 Slug monitoring ...... 47 Statistical analysis ...... 47 Results ...... 49 Plant growth ...... 49 Herbivory damage ...... 49

vi Seedling survival ...... 50 Seed bank regeneration ...... 51 Slug monitoring ...... 51 Discussion ...... 51 Appendix A: Change in plant size index over time ...... 63 Appendix B: Change in leaf number over time ...... 68 Appendix C: Change in herbivore damage over time ...... 73 Literature cited ...... 78

vii LIST OF TABLES

Table Page 1.1 Gastropod families containing slugs ...... 9 1.2 List of slugs found within the state of Hawai‘i ...... 10 1.3 List of native species threatened by alien slugs ...... 11 2.1 Mean acceptability indices (AI) for 25 slug-plant species pairs . . . 30 2.2 ANOVA using AI scores from 25 slug-plant species pairs ...... 31 2.3 Secondary plant compounds repellent to slugs ...... 31 3.1 Seedling height (mm) by species on day 0 of the study ...... 55 3.2 Two-way ANOVA of seedling survival by herbivory treatment . . . . 55 3.3 Number of germinants from seed bank by herbivory treatment . . . 55

LIST OF FIGURES

Figure Page 2.1 Slug collection sites on the Island of O‘ahu ...... 32 2.2 Percent of slugs engaged in feeding ...... 33 2.3 Mean AI scores for each plant species (all slug species) ...... 34 2.4 Distribution of AI scores ...... 35 2.5 Boxplot showing weight distribution of slug species ...... 36 3.1 Location of Kahanahāiki Management Unit ...... 56 3.2 Plant growth in slug-exposed vs. slug-excluded plots ...... 57 3.3 Leaf change in slug-exposed vs. slug-excluded plots ...... 58 3.4 Herbivory damage in slug-exposed vs. slug-excluded plots . . . . . 59 3.5 Survival in slug-exposed vs. slug-excluded plots ...... 60 3.6 Survival over time by plant species ...... 61 3.7 Slug counts at refugia over time ...... 62

viii CHAPTER 1

LITERATURE REVIEW AND DISCUSSION OF HYPOTHESES

INTRODUCTION

The deliberate and accidental introduction of alien species into new habitats where they establish and spread is a significant component of human-caused global change (Vitousek et al. 1997). Through predation on and competition with native species, introduced species accelerate rates and reduce global biodiversity (Gurevitch and Padilla 2004). For instance, approximately 42% of the

species on the U.S. Threatened and lists are at risk

primarily because of the activities of introduced species (Pimentel et al. 2004). In

other regions of the world, introduced species seriously threaten up to 80% of

native endangered species (Armstrong 1995).

The endangerment and loss of native species is particularly acute on oceanic

islands, which have unusually high rates of endemism (Kaneshiro 1988). In the

Hawaiian Islands, where a considerable portion of intact native ecosystems are

under State and Federal protection (Cuddihy and Stone 1990), invasive species

are now the primary threat to the persistence of the native flora and fauna (Loope

1992). For native Hawaiian plants, competition from invasive plants and

herbivory by invasive are two of the most important forces responsible

for declining populations (Bruegmann et al. 2002).

1 Herbivory is central to the organization of biotic communities (Marquis 1992;

Janzen 1970). By influencing plant species composition, it also has indirect effects on most other species in a community. While vertebrate herbivores are frequently the focus of studies regarding plant community dynamics (Kotanen

1995; Stone 1985), the impacts of invertebrates are important as well. Among invertebrate herbivores, terrestrial molluscs, such as slugs and snails, are some of the most important grazers of temperate grassland species (del-Val and

Crawley 2004; Hitchmough 2003). Because these animals are known to target seedlings (Hanley et al. 1995), they can have profound consequences for native plant recruitment and therefore adult species composition (Hanley et al. 2002;

Hanley et al. 1996). They may also play an important role in the success or failure of rare plant restoration efforts.

Study organisms

Slugs are a polyphyletic group of terrestrial gastropods descended from snails.

The absence of a shell gives them their distinctive form, which arose by way of convergent evolution in multiple snail lineages (Runham and Hunter 1970). As a result of this evolutionary history, there exists a continuum of gastropod forms in terms of shell reduction, with snails at one end, slugs at the other, and various intermediate forms, sometimes referred to as “semi-slugs”, in between. Thus, the number of gastropod species considered to be slugs is not definitive. To avoid confusion, my use of the term is limited to those families considered to be slugs by South (1992) (Table 1.1).

2

The repeated evolution of the slug form in several gastropod lineages is a testament to its utility, particularly in calcium-poor, wet environments, such as those found on many volcanic islands. Both slugs and snails are hermaphroditic and many species can self-fertilize (Jordaens et al. 2000) as well as tolerate a wide range of temperatures (Rollo and Shibata 1990). These traits make slugs excellent colonizers and potential invaders when introduced into new areas.

Slugs will feed on a variety of foods including carrion, feces, lichens, a variety of small animals and other slugs (South 1992). With the exception of a few geographically restricted groups, however, most slugs are considered to be generalist herbivores (Rathke 1985).

Impacts of slug herbivory

The exclusion of molluscs from plant communities has been shown to have a marked influence on the survivorship of emerging seedlings. Because seedlings are more sensitive than adults to the removal of small amounts of tissue, the risk of mortality due to slugs is highest when the plant is small. Risk is further enhanced by seedling architecture as slugs more readily attack plants less than

10 cm tall (Albrectsen et al. 2004; Rathke 1985) and find young plants more palatable than older ones (Fenner et al. 1999). Hulme (1994) followed the fate of roughly 13,000 seedlings in temperate grassland. Along with rodents, molluscs were among the most important seedling predators, exploiting on average 30%

3 of all individuals. Less is known regarding slugs’ importance as seedling predators in forest habitats. In one of the few studies of this kind, Christel et al.

(2002) showed that seedling recruitment of a perennial forest herb in deciduous forest in Sweden was significantly greater following the application of molluscicide, and that these effects lasted up to four years following a single treatment. In open coniferous forest, Nystrand and Grandström (1997) demonstrated that high densities of the slug Arion subfuscus (Draparnaud)

(Arionidae) were responsible for a three-fold increase in scots pine seedling mortality.

Though death may not occur, the removal of large amounts of photosynthetic tissue and the damaging of reproductive organs can reduce the fitness of low- lying adult plants. In a limestone grassland in central England, for example,

Breadmore and Kirk (1998) found that the two slugs Deroceras reticulatum

(Mϋller) () and Arion ater (Linneaus) (Arionidae) were the main petal herbivores of a wide range of herbaceous plant species, causing damage exceeding that of arthropods. Mollusc herbivory significantly reduced the number of capitulae produced by an invasive aster, Senecio inaequidens D.C., despite the presence of pyrrolizidine alkaloids (Scherber et al. 2003). Rai and Tripathi

(1985) demonstrated that slug and insect herbivory was responsible for delayed flowering and reduced seed output in the weed Galinsoga quadriradiata Cav.

(Asteraceae).

4 Slugs can have profound effects on single species if they are among their

favored food items. Bruelheide and Scheidel (1999) reported that non-native

slugs removed nearly three-fourths of all leaf tissue from the rare perennial

Arnica montana Linneaus (Asteraceae), and, further, restricted this species to

high elevations where herbivory was less intense. In New Zealand alien slugs

caused extensive defoliation of a native slow-growing fern, thereby facilitating the

invasion of alien plant species (Sessions and Kelley 2002).

Several studies have described the short-term effects of slug grazing in

European pasture, noting a general shift away from the use of highly palatable

herbaceous species towards more herbivore resistant grasses (Wilby and Brown

2001; Hanley et al. 1996). Cates and Orians (1975) found the palatability of 100

species to slugs was strongly correlated with successional status. Annuals were preferred to perennials and early successional species were preferred to later successional species, a finding supported by Briner and Frank (1998) using 78 additional plant species. As a result of this work, it has been suggested that molluscs, and especially slugs, can influence both the rate and direction of succession.

Slugs in Hawai‘i

According to the plant-defense hypothesis, plants evolve anti-herbivore defenses

in response to grazing (Feeny 1992). Thus, plants that evolved in the presence of

herbivores are likely to be better defended than plants that evolved in their

5 absence. The degree to which plants develop, or fail to develop, herbivore

defenses is largely a function of the combined effects of resource availability and

herbivore pressure (Coley et al. 1985). In an environment with few herbivores,

evolution would tend to favor fast growing, highly fecund individuals over well-

defended ones. As a result, when non-native herbivores are introduced into a

plant community, species composition can change rapidly and rare species can

be pushed towards extinction (Coomes et. al 2003; Schreiner 1997). This is a

critical issue in Hawai‘i (Stone 1985), where several important guilds of

herbivores were historically lacking and endemism in the flora and fauna is high

because of the islands’ extreme isolation.

Hawai‘i lacks native slugs, but has a rich native snail fauna (Cowie 1995; Gagné

and Christenson 1985). Native snails have not been reported to eat living plant

tissue; tree snails of the Achatinella Swainson (Achatinellidae), for

example, are believed to feed exclusively on epiphytic algae and fungi (Severns

1981; Hadfield and Mountain 1980). The diets of most groups of native snails, however, have yet to be studied (R. Cowie pers. comm.), and it is therefore unknown to what extent native plants are adapted to mollusc herbivory. At least

12 slugs and one semi-slug, martensi Simroth (), are

now established in Hawai‘i (Table 1.2). This number is conservative, as thorough

surveys for slugs in Hawai‘i have not been undertaken. While no formal studies

have been conducted to investigate the impacts that alien slugs are having on

native flora, they are nevertheless widely regarded among local botanists to be

6 key limiting factors to native seedling survival and responsible for the failure of several restoration efforts. The number of endangered Hawaiian plant species reported by the U.S. Fish and Wildlife Service (USFWS) to be imminently or potentially threatened by alien slugs is alarming (Table 1.3).

Two field trials in Lyon Arboretum on O‘ahu (A. Yoshinaga, unpublished) demonstrated that slugs could reduce the survival of Cyanea angustifolia

(Cham.) Hillebr. (Campanulaceae) seedlings by as much as 80%. In one of the few published reports of slug feeding in Hawai‘i, Gagné (1983) documented the slug gagates (Draparnaud) () feeding on the native greensword

Argyroxiphium grayanum (Hillebrand) Degener (Asteraceae). Managers with the

Nature Conservancy and the Army Natural Resource Center constructed elaborate mollusk barriers after plantings of sandwicense Hook. &

Arnott (), Schiedea kaalae Caum & Hosaka, Schiedea obovata

(Sherff) (Caryophyllaceae), Cyrtandra dentata St. John & Storey and Cyanea superba (Cham.) A. Gray (Campanulaceae) apparently failed because of slugs

(Sailer 2002; Arcand et al. 2002).

HYPOTHESES

My study investigates the potential impacts of introduced slugs on native and introduced plants in Hawai’i. The aim of this research is to identify differences (if any) in plant acceptability to slugs in the laboratory as well as assess the effect of slug feeding on plant survival and seedling regeneration under natural conditions.

7 Hypotheses specifically addressed are:

H1) The acceptability of plant material to slugs differs among different plant species

H2) The acceptability of plant material to slugs differs among different slug species

H3) Slug herbivory is responsible for increased damage to leaf tissue

H4) Slug herbivory is responsible for reduced plant growth

H5) Slug herbivory is responsible for increased seedling mortality

H6) Plant species will differ in the above effects (increased leaf damage, reduced growth, reduced survival) of slug herbivory

H7) The number and/or species composition of seedlings naturally regenerating from the seed bank is altered by slug herbivory.

