Leidyula floridana (Leidy) (Gastropoda: Veronicellidae), also known as Florida leatherleaf slug, was first described from southern Florida, but occurs widely in the Caribbean region, including Cuba, Haiti, Puerto Rico, Dominica, Bahamas, and Jamaica (Baker 1925; Maceira 2003; Rosenberg & Muratov 2006). Pilsbry (1948) concluded that its origin was Cuba, which is logical given its widespread distribution on that island and the tropical nature of the Veronicellidae. The range of L. floridana is expanding, and it is now found in northern Florida, Gulf Coast states (Louisiana and Texas), and northeastern Mexico (Hubricht 1985; Naranjo-Garcia et al. 2007).

Though the dietary habits of this slug are poorly known, L. floridana is polyphagous, feeding on plants of several families. Capinera and Guedes Rodriguez (2015) reported that this slug consumed measurable quantities of about 80% of the plant species provided, but clearly consumed more of some than others; only about 50% of the plants allowed significant slug growth to occur. Leidyula floridana is a large slug, often attaining > 10 g in weight and > 5 cm in length, so it is capable of a considerable amount of foliage consumption, perhaps 20 cm² per d (Capinera & Guedes Rodriguez 2015). It also is the most commonly observed slug in Florida, probably because it is so large. Nevertheless, slugs are predominantly nocturnal, so often are not observed. Slime trails found on sidewalks and vegetation are sometimes the only clues to indicate the cause of plant damage.

The technologies available for slug management have changed in recent yr. After many yr of depending nearly exclusively on highly toxic chemicals for protection of plants from terrestrial molluscs (usually metaldehyde products), less toxic materials, such as iron phosphate-, sodium ferric EDTA-, and sulfur-based baits are now available (Speiser & Kistler 2002; Hollingsworth & Armstrong 2003; Ciomperlik et al. 2013; Smith et al. 2013; Capinera 2018a, b). However, many gardeners prefer not to use pesticides of any type, so determination of host selection behavior could produce tangible benefits, perhaps allowing culture of bedding plants without concern about slugs. Here I report evaluation of the acceptance by L. floridana on the most commonly used bedding plants grown in Florida. The specific goals were to determine which plants were readily accepted, because use of these should be avoided, and which plants were not accepted, because use of these should be encouraged.

Materials and Methods

The slugs used in these studies came from a laboratory colony that has been maintained for about 5 yr. They were cultured in plastic boxes (TriState Plastics, Dixon, Kentucky, USA) measuring 28 × 18 × 10 cm (L, W, H) that contained about 5 cm of moist potting soil (Robin Hood garden soil, Hood Landscaping, Adel, Georgia, USA). Each box contained 20 to 30 slugs, and was maintained at 25.5 °C and 14:10 h (L:D) photoperiod. The boxes were not vented so humidity exceeded 90% RH. They were fed only romaine lettuce prior to evaluation.

Plant choice (‘common garden') tests were conducted in 60 × 60 × 60 cm cages with fine nylon mesh sides, containing about 10 cm of moist soil in the cage bottom. Each cage was planted with 7 different bedding plants, and each plant was photographed prior to exposure to slugs. Plants were selected to be about 15 cm high, and to provide approximately equivalent amounts of foliage. After planting and watering, 20 slugs weighing 4 to 6 g each were introduced to each cage and allowed to feed for 7 d. Environmental conditions were the same as described for slug culture except that the RH was unregulated, falling to about 70% during the d but > 90% at night. After the 7 d period, the bedding plants were compared to their pre-treatment photographs and visual estimates of the levels of defoliation were categorized as 1 = 0 to 19%, 2 = 20 to 39%, 3 = 40 to 59%, 4 = 60 to 79%, and 5 = 80 to 100%. The plants were procured locally at a Lowe's Home Center in Gainesville, Florida, USA, thus representing the plant condition that gardeners might encounter when planting new garden beds. The plants mostly consisted of flowering annuals, though some are perennials that often are grown as annuals under local weather conditions.