To investigate these hypotheses, two experiments were carried out, one in the laboratory and one in the field. The first addressed H1 and H2 (Chapter 2) and was conducted under controlled conditions using five slug species and five plant species. Plant acceptability was ranked from 0-1 using an index developed by

Dirzo (1980) and differences attributable to slug species, plant species and/or an interaction between the two, were quantified. A second experiment (Chapter 3) carried out in the Wai‘anae Mountains, followed the fate of transplanted seedlings (H3-H6), and tracked changes in the number and identity of germinants from the seed bank (H7), when exposed to or protected from slug herbivory.

8 TABLES

Table 1.1 Gastropod families containing slugs (South 1992; new family names from Barker 2001). Class Subclass Order Familiy Gymnomorpha Soleolifera Rathouisiidae Vaginulidae Milacidae Agriolimacidae Boettgerillidae Arionidae Philomycidae

9 Table 1.2. List of slugs found within the state of Hawai‘i, island distribution and year of first record. Species Dist. in Hawai‘i Year first rec. Native range Arion distinctus Hawai‘i (D. Foote, pers. 2003 (D. Foote, Natural range in Europe Mabille comm.) pers. comm.) not known (Barker 1999) A. hortensis Hawai‘i (D. Foote, pers. 2003 (D. Foote, Natural range in Europe Férussac comm.) pers. comm.) not known (Barker 1999) A. intermedius Hawai‘i (Cowie 1999) 1998 (Cowie Central and western Normand 1999) Europe (Barker 1999) Deroceras Hawai‘i, Kaua‘i, Maui, 1897 (Cowie Palearctic (Barker 1999) leave (Mϋller) O‘ahu (Cowie 1997) 1997) D. reticulatum Hawai‘i, Kaua‘i (Cowie 1963 (Cowie Natural range in Europe (Mϋller) 1997), O‘ahu (S. Joe pers. 1997) not known (Barker 1999) obs.) alte Hawai‘i, Moloka‘i, O‘ahu 1900 (Cowie Central Africa (Cowie (Férussac) (Cowie 1997) 1997) 1997) Lehmannia Maui (Cowie 1997), O‘ahu 1982 (Cowie Iberian Peninsula and valentiana (S. Joe pers. obs.), Hawai‘i 1997) Atlantic Islands (Barker (Férussac) (D. Foote pers. comm.) 1999) Limacus flavus Maui, O‘ahu (Cowie 1997) 1948 (Cowie Natural range in Europe (Linnaeus) 1997) not known (Barker 1999) Limax maximus Hawai‘i, Maui, O‘ahu 1931 (Cowie Southern and western Linnaeus (Cowie 1997) 1997) Europe (Barker 1999) Meghimatium Kaua‘i, O‘ahu (Cowie 1846 (Cowie (Cowie 1997) striatum van 1997) 1997) Hasselt Maui, Hawai‘i (Cowie 1896 (Cowie Canary Islands, (Draparnaud) 1997), O‘ahu (Arcand et al. 1997) Mediterranean and Black 2002) Sea region (Barker 1999) Parmarion O‘ahu (Cowie 1997, 1998) 1996 (Cowie Cambodia (Cowie 1997) martensi 1997) Simroth Hawai‘i, O‘ahu (Cowie 1978 (Cowie Central and South America plebeia 1997) 1997) (Rueda et al. 2002) (Fischer) O‘ahu (Cowie 1997) 1985 (Cowie (Cowie 1997) cubensis 1997) (Pfeiffer)

10 Table 1.3. List of native species reported by the USFWS to be currently or potentially threatened by alien slugs. Quotation marks are not used in the comment column, rather, any comments that are not direct quotes but my own interpretation of the material are enclosed in brackets [].

Species Family Comment regarding slugs Citation

Acaena exigua A. Gray .. consumption of vegetative or floral parts of this species by alien USFWS 1997, p. slugs and/or could have been a factor in the decline of the 10 species and could continue to be a critical limiting factor. Clermontia oblongifolia Campanulaceae [Slugs appear in a table of potential threats to this species.] USFWS 1997, Gaud. Table 1, p. 9 C. samuelii C. Forbes Campanulaceae .. slugs (mainly Milax gagetes) are known to eat leaves, stems, and USFWS 1999, p. fruits of other members of this genus, and therefore are a potential 48309 threat. Cyanea Gaud. Campanulaceae Little is known about the predation of certain rare Hawaiian plants by USFWS 1998c, alien snails and slugs. Field botanists have observed slugs preying pp. 27-28 indiscriminately on plants belonging to the bellflower family. The effect of these alien predators on the decline of Cyanea species (which are in the bellflower family) and related species is unclear, although slugs may pose a threat by feeding on seedlings, stems, and fruit, thereby reducing the vigor of the plants and limiting regeneration. C. acuminata (Gaud.) Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Hillebr. Table 3, p. 29 C. asarifolia St. John Campanulaceae Common garden slugs have become widespread in and may USFWS 1995b, pose a major, but little-recognized threat to native plant species. p. 14 Slugs can damage flowers, fruits, stems and seedlings of native plants. Slug damage has been documented on Cyanea asarifolia and slugs are a potential threat to other Kauai cluster taxa. C. asplenifolia (Mann) Campanulaceae This species is threatened by ... slugs that directly prey upon and USFWS 2005, p. Hillebr. defoliate the species. 24912

11 C. calycina (Cham.) Campanulaceae Threats to the species include pigs and goats that degrade and USFWS 2005, p. Lammers destroy habitat, rats and slugs that directly prey upon it. 24880 C. copelandii spp. Campanulaceae .. slugs (mainly Milax gagates) are known to eat leaves, stems, and USFWS 1999, p. haleakalaensis (St. fruits of other members of this genus, and therefore are a potential 48310 John) Lammers threat to this species. C. crispa (Gaud.) Campanulaceae The major threats to [this species] are .. suspected predation by .. USFWS 1998c, Lammers, Givnish & slugs. p. 57-58 Sytsma C. dunbariae Rock Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998b, Table 3, p. 7 C. eleeleensis (St. John) Campanulaceae This species is highly threatened by ... slugs that eat this plant. USFWS 2005, p. Lammers 24912 C. glabra (F. Wimmer) Campanulaceae Slugs are the primary threat to C. glabra, shown by recent USFWS 1998b, St. John observations of slug damage on both juveniles and adults. p. 11 C. grimesiana Campanulaceae The major threats to [this species] are .. predation of seeds or fruits USFWS 1998c, subsp. obatae (St. John) by introduced slugs. p. 59 Lammers C. humboldtiana (Gaud.) Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 C. koolauensis Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Lammers, Givnish & Table 3, p. 29 Sytsma C. kuhihewa Lammers Campanulaceae This species is highly threatened by ... slugs that eat this plant. USFWS 2005, p. 24913 C. kunthiana Hillebr. Campanulaceae Slugs have also been observed on C. kunthiana plants, which had USFWS 1998b, extremely damaged leaves. p. 11 C. lanceolata (Gaudich.) Campanulaceae Threats to the species include pigs, rats, and slugs that prey USFWS 2005, p. Lammers upon [it]. 24880 C. lobata H. Mann Campanulaceae [Slugs appear in a table of potential threats to this species.] USFWS 1997, 12 Table 1, p. 9 C. longiflora (Wawra) Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 C. mceldowneyi Rock Campanulaceae [Slugs appear in a table of potential threats to this species.] USFWS 1997, Table 1, p. 9 C. pinnatifida (Cham.) F. Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Wimmer Table 3, p. 29 C. tritomantha A. Gray Campanulaceae Threats to this species include pigs, rats, and slugs that eat this USFWS 2005, p. plant. 24880 C. obtusa (Gray) Hillebr. Campanulaceae This species is highly threatened by goats, pigs, cattle, rats, USFWS 2005, p. and slugs that eat this plant. 24913 C. recta (Wawra) Hillebr. Campanulaceae The major threats to [this species] are .. slugs that feed on the USFWS 1998a, stems. p. 25 C. remyi Rock Campanulaceae Reasons for the decline of [this species] are .. slugs that feed on the USFWS 1998a, stems. p. 28 C. st.-johnii Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 C. superba (Cham.) A. Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Gray Table 3, p. 29 C. truncata (Rock) Rock Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 C. undulata (C. Forbes) Campanulaceae Introduced slugs have .. been observed feeding on stems and USFWS 1994, p. leaves. 22 Cyrtandra J. R. Forester Gesneriaceae Members of the genus Cyrtandra in the family Gesneriaceae are USFWS 1998c, & G. Forester also thought to be susceptible to slug predation. p. 28 C. crenata St. John & Gesneriaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Storey Table 3, p. 29

13 C. dentata St. John & Gesneriaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Storey Table 3, p. 29 C. kaulantha St. John & Gesneriaceae Threats to the species include pigs and slugs that eat this plant. USFWS 2005, p. Storey 24880 C. polyantha C. B. Gesneriaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Clarke Table 3, p. 29 C. subumbellata Gesneriaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, (Hillebr.) St. John & Table 3, p. 29 Storey C. viridiflora St. John & Gesneriaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Storey Table 3, p. 29 Delissea rhytidosperrna Campanulaceae .. other threats are predation by rats and slugs. USFWS 1995b, H. Mann p. 39 D. subcordata Gaud. Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 Lobelia gaudichaudii Campanulaceae [Slugs appear in a table of potential threats to this species.] USFWS 1998c, subsp. koolauensis Table 3, p. 29 (Hosaka & Fosb.) Lammers L. monostachya (Rock) Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Lammers Table 3, p. 29 L. niihauensis St. John Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 L. oahuensis Rock Campanulaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 Ranunculus hawaiensis Ranunculaceae Threats to the species include competition from nonnative plants, USFWS 2005, p. Gray and damage from slugs. 24884 Ranunculus mauiensis Ranunculaceae This species is threatened by feral pigs and slugs that eat this plant. USFWS 2005, p. 24922 14 Schiedea Ranunculaceae Based on recent unpublished evidence, recruitment of Schiedea USFWS 1997, p. germinants [seedlings] can be catastrophically suppressed by 85 herbivory of alien slugs in the Wai‘anae Mountains of O‘ahu. Schiedea Ranunculaceae There is no evidence of regeneration from seed under field USFWS 1998a, conditions. Seedlings of species of Schiedea occurring in mesic or p. 59-60 wet sites are apparently consumed by introduced slugs and snails ... In contrast to mesic-forest species, Schiedea occurring in dry areas produce abundant seedlings following winter rains, presumably because there are fewer alien consumers in the drier sites. Schiedea haleakalensis Caryophyllaceae [This species] has survived only on precipitous cliff faces USFWS 1997, p. inaccessible to goats. In spite of the removal of goats in the late 84 1980s from habitat of this taxon in Haleakalā National Park, no establishment by seedlings has ever been observed. Slugs may be completely devouring the seedlings. S. kaalae Wawra Caryophyllaceae This species .. reproduces prolifically under greenhouse conditions. USFWS 1998c The lack of seedlings in the field seems, therefore, almost certainly to be the result of grazing by alien snails and slugs. S. kauaiensis St. John Caryophyllaceae Threats to [this species] include ... predation by introduced slugs. USFWS 1998a, p. 60 S. kealiae Caum & Caryophyllaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Hosaka Table 3, p. 29 S. membranacea St. Caryophyllaceae There was no evidence of recruitment in the population, despite the USFWS 1998a, John production of abundant seed during all years of observation (1987, p. 63 1994-1997). Introduced snails have been observed feeding on flowers and developing seed capsules, and garlic snails are common near the plants. It seems very likely that introduced molluscs are responsible for the failure of recruitment. S. stellarioides H. Mann Caryophyllaceae [Slugs appear in a table of current threats to this species.] USFWS 1998c, Table 3, p. 29 Viola lanaiensis Becker Violaceae Slug damage and live slugs have been observed on [this species]. USFWS 1995a, p. 61

15 CHAPTER 2

PALATABILITY OF FIVE NATIVE HAWAIIAN PLANT SPECIES TO FIVE ALIEN

SLUG SPECIES

INTRODUCTION

Hawai‘i has a very high number of endangered plants, with many species

reduced to small and dwindling populations (Wagner et al. 1999). One of the

most important threats to these disappearing species is predation by introduced

herbivores. Introduced ungulates, for example, have devastated plant

communities in a number of natural areas in Hawai‘i and elsewhere (Coblentz

and Van Vuren 1987; Van Vuren and Coblentz 1987; Stone 1985). Terrestrial

slugs (: Pulmonata) are another group of introduced herbivores with no

native representatives in Hawai’i (Cowie 1997). Slugs constitute a polyphyletic

group distinguished from snails by the absence of an external shell (Runham and

Hunter 1970). They are important seedling predators in native (Wilby and Brown

2001; Hulme 1994) and introduced habitats (Ferguson 2004; Fritz et al. 2001) in

many parts of the world. In Hawai‘i, there is copious anecdotal information

suggesting that they harm native plants yet, to date, they have been the focus of

little research.

Theories of plant defense evolution hinge on the assumption that the resistance

conferred by defense mechanisms involves an allocation of resources which, in

the absence of herbivores, could be put towards other fitness-enhancing ends

16 (Coley et al. 1985). Plant defense mechanisms include structural defenses like

spines or hairs that can interfere with feeding (Pollard 1992), as well as chemical

defenses in the form of secondary compounds that can influence plant

palatability or digestibility (Arnold 1980). Although the cost of defense is not always obvious (Siemens et al. 2002), a recent synthesis of 33 studies supports the idea of inherent trade-offs between allocation to plant defense and other fitness-enhancing activities such as growth and reproduction (Strauss et al.

2002). Because Hawaiian plants evolved in the absence of slugs, it could be hypothesized that most native plants will be highly vulnerable to slug herbivory due to the loss of unnecessary defense mechanisms and hence increased palatability. In the Santa Cruz Islands, for example, Bowen and Van Vuren

(1997) found higher acceptability to domestic sheep and lower chemical

defenses in plant species relative to mainland congenerics. But Hawaiian plants

have had to defend themselves against arthropod and avian herbivores (Givnish

et al. 1994), and some of the mechanisms that protect them from these

consumers may also be effective against novel ones. In addition, phylogenetic

inertia may dictate that certain groups of native plants use different defenses, or

retain them for longer periods of evolutionary time, than others. It has long been

recognized, for example, that introduced ungulates prefer certain native plants to

others (Scowcroft and Giffin 1983) suggesting rejected species may have

retained defenses. Likewise, native Hawaiian plants may not be equally attractive

to alien slugs, and further, food preferences may differ among different slug

species.

17 Variability in the feeding relationships between native plants and alien slugs should have important implications for conservation in Hawai’i. Knowledge of which plant species are most palatable, and which slug species are most likely to eat them, will allow more informed management decisions regarding the protection and restoration of endangered plant populations. As a first effort to establish this base of information, I investigated five endangered plant species and five introduced slug species, and used relative acceptability indices (Dirzo

1980) to assess whether there are differences among these taxa in either palatability or feeding preference.

MATERIALS AND METHODS

Study organisms

Cyanea Gaud. (Campanulaceae), the largest genus of plants endemic to the

Hawaiian archipelago, once contained 52 species, of which 14 are now extinct and 15 are listed as endangered by the U.S. federal government (Wagner et al.

1999). Slugs have been observed to eat the leaves, stems and fruits of many

Cyanea species (Chapter 1) and are suspected of contributing to their decline.

Cyanea grimesiana Gaud. ssp. obatae (St. John) Lammers is a , usually unbranched, 1-3.2 m in height, typically found on steep, moist, shaded slopes in diverse mesic to wet forests between 550 and 670 m elevation (Wagner et al.

1999). Endemic to the Wai‘anae Mountains on the island of O‘ahu, only 13 individuals of this subspecies remain in the wild (USFWS 1998c).

18 Brighamia A. Gray, also in the Campanulaceae, contains two species, both

endangered. Brighamia rockii St. John is a stout unbranched succulent, 1-5 m

tall, currently restricted to steep, inaccessible sea cliffs along East Moloka‘i’s

northern coastline between sea level and 470 m elevation (Wagner et al. 1999).

Fewer than 200 plants occur in the wild (USFWS 1996). Slugs may not be a

serious concern for this species if they do not forage on sea cliffs. Nonetheless,

this plant may have had a broader habitat range in the past, and introduced slugs

could have played a role in restricting its range. Slugs readily consume plants

grown under greenhouse conditions and are known to kill adult plants by

burrowing into the fleshy stem and hollowing out the center (M. Sherman pers.

comm.; K. Swift pers. comm.).

Slugs have been observed to feed readily on a number of species of Scheidea

Cham. & Schlechtend., non A. Rich. nec Bartl (Chapter 1), a speciose endemic

genus. Scheidea obovata (Sherff) (Caryophyllaceae) is a branching subshrub

attaining heights of approximately 1 m and possessing thick, fleshy, deep green

leaves that are either opposite or whorled about the stem. This species occurs in

diverse mesic forest from 550 to 800 m elevation on ridges and slopes in the

Wai‘anae Mountains (Wagner et al. 1999). In 1994, the three known wild

populations contained 11 individuals in total (USFWS 1994).

Urera kaalae Wawra (Urticaceae) is a small tree, 3-7 m tall, with distinctive greenish sap that turns black when exposed to the air and has thin, heart-

19 shaped, pale green leaves, serrate at their margins (Wagner et al. 1999).

Endemic to the Wai‘anae Mountains, this species grows on slopes and gulches in diverse mesic forest at elevations of 300-820 m. About a decade ago, an estimated 40 plants remained in the wild (USFWS 1994). In an effort to prevent further decline, the Nature Conservancy has been augmenting natural populations with greenhouse seedlings but reports that slug predation of these plants is high (Sailer 2002).

Solanum sandwicense Hook & Arnott (Solanaceae) is a shrub, 4-5 m tall, with simple unarmed leaves, occurring in diverse mesic forest between 760 and 1220 m elevation on O‘ahu and Kaua‘i. Young parts are densely pubescent and reddish pubescence coats the underside of the leaves (Wagner et al. 1999). The

O‘ahu population, which in 1995 shrank to one plant (USFWS 1995b), now numbers about 10 thanks to captive propagation and outplanting of seedlings by

Nature Conservancy staff.

The five slug species used in the present study are among the most common on

O‘ahu, and at least three of them co-occur with the plant species tested here.

Between 30 and 80 slugs of each species were collected between May and June

2003 from two areas on O‘ahu: at 700 m elevation in the Kahanahāiki

Management Unit (KMU), a preserve managed by the U.S. Army in the northern

Wai‘anae Mountains; and at 60 m elevation in the Kaimuki residential district of

Honolulu (Figure 2.1). Due to an apparent stratification in the distribution of these

20 slugs, the two veronicellid species: (Férussac) and Veronicella

cubensis (Pfeiffer) were more common at elevations < 300 m, in residential areas, while the other three, Limax maximus Linnaeus (Limacidae), Limacus flavus (Linnaeus) (Limacidae), and Meghimatium striatum van Hasselt

(Phylomicidae) were more common in forested areas > 300 m (S. Joe pers. obs.). Despite these differences, all species survived well in Kaimuki, where they were housed outdoors in the shade and exposed to natural weather and light conditions through February 2004, after which they were killed.

Feeding trial protocol

Following capture, each slug was weighed and housed in a single 266 ml plastic

cup containing a substrate of fine cinder 5 cm deep. A mesh top secured by a

rubber band prevented escape. Holes punched into the bottom of the cups

allowed for water drainage and a wet cloth placed over the top prevented

dehydration. Study slugs probably included both juveniles and adults, as maturity

was not confirmed prior to the feeding trials. Species identifications were made

using external morphology.

Slugs can remember unacceptable food for up to three weeks after a single

encounter (Gelperin 1975); therefore, an interval of at least three weeks was

allowed between feeding trials. In the intervals, slugs received a diet of Beneful

® (Nestlé Purina PetCare Company, St. Louis, Missouri, USA) dog kibble and

Aloha brand (Aloha Seed & Herb Paia, Hawai’i, USA) buttercrunch lettuce

21 (Lactuca sativa L.). Palatability testing began in August 2003 and ended 6 months later.

An acceptability index (AI) was used to measure plant palatability (Dirzo 1980).

On the day of testing, each slug was weighed, placed in a fresh cup containing only moistened filter paper and a pair of 15 cm2 leaf discs, one from the test species and one from a highly acceptable control species, in this case, lettuce, and allowed to feed for 48 hours, after which they were returned to their original cups. Although Dirzo (1980) allowed slugs to feed for only 12 hours, they were starved 24 hours prior to testing. In order to reduce the labor required to transfer slugs into new containers, slugs were not starved prior to testing as in Dirzo

(1980). Instead the duration of each trail was increased. Uniformity of disc size was maintained using a hole puncher, and measurement of the leaf surface area both before and after slug feeding was made with a LI-3000A Portable Area

Meter (LI-COR Environmental, Lincoln, Nebraska USA). The index was calculated as the area eaten from the test species leaf disc divided by that eaten from the lettuce disc. Thus, in order for a trial to be considered valid, some or all of the lettuce disc must be consumed by the slug. Instances in which the slug ate neither disc were discounted, along with instances in which the slug ate only the test disc. Multiple test plant species were then ranked by relative palatability according to their AI scores. AI scores have the benefit of controlling for slug size since the ratio remains the same for a small slug eating a small amount of food and a larger slug eating a large amount.

22

All plants with the exception of B. rockii, which came from the Hui Ku Maoli Ola

Native Hawaiian Plant Nursery in Waimanalo, O‘ahu, and lettuce, which was

grown by myself from seed, were borrowed from the Nature Conservancy’s rare

plant nursery in Wahiawa. Plant age can affect palatability (Fenner et al. 1999);

thus I only harvested mature leaves from adult (i.e. reproductive) plants for use in

feeding trials. No more than three plants per species (not including lettuce, which

came from approximately 12 plants) served as leaf disc donors because of their

extreme rarity. Intraspecific variation in plant palatability, if it exists, was not

addressed in this study. After feeding trials had concluded, plants were returned

to the Wahiawa nursery for use in restoration projects. None died as a result of

leaf harvesting.

Because test plant material was limited to a one-time harvest, it was necessary

to present each plant species to all slugs on a single occasion rather than

randomizing the order of plant presentation for each slug. Thus, the order of plant

species presentation, determined randomly, and commencing on the date in

parentheses, was as follows: C. grimesiana ssp. obatae (27 August 2003), S. sandwicense (27 September 2003), U. kaalae (31 October 2003), B. rockii (26

November 2003), and S. obovata (28 December 2003). Because of the long interval between feeding trials, some slugs died before being offering all five plant species. Because these were not replaced, there was a very slight attrition in slug number over the study period. Mortality did not exceed two slugs per

23 species, a remarkably low number considering reported losses of greater than

50% over three weeks for other studies of this type (Gebauer 2002; Glen et al.

2000; Wilson et al. 1999).

Statistical Analysis

Because slug size was expected to differ among species, differences in slug weights were analyzed using a one-way ANOVA. The effects of plant species, slug species, and plant x slug interaction on AI scores were analyzed with a two- way ANOVA, with all factors fixed. The AI scores did not meet all the assumptions of ANOVA, that is, residuals were not normally distributed and were heteroscedastic. However, variances among the different groups were not significantly different (Levene’s Test; P=0.243). Log transformation of the data yielded results almost identical to those of the untransformed data and did not improve the distribution of the model’s residuals. I judged that this violation of residual normality was preferable to using a combination of tests on subsets of the data. All statistical analyses were performed with Minitab® Release 14 software (Ryan et al. 2005).