Two series of plant choice trials were conducted. Trials 1 to 5 each consisted of 1 marigold plant and 1 plant of 6 other species (n = 7 plants per cage; n = 31 species evaluated). Except for marigold, each plant species was tested only in 1 trial. Marigold is readily accepted as a food plant by molluscs (Raut & Ghose 1983; Dickens et al. 2017; Wilen & Flint 2018), and was included in each trial to gauge the overall hunger of the slugs, assuring that herbivory pressure was comparable across trials. Trials 6 and 7 were conducted in the same manner, but the plant species were selected from those that previously had been shown to be not readily accepted. One plant species in each of the latter 2 ‘unacceptable’ trials (sweet potato or purslane) was, in fact, selected because it displayed a modest level of acceptability. This was included for the same reason marigold was included in the first series, to gauge the willingness of the slugs to feed on acceptable food, though of course the expectations of ingestion were more limited.

Each trial (1–7) was conducted in 4 cages, but the trials were staggered over time so that each cage (replicate) contained plants that were of slightly different ages. The plant species and the distribution of plant choices are shown in Table 1. Plant damage ratings after 7 d were transformed to square root values and analyzed by 1-way analysis of variance (ANOVA) using GraphPad Prism (GraphPad Software, San Diego, California, USA). The mean values within each trial were compared using the Tukey-Kramer multiple comparison test.

The Lissachatina fulica (Bowditch) (Gastropoda: Achatinidae) growth study conducted by Dickens et al. (2017) overlapped, in part, with the choices displayed herein by L. floridana. The abilities of L. fulica and L. floridana to choose among some plants commonly grown in Florida was compared between these 2 studies using Spearman's correlation coefficient through application of GraphPad Prism.

Results

In the first series of evaluations, consisting of 5 trials to assess acceptance, L. floridana strongly discriminated among bedding plants (P < 0.001 in all cases), and most plants were readily categorized as acceptable or unacceptable (Table 1). The most unacceptable species (damage rating = 1.00) were: lantana (Lantana camara L.; Verbenaceae), tickseed (Coreopsis spp.; Asteraceae), torenia (Torenia fournieri Linden ex E. Fourn.; Linderiaceae), angelonia (Angelonia angustifolia Benth.; Plantaginaceae), and snapdragon (Antirrhinum majus L.; Plantaginaceae). Additional plant species that were not very acceptable (damage rating of between 1.00 and 1.50) were blue daze (Evolvulus globeratus Choisy; Convolvulaceae), penta (Pentas lanceolata [Forssk.] Deflers; Rubiaceae), dusty miller (Centaurea cineraria (L.) Jacq. ex Nym.; Asteraceae), viola (Viola spp.; Violaceae), celosia (Celosia argentea L.; Amaranthaceae), and geranium (Geranium spp.; Geraniaceae). Plant species that seem to be at considerable risk of damage (damage rating > 3) by L. floridana were: French marigold (Tagetes patula L.; Asteraceae), Madagascar periwinkle (Catharanthus roseus (L.) G. Don; Apocyanaceae), coleus (Plectranthus scutellarioides (L.) R. Br.; Laminaceae), petchoa (Petunia × Calibrachoa; Solanaceae), zinnia (Zinnia elegans Jacq.; Asteraceae), polka dot plant (Hypoestes phyllostachya Baker; Acanthaceae), chrysanthemum (Chrysanthemum morifolium Ramat; Solanaceae), petunia (Petunia spp.; Solanaceae), Stokes' aster (Stokesia laevis (Hill) Greene; Asteraceae), scarlet sage (Salvia splendens Sellow ex Nees; Lamiaceae), butter daisy (Melampodium paludosum Kunth; Asteraceae), and verbena (Verbena spp.; Geraniaceae). Examples of plant damage can be seen in Fig. 1, where after 2 d of feeding plant damage at the front of the image shows extensive French marigold damage (all leaves removed, though blossoms are present), whereas plant damage in the rear of the image shows lack of plant damage.

Table 1.

Damage ratings of bedding plants provided to Leidyula floridana for 7 d in multiple choice (‘common garden') tests. Ratings with a value of 1 experienced 0 to 19% consumption, whereas 5 experienced 80 to 100% leaf consumption (see Materials and Methods for details). Seven separate trials, each consisting of 7 plants, were conducted. Trials 1 to 5 were designed to identify preferred plants, whereas trials 6 and 7 were designed to assess nonpreferred plants. ANOVA statistics are found beneath each trial. Mean damage ratings followed by the same lower-case letter are not significantly different (P > 0.05) using the Tukey-Kramer multiple comparison test).