RESULTS

Feeding activity was low during each of the five trials regardless of slug species, with, on average, only one in every five slugs feeding (Fig. 2.2). This allowed estimation of AI scores based on sample sizes ranging from four to 20 AI replicates for each slug\plant species pair (Table 2.1). An AI of 0 indicates

24 complete rejection of the test species. A two-way ANOVA identified a significant

difference in palatability among plant species, but no statistical difference in

overall feeding among slug species, and no plant-by-slug species interaction

(Table 2.2). Post-hoc testing indicated that palatability differences among plant

species were significant (Tukey’s HSD) only between U. kaalae and the other

four plant species, with U. kaalae being more palatable than the other species

(Figure 2.3). The high frequency of AI scores close to zero (Figure 2.4) shows most of the test plant species offered to slugs were rejected and were not nearly as acceptable as the control.

As expected, slug taxa differed significantly by weight (one-way ANOVA;

F=42.23; P<0.0001), with groups differing at the family level (Figure 2.5). The two limacid species were heaviest, followed by the two veronicellids in the midrange and M. striatum the lightest of the group.

DISCUSSION

The results of this experiment indicate that there is significant variation in palatability among native Hawaiian plants, with U. kaalae significantly more palatable than the other four plant species tested. Urera kaalae was considerably more palatable, even, than the two Campanulaceae species (C. grimesiana and

B. rockii), a plant family previously noted to be especially attractive to slugs in

Hawai‘i (Sailer 2002).

25 Somewhat surprisingly, there was no significant difference among slug species in food preference. This may reflect similarities these slugs share as colonizing species, such as non-specialist feeding behavior stemming from the absence of tightly co-evolved relationships with local plants. Alternatively, small sample sizes, combined with high variation among trials, may have obscured more subtle differences in plant preference among the slug species.

Even though I started with relatively large numbers of slug replicates, the majority of trials resulted in no feeding and had to be discounted. While failure to eat caused many trails to be omitted from analysis, no omissions occurred as a result of slugs eating the test plant but not the control. In order to account for this possibility, others (Kelly and Hanley 2005) have calculated AI from the area of the test disc eaten divided by the area of both discs eaten. Feeding might have been observed in a greater proportion of trails by starving the slugs for a period of time prior to testing or by increasing the length of each trail to more than 48 hours. Gebauer (2002), for example, exposed test foods to slugs for 7 days. In addition, I suspect the slugs were overfed prior to and in between testing, as all gained weight and continued to gain weight during their months in captivity.

Perhaps more importantly, the data reveal large variation in feeding preferences by individual slugs. For example, a single L. flavus ate 31 times more U. kaalae than lettuce, but failed to eat anything in any of the other trials. To a lesser degree, other individual slugs showed distinct preferences that may, or may not, have been in line with the average for their species. Other studies have typically

26 dealt with this problem by omitting “anomalous” (Brooks et al. 2003) or “deviant”

(Dirzo 1980) slugs from the analysis. The frequency of such behavior, however,

imply that while there may be valid preferences for a slug species as a whole,

there is a lot of meaningful intraspecific variation that should be recognized.

The distribution of AI scores close to 0 (Figure 2.5) was similar to that observed

by Dirzo (1980) who found that slugs rejected, on average, 42% of the 30 plant

species offered. He went on to characterize slug food preferences as falling

within one of the following three categories: rejected or rarely chosen, moderately

acceptable and very acceptable. Why so many species were rejected for rarely

chosen remains the subject of much speculation. Slugs have complex feeding

behavior. Food choice is driven not only by metabolic requirements and plant

defenses, but is also subject to more subtle influences such as the prior

experience of the animal (Gouyon et al. 1983). It has been demonstrated, for

example, that the slug Limax maximus is capable of associative learning that

subsequently guides its food choices (Sahley et al. 1981). Food preferences

among slugs are not static, and depend in part upon the availability of other

foods and the frequency with which they are encountered. Slugs often prefer

novel foods over familiar ones (Cottam 1985; Mølgaard 1986a), and may consume up to 270% more when offered a choice of foods (Peters et al. 2000).

This behavior has been described as “neophilic” by some authors (Cook et al.

2000) and even extends to foods that are known to contain noxious compounds.

Slugs tend to eat such foods too, at least until the novelty wears off (Whelan

27 1982). Thus, a rare plant may be more likely to be attacked by slugs just by virtue

of being rare.

The level of risk incurred by a native plant will largely depend on how palatable it

is relative to other available plants. Unfortunately, few of these neighbors are

likely to be lettuce. Though S. obovata had the lowest mean acceptability among

the species tested, the area where it occurs is dominated by invasive strawberry

guava, Psidium cattleianum Sabine. Herbivore damage to P. cattleianum in the

field is negligible (S. Joe pers. obs.). Leaf extracts from Psidium guajava (L.), a

related species, contain biologically active chemical compounds which can have

an analgesic effect (Somchit et al. 2004) and disrupt reproduction in mammals

(Lapçik et al. 2005) and these may have strong anti-herbivory properties.

Whether seedling tissue in my test species is more acceptable to slugs than adult

tissue, something that has been found to be true for other species (Fenner et al.

1999), remains unknown; however, anecdotal reports suggest that adult plant

parts are vulnerable as well (see Study Organisms Materials and Methods).

Because of the rarity of plant species tested, and because the harvesting of seedling tissue would require the destruction of plants, only adult tissue was used in the present study. In order to explore differences in adult and seedling palatability, future studies should draw from a pool of non-endangered plant species the seedlings of which can be sacrificed.

28 The complex behavior of slugs and the variation in plant community composition

make it difficult to extend results from lab-based preference trials to feeding

behavior in nature. At the least, this study suggests that all five of the native

species tested are likely to be fed upon to some degree in nature. It also

identifies U. kaalae as a species that may be exceedingly vulnerable to slug herbivory. The predictive power of palatability trials such as this one, however, could be strengthened if correlated on a broad scale with plant defense structures and species-specific secondary compound content. For example, pubescence has been shown to deter slug herbivory in some plant species

(Westerbergh and Nyberg 1995). The pubescence on S. sandwicense may have

reduced its acceptability, but it clearly did not prevent slug feeding entirely. A

number of secondary compounds are also known to deter slug herbivory (Table

2.3). Swanholm et al. (1959; 1960) found alkaloids in a significant number of

Hawaiian plants in various families, including the Campanulaceae species

Cyanea angustifolia and Clermontia kakeana. Despite this finding, the two

Campanulaceae species were eaten by slugs in this study, and Cyanea suberba

is heavily fed upon by slugs in field trials (Chapter 3). It may be that these

alkaloids are not especially effective in deterring slug herbivory, or there may be

large intrafamilial or even intrageneric differences in chemical content. With

additional research, some of these questions could be resolved.

29 TABLES

Table 2.1. Mean acceptability indices (AI) for 25 slug-plant species pairs. N equals the number of slugs that consumed any part of one or both leaf discs. In order for a plant species to be completely rejected (AI=0) all or a portion of the lettuce must have been consumed. Slug species Plant species N AI sd ± Laevicaulis alte Brighamia rockii 9 0.43203 0.9157 Cyanea grimesiana 16 0.39834 0.5158 Schiedea obovata 7 0 0 Solanum sandwicense 16 0.09772 0.1614 Urera kaalae 13 1.08496 1.1489 Limacus flavus Brighamia rockii 9 0.28909 0.8673 Cyanea grimesiana 7 0.04579 0.0682 Schiedea obovata 9 0 0 Solanum sandwicense 8 0.24592 0.3174 Urera kaalae 10 3.74690 10.3447 Limax maximus Brighamia rockii 7 0 0 Cyanea grimesiana 9 0.08891 0.1077 Schiedea obovata 6 0 0 Solanum sandwicense 14 0.36335 0.4483 Urera kaalae 10 0.34475 0.4437 Meghimatium striatum Brighamia rockii 17 0 0 Cyanea grimesiana 10 0.14383 0.3247 Schiedea obovata 20 0.223947 0.7341 Solanum sandwicense 15 0.07970 0.1109 Urera kaalae 16 0.28902 0.8327 Brighamia rockii 7 0.60608 0.3270 Cyanea grimesiana 4 0.13342 0.2203 Schiedea obovata 6 0.03685 0.0824 Solanum sandwicense 8 0.17355 0.1434 Urera kaalae 7 1.77946 1.0605

30 Table 2.2. ANOVA using AI scores from 25 slug-plant species pairs. Source of Adj MS df F-ratio P variation Slug species 4.127 4 1.03 0.391 Plant 16.236 4 4.06 0.003 species Slug x plant 4.275 16 1.07 0.386 species Error 4 231

Table 2.3. Secondary plant compounds repellent to slugs. Slug Plant Repellent Reference Arion subfuscus Salix eriocephala condensed tannins Albrectsen et al. 2004 Deroceras reticulatum Conium maculatum, alkaloids Birkett et al. 2004 Coriandrum sativum, Petroselinum crispum Deroceras reticulatum Trifolium repens cyanogenic Raffaelli and Mordue glycosides 1990 Deroceras reticulatum Plantago major ssp. caffeic acids Mølgaard 1986b Arion ater pleiosperma Deroceras reticulatum Thymus vulgaris monoterpenes Gouyon et al. 1983

31 FIGURES

Figure 2.1. Slug collection sites. Laevicaulis alte and Veronicella cubensis were collected in Kaimuki, while the remainder were collected in Kahanahāiki Management Unit, a forested area managed by the U.S. Army in the Wai‘anae Mountains.

32

Figure 2.2. Percent of slugs (bars are one SEM) engaged in feeding (i.e. ate all or a portion of the control species), averaged among five feeding trials.

33

Figure 2.3. Mean AI scores (bars are one SEM) for each plant species (all slug species pooled). The asterisk (*) indicates that this group differs significantly (P<0.05) from the rest (Tukey’s HSD).

34

Figure 2.4. Distribution of AI scores (all slug species and plants combined) (single data point of AI = ~31 not shown).

35

Figure 2.5. Boxplot showing weight distribution of slug species. Outliers are marked with asterisks (*). Circles indicate means.

36 CHAPTER 3

IMPACT OF ALIEN SLUGS ON NATIVE PLANT SEEDLINGS IN A DIVERSE

MESIC FOREST, O‘AHU, HAWAI‘I

INTRODUCTION

Plant seedling survival in a natural environment depends on a number of factors, including growth rate, level of competition for light and nutrients, and the magnitude of herbivory. The relative rate of seedling survival among species, in turn, is a key factor influencing the composition of a plant community (Fritz et al.

2001, Buckland and Grime 2000). Herbivores that target seedlings, therefore, affect not only individual plants and plant species, but also influence the make-up of mature plant assemblages. In response to herbivory, plants often protect their seedlings using defense mechanisms, such as the production of secondary compounds that deter feeding. This evolutionary interplay between plants and seedling herbivores is a central dynamic shaping biotic communities. It follows that the introduction of novel herbivores may have profound consequences that reverberate throughout the community.

Slugs are generalist herbivores (Rathke 1985) that feed principally on plant seedlings and low-lying herbs, yet they are not completely indiscriminate in their choices of foods (Dirzo 1980). In their native ranges, slugs and other molluscs are known to be important herbivores that influence seedling survival and plant community species composition. In European grasslands, molluscs, and

37 especially slugs, affect seedling survival (Hulme 1994), shift relative abundances of palatable versus resistant species (Hanley et al. 1996; Wilby and Brown 2001), and may influence both the rate and direction of plant succession (Briner and

Frank 1998, Cates and Orians 1975). In forest habitats, Christel et al. (2002), found that seedling recruitment of a perennial forest herb was significantly greater following application of molluscicide, and Nystrand and Grandström

(1997) demonstrated that high densities of the slug Arion subfuscus (Drap.) were responsible for a three-fold increase in Scots pine seedling mortality. Even among low-statured adult plants, the removal of large amounts of photosynthetic tissue and the damaging of reproductive organs by slugs can reduce plant fitness

(Scherber et al. 2003, Breadmore and Kirk 1998, Rai and Tripathi 1985).

Hawai‘i has no native slugs, but has a rich native snail fauna (Cowie 1995;

Gagné and Christenson 1985). Native snails in the Achatinellidae and

Amastridae are thought to eat decaying plant material and fungus (Severns

1981; Hadfield and Mountain 1980). Outside these families, little is known regarding the feeding behavior of native snails (Cowie pers. comm.).

Nonetheless, there is presently no evidence that native snails eat live vascular plants. The dozen or more species of slugs now established in Hawai’i (Chapter

1), therefore, may represent an entirely new guild of herbivores. Although native

Hawaiian plants have had to defend themselves against avian, insect and possibly snail herbivory, the defense mechanisms evolved by Hawaiian plants may not be very effective against introduced slugs. Slugs, in fact, are widely

38 regarded by local botanists as key limiting factors in native seedling survival, especially among the Campanulaceae, and are believed to be responsible for the failure of restoration efforts (Chapter 1). Detailed research of these impacts in

Hawai‘i, however, is lacking.

Because of their ability to shift plant species composition through selective feeding, slugs could serve as an important pressure favoring alien plants over native species. In New Zealand, for example, alien slugs caused extensive defoliation in a slow-growing native fern, thereby facilitating the invasion of alien plant species (Sessions and Kelley 2002). In Hawai‘i, slugs were found to prefer certain native plants over others (Chapter 2), and, for the following reasons, it seems likely that palatability asymmetries will be magnified with respect to naturalized alien plants. Most introduced plant species have evolved with slugs, and may therefore be better-defended and less palatable, in general, than their now sympatric native competitors. Although other factors, such as life history traits, will influence competitive outcomes between native and alien plant species, seedling herbivory by slugs may be important.

I used a replicated experimental design to investigate the effects of slug herbivory on native and alien seedling growth and survival in a natural field setting on O‘ahu, Hawai‘i. I chose three native and two alien plant species that occur at the field site, and measured plant growth, degree of feeding damage and overall survival of outplanted seedlings in slug-accessible and slug-free

39 plots. I also monitored natural plant germination and growth from the seed bank in the replicated plots. Finally, I monitored slug species composition and abundance over the course of the six-month study in order to document ambient slug population levels at this site.

MATERIALS AND METHODS

Field Site

Slugs are believed to be major seedling predators of several rare and endangered plant species on military land in Hawai‘i under management by the

U.S. Army Garrison Natural Resources Division (AGNRD) (Arcand et al. 2002).

Of particular biological significance are areas encompassed by the Kahanahāiki

Management Unit (KMU), which was therefore chosen for this study. KMU is situated at 700 m elevation on the northeast rim of Mākua Valley, in the Wai‘anae

Mountains on the island of O‘ahu (Figure. 3.1). This area is classified as Montane

Diverse Mesic Forest (Gagné and Cuddily 1999) as it receives approximately

3000 mm of rainfall annually (Arcand et al. 2002).

The KMU harbors 12 endangered plant species, two endangered animal species and is the site of the first endangered species restoration effort by the military in

Hawai‘i. In consequence, AGNRD staff have taken steps to minimize damage from alien ungulates have been taken, including the construction of a fence in

1996 and the removal of feral goats and pigs within the fenced enclosure.

Rodents are controlled using snap traps and bait stations arranged in a grid

40 throughout the enclosure. Such activities are important to the present study as they reduce the level of herbivory and plant deaths not attributable to slugs.

The predominant canopy tree is the alien Psidium cattleianum Sabine

(Myrtaceae), followed by another invasive tree, Schinus terebinthifolius Raddi

(Anacardiaceae). Also common is a Polynesian introduction, Aleurites moluccana

(L.) Willd (Euphorbiaceae). While approximately 90% of the canopy is dominated by these introduced species, the remainder consists mainly of native trees (S.

Joe pers. obs.). Most frequently encountered are Acacia koa A. Gray (Fabaceae) and Metrosideros polymorpha Gaud. (Myrtaceae), followed by Pisonia brunoniana Endl. (Nyctaginaceae) and Nestegis sandwicensis (Gray) O.& I. Deg.

& L. Johnson (Oleaceae). Other, less common trees include Charpentiera obovata Gaud. (Amaranthaceae) and Pouteria sandwicensis (Gray) Baehni & O.

Deg. (Sapotaceae). Common understory plants include the invasive species

Clidemia hirta (L.) D. Don (Melastomataceae) and Rubus argutus Link

(Rosaceae), interspersed with native species of Hedyotis L. (Rubiaceae) and

Melicope (J.R. & G. Forst.) T.G. Hartley & B.C. Stone (Rutaceae). Natural populations of endangered natives occur throughout the area, including Cyanea superba (Cham.) A. Gray (Campanulaceae) and Schiedea obovata (Sherff)

(Caryophillaceae), which are augmented annually with greenhouse reared individuals.

41 With the assistance of five AGNRD staff, I conducted daytime searches for slugs

on three separate occasions in August 2003. Searches began at 1:00 PM and

ended at 2:00 PM. Through these surveys, I confirmed the presence of at least

four slug species at KMU (S. Joe pers. obs.), listed here from most to least

common: Deroceras Rafinesque sp. (Agriolimacidae), Limax maximus Linnaeus

(Limacidae), Meghimatium striatum van Hasselt (Philomycidae) and Limacus

flavus (Linnaeus) (Limacidae).

Study species

Three native (Cyanea superba, Schiedea obovata and Nestegis sandwicensis) and two alien (Clidemia hirta and Psidium cattleianum) species were chosen for the seedling growth and survival experiment. Cyanea superba is a palm-like tree reaching heights of 4-6 m when mature. Although two subspecies of C. superba are recognized, I do not distinguish them here because one of them C. superba regina (Wawra) Lammers has not been collected since 1932, and is likely extirpated (Wagner et al. 1999) and was not known from the Wai‘anae

Mountains. The remaining extant subspecies, C. superba superba, is referred to throughout this paper simply as C. superba. After its collection in 1870, there were no further documented sightings of C. superba until its rediscovery in the

Wai‘anae Mountains in 1971 (Wagner et al. 1999). Presently it is known from only one small population in the KMU, which, in 1998, numbered only 5 plants

(USFWS 1998c). Schiedea obovata, is a branching shrub growing to 3-10 dm height. Historically, S. obovata was found throughout the Wai’anae mountain

42 range scattered on ridges and slopes in diverse mesic forest, at elevations of

550-800 m (Wagner et al. 1999). Currently, the KMU is one of only three sites

were S. obovata can be found. As of 2000, the total number of plants was estimated to be about 30 (Arcand et al. 2002). Nestegis sandwicensis is an

endemic tree 8-25 m tall found in dry to mesic forest on all of the main Hawaiian

islands (except Ni‘ihau) (Wagner et al. 1999). It is locally common in the KMU

where it serves as the plant host to native snails in the genus Achatinella (Arcand

et al. 2002). Both C. hirta and P. cattleianum are highly invasive weeds native to

Central and South America. In Hawai‘i, they form dense, monotypic stands in

mesic to wet areas at 10-1500 m elevation. Clidemia hirta is a perennial shrub

0.5-3 m tall, while P. cattleianum is a short tree 2-6 m tall.

Seedling preparation

Schiedea obovata and C. superba were grown from seed by AGNRD staff at the

Hawai‘i Department of Land and Natural Resources (DLNR) Native Plant Nursery

adjacent to the KMU and located in the Wai‘anae Mountains at an elevation of

700 m. One month prior to planting, plants were moved on site but remained in

pots elevated on flats to prevent slug predation. At this time seedlings from C.

hirta, N. sandwicensis and P. cattleianum, all abundant on site, were dug up,

planted into pots and moved onto flats. Plants were selected by size, number of

leaves (2-4) and absence of any herbivore damage. Height and number of

seedlings by species is given in Table 3.1. In an effort to maximize survival

43 among native plant species, these were planted at larger sizes than either C. hirta or P. cattleianum.

Experimental design

Thirty 1 m2 plots were established along a contour close to the Kahanahāiki gulch bottom in February of 2004. Plots fell within an area roughly 0.6 ha, but random placement within this area was impossible due a steep slope and presence of dense stands of P. cattleianum. Instead, plots were arranged wherever trees did not interfere with placement, the slope was less than 35˚ and the nearest plot was at least 5 m away. Half of these plots (n = 15) were then randomly selected to receive physical and chemical barriers to slugs, while the remainder were exposed to natural levels of slug herbivory. In this paper, I refer to the former as the ‘slug-excluded’ treatment and the latter as the ‘slug-exposed’ treatment.

A copper mesh fence 15 cm high, buried to a depth of 5 cm and topped with a 5 cm strip of zinc tape enclosed all slug-excluded plots to prevent incursion of new slugs. For the slug-exposed treatment, a galvanized steel mesh fence of similar dimensions enclosed the plots, but 5 X 5 cm holes were cut into the bottom at 10 cm intervals to allow entry by slugs. Both the copper and galvanized steel hardware cloth were purchased from TWP Wire Mesh Inc. Berkeley, CA, had wire diameters of 0.71 mm and a mesh density of 3 squares per cm. Zinc tape was purchased from BAC Corrosion Control Ltd. (Telford, UK). While copper

44 barriers are known to repel slugs better than those constructed of other materials

(Hata et al. 1997), the effect is enhanced when copper is combined with zinc (S.

Joe unpub. data). In order to test the efficacy of this method I recorded the length of time it took for a slug to cross a line of tape 5 cm wide laid in a circle with the slug at its center. Circle boundaries were delineated using zinc and copper tape alone and in combination or masking tape (the control). Slugs were encouraged to escape the circle by shining a bright desk lamp on them. All slugs succeeded in escaping the control, the zinc and the copper circles within 5 minutes. In contrast, 80% of L. maximus (n = 30), 60% of L. flavus (n = 10) and 10% of M. striatum (n = 20) failed to cross a combination zinc/copper barrier within 5 minutes. Why some species were more sensitive than others is not known, but may be due to differences in mucosal conductivity.

A waterproof bait station containing the molluscicide Corry’s Slug and Snail

Death® (Corry & Co. Limited, North Bend, Washington, USA) was placed at the center of each slug-excluded plot to eliminate any slugs that managed to breach the barrier as well as any pre-existing, resident slugs. Bait was replenished every

20 days. Empty bait stations were placed at the centers of slug-exposed plots.

All plots were cleared of pre-existing vegetation and raked. Each plot was split in half; one half received transplanted seedlings (S. obovata (n = 3), N. sandwicensis (n = 3), C. superba (n = 5), C. hirta (n = 5) and P. cattleianum (n =

5) and the other was left fallow and natural germination from the seed bank was

45 monitored. In the half of the plots receiving seedlings, species were arranged randomly in 3 columns of 7 plants. Columns were spaced 8.3 cm apart and plants within each column were spaced 7.1 cm from one another. With assistance from AGNRD staff, all seedlings were planted on 23 February 2004 and monitored through 1 September 2004, for a total of 190 days. Whereas growth and survival of transplanted seedlings were recorded approximately every

10 days, natural regeneration was assessed only once at the end of the study.

This was because very few seedlings in total germinated in the cleared plots and, thus, were easily counted and identified. Germinants coming up among the transplanted seedlings were pulled to prevent crowding, however this was rarely necessary as very few were found. These were noted but not included in the counts of seedlings regenerating from the fallow half of the plot. Each plot received 1.9 L of water every 10 days to minimize transplant losses and to encourage germination from the seed bank in the fallow side of the plot.

The following measurements were recorded for transplanted seedlings.

1. Survival status. If dead, no further data were recorded.

2. Number of leaves.

3. Herbivore damage. This was the proportion of leaf removed. Damage

categories were: 0 or no damage, 1-25%, 26-50%, 51-75%, 76-95%.

4. Discoloration and senescence of the leaves caused by disease and/or

nutrient deficiency , were noted but not quantified.