Not surprisingly, however, some plant species were somewhat intermediate in acceptance. Among those intermediate in acceptance were: impatiens (Impatiens hawkeri W. Bull; Balsamaniaceae), wax begonia (Begonia × Semperflorens × Cultorum; Begoniaceae), sweet potato vine (Ipomoea spp.; Convolvulaceae), firecracker flower (Crossandra infundibuliformis (L.) Nees; Acanthceae), sweet William (Dianthus barbatus L.; Caryophyllaceae), pansy (Viola × Wittrochinana; Violaceae), purslane (Portulaca oleraceae L.; Portulacaceae), and alyssum (Lobularia maritima (L.) Desv.; Brasscaeae).

Fig. 1.

Examples of plant damage by Leidyula floridana can be seen after 2 d of plant feeding. At the front of the image, extensive French marigold damage (all leaves removed, though blossoms are present) is evident, whereas plant damage at the rear of the image to torenia shows lack of plant damage (lack of damage to both leaves and blossoms).

In the second series, consisting of 2 trials using plant species that had been shown to be relatively unacceptable, there were few differences in acceptance; nearly all were refused. Only sweet potato vine was consumed significantly more than other plant species, and the level of consumption was low to intermediate. Overall, plant species that were relatively unacceptable in the first series maintained their unacceptable status in the second series, despite the absence of plants on which to feed.

Discussion

Plant acceptance displayed by the slugs was quite consistent, so even though the number of replicates (4) is quite minimal, significant differences in feeding were readily apparent. Furthermore, plant species that were found to be unacceptable in the first series of trials maintained their lack of acceptability even when slugs (in the second series of trials) had very limited access to acceptable food. Thus, plant species that were evaluated and received a damage rating of ‘1' in the series 1 trials should be considered most unacceptable to L. floridana. The most unacceptable species were: lantana, tickseed, torenia, angelonia, and snapdragon. Additional plant species that were not very acceptable (damage rating of 1.00 to 1.50) were blue daze, penta, dusty miller, pansy, viola, celosia, and geranium. Thus, these plant species provide a dozen choices available to gardeners who maybe concerned about slug damage and wish to minimize the risk of plant injury. Plant species that seem to be at considerable risk of damage (damage rating > 3) by L. floridana were: French marigold, Madagascar periwinkle (often sold as ‘Vinca'), coleus, petchoa, zinnia, polka dot plant, chrysanthemum, petunia, Stokes' aster, scarlet sage, butter daisy, and verbena. Unfortunately, these ‘at risk’ plants are among the most popular bedding plants currently used for gardens in Florida. French marigold often is considered to be resistant to insects and slugs, though it is susceptible to both forms of herbivores (Raut & Ghose 1983; Wilen & Flint 2018).

Dickens et al. (2017) conducted tests of ornamental plant suitability to the invasive giant African land snail, L. fulica, using plants recommended for the Miami region of Florida. They measured growth and survival of these snails fed 1 of 21 ornamental plants, and showed that plant species varied greatly in suitability and that, as would be expected of a polyphagous animal, there was not a strong affinity by the snails for a particular plant taxon. There was a statistically significant association of high growth rates and high levels of survival with annual plants, however, relative to perennial plants.

There also was overlap between the Dickens et al. (2017) growth study and the study reported herein, each consisting of 8 plant species. The species found in both studies were: French marigold, scarlet sage, purslane, zinnia, dusty miller, coleus, lantana, and snapdragon. If the final size of the L. fulica snails from the growth study is correlated with the acceptance evaluation from the L. floridana study, the relationship is positive and statistically significant (Spearman's r = 0.7765; P = 0.028). Thus, despite the use of different molluscs in the 2 studies, we might expect the slugs to thrive on the plants they most avidly accepted. This is the most common outcome when acceptance and suitability are assessed among polyphagous herbivores. Herbivores typically select not only what ‘tastes good' but actually what ‘is good’ for growth and reproduction; this is a fundamental aspect of plant-herbivore co-evolution. Importantly, it suggests that different polyphagous molluscs may respond to the same or similar stimuli, and that if additional assessments with other molluscs are performed, a pattern may emerge that will allow us to predict what plants might be more susceptible to damage from molluscs, many of which are newly invasive.