46 5. Plant growth. An index of overall plant size, in dm3, was calculated based

on the formula for a cylinder (V = π r2 h ) where h is plant height and r is

length of the longest leaf.

Slug monitoring

Slug species composition and abundance were assessed using a total of 30

simple refugium traps consisting of 1 m2 pieces of unwaxed cardboard (Hawkins

et al.1998) placed on the ground within 1 m of each study plot. The traps

degraded quickly, so they were replaced once a month following a count of slugs

on top of or underneath the cardboard square. The use of traps to estimate

population densities can yield biased results. Compared to absolute sampling

methods, such as hand searching of soil and litter samples, large slug species

tend to be over-represented and small slug species (< 3.5 mm) underrepresented

(McCoy 1999). This is presumably because larger slugs can travel greater

distances and are therefore more likely to find the refugia. However, the traps

should provide fairly accurate information, within slug species, about relative

abundances across space and time. This relative abundance information was

used to assess uniformity of slug abundance among the plots and to provide

information on slug population fluctuation throughout the study period.

Statistical analysis

At the end of 190 days, seedling measurements were averaged for all individuals

of each species in each plot, because conspecific individuals in the same plot

47 should not be treated as being independent. If no representatives of a given species remained within the plot, it was omitted from the analysis. Thus, the maximum number of replicates for any species within each treatment was 15, but some were as low as 13 (when, for example, there were no survivors of a particular species in each of two plots). The plant size index and number of leaves, as well as amount of herbivory damage were averaged for all individuals of each species in each plot, whereas survival was calculated as the fraction of conspecifics within the same plot still alive after 190 days.

Two-way ANOVAs were used to assess the effects of plant species, the treatment (slug-exposed and slug-excluded) and their interaction on the responses listed above. The plant size index was log-transformed prior to analysis in order to correct for disparities in variance among groups. Herbivore damage was calculated as amount of leaf tissue removed (total from all leaves) divided by the number of leaves on the plant (both whole and grazed). Within each damage category only the high estimate (e.g. 25% in the 1-25% category) was used, thus, herbivory damage estimates were not conservative. The variance among groups was similar and residuals were normally distributed. All statistical analyses were performed with Minitab® Release 14 (Ryan et al. 2005).

48 RESULTS

Plant growth

Final plant size (as measured using a size index) and number of leaves per plant

varied among species (Figures 3.2 and 3.3; for volume, F4,133 = 34.06, P < 0.001,

for number of leaves, F4,133 = 11.0, P < 0.001). But for both responses, neither

slug treatment nor slug treatment by plant species interaction contributed

significantly to total variation (for log transformed plant volume, F1,133 = 0.03, P =

0.859 for treatment and F4,133 = 0.21, P = 0.930 for treatment x species; for number of leaves, F1,133 = 0.64, P = 0.426 for treatment and F4,133 = 0.83, P =

0.509 for treatment x species). Changes in plant volume and number of leaves

over time for each species and treatment are given in Appendix A and B

respectively.

Herbivory damage

Average damage per leaf at the end of the study period ranged from 10 to 30%

(Figure 3.4). There were no significant differences in final leaf damage between

the slug treatments (F1,133 = 0.36, P = 0.552), among plant species (F4,133 = 0.96,

P = 0.431), or as a result of a treatment by species interaction (F4,133 = 0.21, P =

0.933). Changes in leaf damage over time are given in Appendix C.

49 Seedling Survival

Within species effects

All plant species had increased mean survival in the slug-exclusion treatment, but the magnitude of this increase was only significant for two of the five species

(Figure 3.5). Results of the ANOVA showed a significant effect due to the slug treatment as well as a significant interaction between species and slug treatment

(Table 3.1), indicating that different plant species responded to slug herbivory to differing degrees. When exposed to slugs, both endangered natives (C. superba

and S. obovata) experienced significantly higher mortality (Tukey’s HSD, P =

0.0065 and P = 0.0002, respectively). The majority of plant deaths occurred

within the first two months after planting (Figure 3.6).

Comparisons across species

While differences between the two tree species, N. sandwicensis and P. cattleianum, were non-significant in the slug-excluded treatment (Tukey’s HSD, P

> 0 .05), in the slug-exposed treatment the native N. sandwicensis had significantly lower survival than the alien P. cattleianum (Tukey’s HSD, P =

0.0315). A similar trend was observed for S. obovata, which when exposed to

slugs had significantly higher mortality than C. hirta (Tukey’s HSD, P = 0.0488)

and P. cattleianum (Tukey’s HSD, P < 0.0001), but under slug-excluded

conditions did not. In the slug-exposed treatment, C. superba differed

significantly only from P. cattleianum (Tukey’s HSD, P < 0.0001), and, again, this

50 difference in survival was eliminated when slugs were excluded (Tukey’s HSD, P

= 0.9505).

Seed bank regeneration

Very few seedlings emerged naturally within the plots during the six month course of the study (Table 3.3). Clidemia hirta was most common, but it averaged less than two seedlings per plot (four per m2). Although more seedlings established in the slug-exposed plots than in the slug-excluded plots, this was not statistically significant (Mann-Whitney, P = 0.88). For all other species, no more than four seedlings total were found.

Slug monitoring

There was no significant difference in the number of slugs counted adjacent to slug-exposed vs. slug-excluded plots (paired-T test, t = -1.19, P = 0.255).

Changes in total slug numbers (Figure 3.7) approximately tracked monthly rainfall (National Weather Service Forecast Office 2004) with a correlation of r =

0.64, however this correlation was not statistically significant (P = 0.244).

DISCUSSION

In the Kahanahāiki Management Unit, slugs appear to be responsible for substantial seedling mortality of certain native plant species. Of three native species studied, two had significantly higher seedling mortality when exposed to slugs. Both of these species (C. superba and S. obovata) are critically

51 endangered, and the 49-53% decrease in mean seedling survival as a result of

slug predation is probably an important factor underlying their current status. In

comparison, the seedlings of both introduced species (C. hirta and P.

cattleianum), which are highly abundant at the study site, were not significantly impacted by slugs. The third native species tested, N. sandwicensis, also had

similar seedling survival in slug-excluded and slug-exposed plots. While this

species is not common at Kahanahāiki, adults make up a small proportion of the

canopy and subcanopy, and numerous seedlings were observed naturally

germinating underneath parent plants. The reasons that the latter three species

escape slug predation are unknown, but leaf toughness, which can influence slug

feeding (Dirzo 1980) may help protect both N. sandwicensis and P. cattleianum

while C. hirta grows rapidly enough to replace leaves lost to herbivory (S. Joe pers. obs.). The chemical characteristics of these species, in addition to other resistant species, should be investigated. Interestingly, none of the plant species exhibited sub-lethal signs of slug herbivory: there were no significant differences in plant size index, leaf number or herbivory damage scores among surviving plants in the two treatments. Within the first two months, seedlings in the slug- exposed treatment experienced high levels of mortality (Figure 3.6) suggesting the plants were small enough that a single feeding event could result in death.

Additionally, this may indicate that slugs feed on most or all of a seedling before moving on, killing the plant in a short amount of time.

52 Despite the small number of species tested, as well as their non-random

selection, my results suggest that slug herbivory may be skewing species

abundance in favor of non-native plants. In plots exposed to slug herbivory, the

rank order of mean seedling survival rates (with means in parentheses) was P.

cattleianum (90.7%), C. hirta (64.3%), N. sandwicensis (60.6%), C. superba

(37.3%) and S. obovata (35.6%). But when slugs were excluded, the rank order changed to P. cattleianum (92.0%), C. superba (80.3%), C. hirta (77.3%), S. obovata (70.0%) and N. sandwicensis (66.7%). Probably even more important than the rank order is the fact that all native species survival rates were high, and comparable to the introduced species. While other factors affect the rate at which surviving seedlings will compete and persist to maturity, these results show that slugs affect one important aspect of the process.

The low rate of natural seedling regeneration in both treatments points towards

additional factors impeding native seedling recruitment. These factors potentially

include reduced native seed rain (Moles and Drake 1999), low seed viability

(Baskin et al. 2004), lack of persistence in the seed bank (Drake 1998), the

destruction of seeds by predators (Garcia et al. 2005) and alteration of the soil

microclimate or chemical make-up (for example through allelopathy) (Macharia

and Peffley 1995). It is telling that the highly invasive C. hirta germinated the

most seedlings by far, and that its naturally germinated seedlings, like the

outplanted ones, were not impacted by slugs. A single individual can produce

over 500 fruits each season, each containing well over 100 seeds (Smith 1992);

53 this characteristic, combined with slug tolerance, helps explain its success at

Kahanahāiki and probably elsewhere.

Individual slug species appeared to respond to seasonal cues differently with M.

striatum appearing in May just as Deroceras sp. was disappearing. In contrast, populations of D. reticulatum in temperate areas are known to peak between May and June, after which they experience a decline until September (Hunter 1966), when a second generation emerges, reaching its peak around October.

Based on the fact that the two native species negatively impacted by slugs in this study are unrelated, it seems likely that other native species are also impacted by slugs to some degree. As with other types of impact, the effects of slugs are likely to be greatest on rare species. The implications are especially compelling for rare plant restoration: outplanted seedlings that are unprotected from slug predation may suffer significant mortality. Slugs now seem to occur in nearly all mesic to wet habitat types in Hawai’i. A lack of slug distributional and abundance data, however, makes it unclear whether slug densities at Kahanahāiki are typical or anomalous, and therefore how representative these results are for other natural areas. Additional studies should be conducted in other habitat types, and with different focal plant species, to better understand the impact of slugs in natural areas of Hawai‘i.

54 TABLES

Table 3.1. Seedling height (mm) by species on day 0 of the study. Seedling Species Mean SEM SD Minimum Maximum (Count) 90 Schiedea 43.1 2.28 21.73 10 100 obovata 150 Cyanea 28.22 1.34 16.38 5 100 superba 150 Clidemia 17.93 0.647 7.902 5 49 hirta 90 Nestegis 40.53 1.9 18.08 10 84 sandwicensis 150 Psidium 20.92 1.22 14.93 5 80 cattleianum

Table 3.2. Two-way ANOVA of seedling survival in slug-exposed and slug- excluded treatments. Source of Adj MS df F-ratio P variation Treatment 1.43734 1 24.05 < 0.000 Plant species 0.66459 4 11.12 < 0.000 Treatment x plant 0.24883 4 4.16 < 0.003 species Error 0.05975 140

Table 3.3. Number and identity of natural seedlings found in slug-exposed and slug-excluded plots over 190 days. Species Count (Slug- Count (Slug- exposed) excluded) Aleurites moluccana 2 0 Rubus rosifolius Sm. 4 0 R. rosifolius 0 3 Buddleia asiatica L. 0 1 Clidemia hirta 35 22 Pipturus albidus (Hook. & Arn.) 0 1 Gray

55

FIGURES

Figure 3.1. Location of Kahanahāiki Management Unit on the Island of O‘ahu. 56

Figure 3.2. Plant growth (using size index described in Materials and Methods) after 190 days in the slug-exposed vs. slug-excluded treatment. No significant difference between treatments was found. Bars are ± one SEM.

57

Figure 3.3. Change in the number of leaves per plant after 190 days in the slug-exposed vs. slug-excluded treatment. No significant difference between treatments was found. Bars are ± one SEM.

58

Figure 3.4. Average damage per leaf per plant in the slug-exposed vs. slug-excluded treatment grouped by plant species. No significant difference between treatments or species was found. Bars are ± one SEM.

59

Figure 3.5. Survival by species after 190 days in the slug-exposed vs. slug-excluded treatment. Significantly higher survival was observed for S. obovata and C. superba in the slug-excluded treatment (P < 0.05 *). Bars are ± one SEM.

60

Y axis = Survival (%)

X axis = Time elapsed from start of experiment (days)

Figure 3.6. Seedling survival over time in slug-exposed vs. slug-excluded plots by species. Survival between monitoring events is cumulative. Day 0 = 23 February 2004 and Day 190 = 1 September 2004. Bars are ± one SEM.

61

Figure 3.7. Counts of slugs frequenting cardboard traps April – August 2004 showing seasonal changes in the abundance of different species and monthly rainfall.

62

APPENDIX A. Cumulative change in plant size index over time by species and slug herbivory treatment. Bars are ± one SEM. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

63

Cumulative change in plant size index over time by species and slug herbivory treatment. Bars are ± one SEM. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

64

Cumulative change in plant size index over time by species and slug herbivory treatment. Bars are ± one SEM. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

65

Cumulative change in plant size index over time by species and slug herbivory treatment. Bars are ± one SEM. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

66

Cumulative change in plant size index over time by species and slug herbivory treatment. Bars are ± one SEM. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

67

APPENDIX B. Cumulative change in number of leaves per plant by species and slug herbivory treatment over time. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

68

Cumulative change in number of leaves per plant by species and slug herbivory treatment over time. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

69

Cumulative change in number of leaves per plant by species and slug herbivory treatment over time. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

70

Cumulative change in number of leaves per plant by species and slug herbivory treatment over time. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

71

Cumulative change in number of leaves per plant by species and slug herbivory treatment over time. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

72

APPENDIX C. Change in damage per leaf by species and slug herbivory treatment over time. Gains and losses in leaf surface area are cumulative. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

73

Change in damage per leaf by species and slug herbivory treatment over time. Gains and losses in leaf surface area are cumulative. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading. 74

Change in damage per leaf by species and slug herbivory treatment over time. Gains and losses in leaf surface area are cumulative. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

75

Change in damage per leaf by species and slug herbivory treatment over time. Gains and losses in leaf surface area are cumulative. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

76

Change in damage per leaf by species and slug herbivory treatment over time. Gains and losses in leaf surface area are cumulative. Day 0 = February 22, 2004 and Day 190 = September 1, 2004. Bars are ± one SEM. Slug Exposed vs. Slug Excluded treatments are illustrated with dark vs. light grey shading.

77 LITERATURE CITED

Albrectsen, B.R., H. Gardfjell, C.M. Orians, B. Murray and R.S. Fritz (2004) Slugs, willow seedlings and nutrient fertilization: intrinsic vigor inversely affects palatability. Oikos 105: 268-278.

Arcand, N., J. Beachy, M. Burt, V. Costello, S. Fyrberg, M. Keir, D. Palumbo, J. Rohrer, L. Salbosa, D. Souza and D. Toibero (2002) Natural Resources Status and Management, U.S. Army Garrison Hawai‘i, O‘ahu Training Areas, Annual Report. The Pacific Cooperative Park Studies Unit, Honolulu, HI, 329 pp.

Armstrong, S. (1995) Rare plants protect Cape’s water supplies. New Scientist 11: 8.

Arnold, G.W. (1980) The influence of odor and taste on the food preferences and food intake of sheep. Australian Journal of Agricultural Research 31: 657-666.

Barker, G.M. (1999) Naturalised terrestrial Stylommatophora (Mollusca: Gastropoda). Landcare Research, Hamilton, NZ, pp. 254.

Barker, G.M. (2001) Gastropods on land: phylogeny, diversity and adaptive morphology. pp. 1-46 in Barker, G.M. (ed.) The biology of terrestrial molluscs. CABI Publishing, New York, NY, 580 pp.

Baskin, J.M., B.H. Davis, C.C. Baskin, S.M. Gleason and S. Cordell (2004) Physical dormancy in seeds of Dodonaea viscosa (Sapindales, Sapindaceae) from Hawai‘i. Seed Science Research 14: 81-90.

Birkett, M.A., C.J. Dodds, I.F. Henderson, L.D. Leake, J.A. Pickett, M.J. Selby and P. Watson (2004) Antifeedant compounds from three species of Apiaceae against the field slug, Deroceras reticulatum (Müller). Journal of Chemical Ecology 30: 563-576.

Bowen, L. and D. Van Vuren (1997) Insular endemic plants lack defenses against herbivores. Conservation Biology 11: 1249-1254.

Breadmore, K.N. and W.D.J. Kirk (1998) Factors affecting floral herbivory in a limestone grassland. Acta Oecologica 19: 501-506.

Briner, T. and T. Frank (1998) The palatability of 78 wildflower strip plants to the slug Arion lusitanicus. Annals of Applied Biology 133: 123-133.

Brooks, A.S., M.J. Crook, A. Wilcox and R.T. Cook (2003) A laboratory evaluation of the palatability of legumes to the field slug, Deroceras reticulatum. Pest Management Science 59: 245-251.

78

Bruelheide, H. and S. Scheidel (1999) Slug herbivory as a limiting factor for the geographical range of Arnica Montana. Journal of Ecology 87: 839-848.

Bruegmann, M.M., V. Caraway and M. Maunder (2002) A safety net for Hawai‘i's rarest plants. Endangered Species Bulletin 27: 8-10.

Buckland, S.M. and J.P. Grime (2000) The effects of trophic structure and soil fertility on the assembly of plant communities: a microcosm experiment. Oikos 91: 336-352.

Cates, R.G. and G.H. Orians (1975) Successional status and the palatability of plants to generalized herbivores. Ecology 56: 410-418.

Christel, G., E. Johan and E. Ove (2002) Recruitment in Dentaria bulbifera: the roles of dispersal, habitat quality and mollusc herbivory. Journal of Vegetation Science 13: 719-724.

Coblentz, B.E. and D. Van Vuren (1987) Effects of feral goats (Capra hircus) on Aldabra Atoll. The Atoll Research Bulletin 306: 1-6.

Coomes, D.A. R.B. Allen, D.M. Forsyth and W.G. Lee (2003) Factors preventing the recovery of New Zealand forests following control of invasive deer. Conservation Biology 17: 450-459.

Coley, P.D., J.P. Bryant and F.S. Chapin III (1985) Resource availability and plant antiherbivore defense. Science 230: 895-899.

Cook, R.T., S.E.R. Bailey, C.R. McCrohan, B. Nash and R.M. Woodhouse (2000) The influence of nutritional status on the feeding behavior of the field slug, Deroceras reticulatum (Müller). Animal Behavior 59: 167-176.

Cottam, D.A. (1985) Frequency-dependent grazing by slugs and grasshoppers. Journal of Ecology 73: 925-933.

Cowie, R.H. (1995) Variation in species diversity and shell shape in Hawaiian land snails: In situ speciation and ecological relationships. Evolution 49: 1191- 1202.

Cowie, R.H. (1997) Catalog and bibliography of the nonindigenous nonmarine snails and slugs of the Hawaiian Islands. Bishop Museum Occasional Papers 50: 66 pp.

Cowie, R.H. (1998) New records of nonidigenous land snails and slugs in the Hawaiian Islands. Bishop Museum Occasional Papers 56: 60-61.

79

Cowie, R.H. (1999) New records of alien nonmarine mollusks in the Hawaiian Islands. Bishop Museum Occasional Papers 59: 48.

Cuddihy, L.W. and C.P. Stone (1990) Alteration of Native Hawaiian Vegetation Effects of Humans, Their Activities and Introductions. Distributed by the University of Hawai‘i Press, Honolulu, HI, 138 pp. del-Val, E. and M.J. Crawley (2004) Importance of tolerance to herbivory for plant survival in a British grassland. Journal of Vegetation Science 15: 357-364

Dirzo, R. (1980) Experimental studies on slug-plant interactions I. The acceptability of thirty plant species to the slug Agriolimax caruanae. Journal of Ecology 68: 981-998.

Drake, D.R. (1998) Relationships among the seed rain, seed bank and vegetation of a Hawaiian forest. Journal of Vegetation Science 9: 103-112.

Feeny (1992) Evolution of plant chemical defense against herbivores. pp. 1-35 in Rosenthal, G.A. and M.R. Berenbaum (eds.) Herbivores: their interactions with secondary plant metabolites, Vol. 2: ecological and evolutionary processes. Academic Press, New York, NY, 493 pp.

Fenner, M., M.E. Hanley and R. Lawrence (1999) Comparison of seedling and adult palatability in annual and perennial plants. Journal of Functional Ecology 13: 546-551.

Ferguson, S.H. (2004) Effects of poisoning nonindigenous slugs in a boreal forest. Canadian Journal of Forestry Research 34: 449-455.

Fritz, S., C.G. Hochwender, D.A. Lewkiewicz, S. Bothwell and C.M. Orians (2001) Seedling herbivory by slugs in a willow hybrid system: developmental changes in damage, chemical defense and plant performance. Oecologia 129: 87-97.

Gagné, W.C. (1983) New invertebrate host associates of the greensword (Argyroxiphium virescens). Proceedings of the Hawaiian Entomological Society 24:190.

Gagné, W.C. and C.C. Christenson (1985) Conservation status of native terrestrial invertebrates in Hawai’i. pp. 105-126 in: Stone, C.P. and M.K. Scott (eds.) Hawai‘i's terrestrial ecosystems: Preservation and management. Proceedings of a Symposium Held June 5-6, 1984 at Hawai‘i Volcanoes National Park. Distributed by University of Hawai‘i Press, Honolulu, HI, 616 pp.

80

Gagné, W.C. and L.W. Cuddihy (1999) Classification of Hawaiian plant communities. pp. 45-85 in Wagner, W.L., D.R. Herbst and S.H. Sohmer (Eds.) Manual of the flowering plants of Hawai‘i revised edition. Bishop Museum Press, Honolulu, HI, 2000 pp.

Garcia, D., J.R. Obeso and I. Martinez (2005) Rodent seed predation promotes differential recruitment among bird-dispersed trees in temperate secondary forests. Oecologia 144: 435-446.

Gebauer, J. (2002) Survival and food choice of the grey field slug (Deroceras reticulatum) on three different seed types under laboratory conditions. Journal of Pest Science 75: 1-5.

Gelperin, A. (1975) Rapid food-aversion learning by a . Science 189: 567-570.

Givnish, T.J., K.J. Sytsma, J.F. Smith and W.J. Hahn (1994) Thorn-like prickles and heterophylly in Cyanea: Adaptations to extinct avian browsers on Hawai‘i? Proceedings of the National Academy of Sciences 91: 2180-2814.

Glen, D.M., M.J. Wilson, P. Brain and G. Stroud (2000) Feeding activity and survival of slugs, Deroceras reticulatum, exposed to the rhabditid nematode, Phasmarhabditis hermaphrodita: a model of dose response. Biological Control 17: 73-81.

Gouyon, P.H., P. Fort and G. Caraux (1983) Selection of seedlings of Thymus vulgaris by grazing slugs. Journal of Ecology 71: 299-306.

Gurevitch, J. and D.K. Padilla (2004) Are invasive species a major cause of ? Trends in Ecology and Evolution 19: 470-474.

Hadfield, M.S. and B.S. Mountain (1980) A field study of a vanishing species, Achatinella mustelina (Gastropoda, Pulmonata), in the Wai‘anae Mountains of O‘ahu. Pacific Science 34: 345-358.

Hanley, M.E., M. Fenner, and P.J. Edwards (1995) The effect of seedling age on the likelihood of herbivory by the slug Deroceras reticulatum. Functional Ecology 9: 754-759.

Hanley, M.E., M. Fenner and P.J. Edwards (1996) The effect of mollusc grazing on seedling recruitment in artificially created grassland gaps. Oecologia 106: 240-246.

81 Hanley, M.E., M. Fenner and P.J. Edwards (2002) An experimental field study of the effects of mollusc grazing on seedling recruitment and survival in grassland. Journal of Ecology 83: 621-627.

Hata, T.Y., A.H. Hara and B.K.-S. Hu (1997) Molluscicides and mechanical barriers against slugs, plebeia Fischer and Veronicella cubensis (Pfeiffer) (Stylommatophora: Veronicellidae). Crop Protection 16: 501-506.

Hawkins, J.W., M.W. Lankester and R.R.A. Nelson (1998) Sampling terrestrial gastropods using cardboard sheets. Malacologia 39: 1-9.

Hitchmough, J.D. (2003) Effects of sward height, gap size and slug grazing on the emergence and establishment of Trollius europaeus (Globeflower). Restoration Ecology 11: 20-28.

Hulme, P.E. (1994) Seedling herbivory in grassland: relative impact of vertebrate and invertebrate herbivores. Journal of Ecology 82: 873-880.

Hunter, P.J. (1966) The distribution and abundance of slugs on an arable plot in Northumberland. Journal of Animal Ecology 35: 543-557.

Janzen, D.H. (1970) Herbivores and the number of tree species in tropical forests. The American Naturalist 104: 501.

Jordeans, K., J. Scheirs, R. Verhagen and T. Backeljau (2000) Is there a geographical pattern in the breeding system of a complex of hermaphroditic slugs (Mollusca: Gastropoda: Carinarion)? Heredity 85: 571-579.

Kaneshiro, K.Y. (1988) The uniqueness of Hawai‘i’s biota. pp. 7-10 in: Stone, C.P. and D.B. Stone (eds.) Conservation biology in Hawai‘i. University of Hawai‘i Press, Honolulu, HI, 280 pp.

Kelly, C.K. and M.E. Hanley (2005) Juvenile growth and palatability in co- occurring, congeneric British herbs. American Journal of Botany 92: 1586-1589.

Kotanen, P.M. (1995) Responses of vegetation to a changing regime of disturbance: effects of feral pigs in a California coastal prairie. Ecography 18: 190-197.

Lapçik, O., B. Klejdus, L. Kokoska, L. Ladislav, M. Davidova, K. Afandi, V. Kuban and R. Hampl (2005) Identification of isoflavones in Acca sellowiana and two Psidium species (Myrtaceae). Biochemical Systematics and Ecology 33: 983- 992.

82 Loope, L.L. (1992) An overview of problems with introduced plant species in national parks and biosphere reserves of the United States. pp. 3-28 in: Stone, C.P., C.W. Smith and J.T. Tunison (eds.) Alien plant invasions in native ecosystems of Hawai‘i: Management and research. University of Hawai‘i Press, Honolulu, HI, 904 pp.

Marquis, R.J. (1992) The selective impact of herbivores. pp. 301-325 in Fritz, R.S. and E.L. Simms (eds.) Plant resistance to herbivores and pathogens. University of Chicago Press, Chicago, IL, 600 pp.

McCoy, K.D. (1999) Sampling terrestrial gastropod communities: using estimates of species richness and diversity to compare two methods. Malacologia 41: 271- 281.

Moles, A.T. and D.R. Drake (1999) Potential contributions of the seed rain and seed bank to regeneration of native forest under plantation pine in New Zealand. New Zealand Journal of Botany 37: 83-93.

Mølgaard, P. (1986a) Food plant preferences by slugs and snails: a simple method to evaluate the relative palatability of food plants. Biochemical Systematics and Ecology 14: 113-121.

Mølgaard, P. (1986b) Population genetics and geographical distributin of caffiec acid esters in leaves of Plantago major in Denmark. Journal of Ecology 74: 1127- 1137.

Macharia, C. and E.B. Peffley (1995) Suppression of Amaranthus spinosus and Kochia scoparia: Evidence of competition or allelopathy in Allium fistulosum. Journal of Crop Protection 14: 155-158

Nystrand, O. and A. Granström (1997) Forest floor moisture control predator activity on juvenile seedlings of Pinus sylvestris. Canadian Journal of Forestry Research 27:1746-1752.

National Weather Service Forecast Office (1994) Hydronet system archived data for Oahu Wai‘anae gage HI-17. http://www.prh.noaa.gov/hnl/hydro/hydronet/hydronet-data.php

Peters, H.A., B. Baur, F. Bazzaz and C. Körner (2000) Consumption rates and food preferences of slugs in a calcereous grassland under current and future CO2 conditions. Oecologia 125: 72-81.

Pimentel, D., R. Zuniga and D. Morrison (2004) Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52: 273-288.

83

Pollard, J.P. (1992) The importance of deterrence: responses of grazing animals to plant variation. pp. 226-227 in Fritz, R.S. and E.L. Simms (eds.) Plant resistance to herbivores and pathogens. University of Chicago Press, Chicago, IL, 600 pp.

Raffaelli, D. and A. J. Mordue (1990) The relative importance of molluscs and insects as selective grazers of acyanogenic white clover (Trifolium repens). Journal of Molluscan Studies 56: 37-45.

Rai, J.P.N. and R.S. Tripathi (1985) Effect of herbivory by the slug, dussumieri, and certain insects on growth and competitive success of two sympatric annual weeds. Agriculture, Ecosystems and Environment 13: 125-137.

Rathke, B (1985) Slugs as generalist herbivores: Tests of three hypotheses on plant choice. Ecology 66: 828-836.

Rollo, C. D. and D. M. Shibata (1990) Resilience, robustness, and plasticity in a terrestrial slug with particular reference to food quality. Canadian Journal of Zoology 69: 978-987.

Rueda, A., R. Caballero, R. Kaminsky and K. L. Andrews (2002) Vaginulidae in Central America, with emphasis on the bean slug (Fischer). pp. 115-144 in Barker, G.M. (ed.) Molluscs as crop pests. CABI Publishing, New York, NY, 576 pp.

Runham, N.W. and P.J. Hunter (1970) Terrestrial slugs. Hutchinson & Co. Ltd., London, UK, 184 pp.

Ryan, B., B. Joiner and J. Cryer (2005) Minitab Handbook, Fifth Edition. Thomson Brooks/Cole, Belmont, CA, 505 pp.

Sahley, C., J.W. Rudy and A. Gelperin (1981) An analysis of associative learning in a terrestrial mollusc. Journal of Comparative Physiology 144: 1-8.

Sailer, D (2002) Slug barrier trail for Solanum sandwicense, Urera kaalae and Schiedea kaalae. The Nature Conservancy, Honolulu, HI, 10 pp.

Schreiner, I.H. (1997) Demography and recruitment of selected trees in the limestone forest of Guam in relation to introduced ungulates. Micronesica 30: 169-181.

Scherber, C., M.J. Crawley and S. Porembski (2003)The effects of herbivory and competition on the invasive alien plant Senecio inaequidens (Asteraceae). Diversity and Distributions 9: 415-426.

84

Scowcroft, P.G. and J.G. Giffin (1983) Feral herbivores suppress mamane and other browse species on Mauna Kea, Hawai’i. Journal of Range Management 36: 638-645.

Sessions, L. and D. Kelly (2002) Predator-mediated apparent competition between an introduced grass, Agrostis capillaris, and a native fern, Botrychium australe, in New Zealand. Oikos 96: 102-109.

Severns, R.M. (1981) Growth rate determinations of Achatinella lila, a Hawaiian USA tree snail. Nautilus 95: 140-144.

Siemens, D.H., S.H. Garner, T. Mitchell-Olds and R.M. Callaway (2002) Cost of defense in the context of plant competition: Brassica rapa may grow and defend. Ecology 83: 505-517.

Smith, C.W. (1992) Distribution, status, phenology, rate of spread and management of Clidemia in Hawai‘i. pp. 241-253, in: Stone, C.P., C.W. Smith and J.T. Tunison (eds.) Alien Plant Invasions in Native Ecosystems of Hawai‘i: Management and Research. Cooperative National Park Resources Studies Unit, University of Hawai‘i Press, Honolulu, HI, 887 pp.

Somchit, M.N., M.R. Sulaiman, Z. Ahmad, D.A. Israf and H. Hosni (2004) Non- Opioid anti-nociceptive effect of Psidium guajava leaves extract. Journal of Natural Remedies 4: 174-178.

South, A. (1992) Terrestrial Slugs, biology, ecology and control. Chapman & Hall, London, UK, 428 pp.

Stone, C.P. (1985) Alien animals in Hawai‘i’s native ecosystems: toward controlling the adverse effects of introduced vertebrates. pp. 251-297 in: Stone, C.P. and M.K. Scott (eds.) Hawai‘i's terrestrial ecosystems: Preservation and management. Proceedings of a Symposium Held June 5-6, 1984 at Hawai‘i Volcanoes National Park. University of Hawai‘i Press, Honolulu, HI, 616 pp.

Strauss, S.Y., J.A. Rudgers, J.A. Lau and R.E. Irwin (2002) Direct and ecological costs of resistance to herbivory. Trends in Ecology and Evolution 17: 278-285.

Swanholm, C.E., H. St. John and P.J. Scheuer (1959) A survey for alkaloids in Hawaiian plants. I. Pacific Science 13: 295-305.

Swanholm, C.E., H. St. John and P.J. Scheuer (1960) A survey for alkaloids in Hawaiian plants. II. Pacific Science 14: 68-74.

85 U.S. Fish and Wildlife Service (1994) Recovery plan for the Wahiawa plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 51 pp + 3 pp Appendix B.

U.S. Fish and Wildlife Service (1995a) Lāna‘i plant cluster recovery plan. U.S. Fish and Wildlife Service, Portland, OR, 138 pp.

U.S. Fish and Wildlife Service (1995b) Recovery plan for the Kaua‘i plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 270 pp.

U.S. Fish and Wildlife Service (1996) Recovery Plan for the Moloka‘i plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 143 pp.

U.S. Fish and Wildlife Service (1997) Recovery plan for the Maui plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 130 pp. + appendices.

U.S. Fish and Wildlife Service.(1998a) Kaua‘i II: addendum to the recovery plan for the Kauai plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 84+pp.

U.S. Fish and Wildlife Service (1998b) Moloka‘i II: addendum to the recovery plan for the Molokai plant cluster. U.S. Fish and Wildlife Service, Portland, OR, 52 pp.

U.S. Fish and Wildlife Service (1998c) Recovery plan for O‘ahu plants. U.S. Fish and Wildlife Service, Portland, OR, 207 pp., plus appendices.

U.S. Fish and Wildlife Service (1999) Endangered and threatened wildlife and plants; final endangered status for 10 plant taxa from Maui Nui, HA. Federal Register 64: 48307-48324.

U.S. Fish and Wildlife Service (2005) Endangered and threatened wildlife and plants; review of native species that are candidates or proposed for listing as endangered or threatened. Federal Register 70: 24880-24922.

Van Vuren, D. and B.E. Coblentz (1987) Some ecological effects of feral sheep on Santa Cruz Island. Biological Conservation 41: 253-268.

Vitousek, P.M., C.M. D’Antonio, L.L. Loope, M. Rejmanek and R. Westbrooks (1997) Introduced species: a significant component of human-caused global change. New Zealand Journal of Ecology 21: 1-16.

Wagner, W.L., D.R. Herbst and S.H. Sohmer (1999) Manual of the flowering plants of Hawai‘i revised edition. Bishop Museum Press, Honolulu, HI, 1918 pp.

86 Westerbergh, A. and A.B. Nyberg (1995) Selective grazing of hairless Silene dioica plants by land gastropods.Oikos 73: 289-298.

Whelan, R.J. (1982) Response of slugs to unacceptable food items. Journal of Applied Ecology 19: 79-87.

Wilby, A. and V.K. Brown (2001) Herbivory, litter and soil disturbance as determinants of vegetation dynamics during early old-field succession under set- aside. Oecologia 127: 259-265.

Wilson, M.J., L.A. Hughes, D. Jefferies and D.M. Glen (1999) Slugs (Deroceras reticulatum and Arion ater agg.) avoid soil treated with the rhabditid nematode Phasmarhabditis hermaphrodita. Biological Control 16: 170-176.

87