The interaction of a Botanophila fly species with an exotic Epichloë fungus in a cultivated grass: fungivore or mutualist?

Epichloë spp. (Ascomycetes: Clavicipitaceae) are endophytic fungii of Pooid grasses that cause choke disease, the suppression of seed production. They also host Botanophila spp. (Diptera: Anthomyiidae), the larvae of which feed on the fungus. Studies on Epichloë elymi on wild grasses indicate that the flies transfer spermatia between Epichloë mating types, thereby affecting cross‐fertilization, suggesting that the fungus–fly interaction reflects obligatory mutualism. Epichloë typhina, inadvertently introduced into western USA, was first detected in cultivated Dactylis glomerata L. fields in 1996. It spread rapidly, raising concerns about impacts on seed production. The present study was conducted to address questions pertaining to the occurrence and nature of the fungus–fly interaction in the new habitat of E. typhina. The first report of an endemic Botanophila species associated with E. typhina in Oregon is presented here. Surveys of D. glomerata fields indicated no correlation between fly abundance and fungal fertilization. In one field, no fly eggs or larvae were detected, but fertilized stromata were universally present. The fly was established in the remaining 12 fields surveyed, but while the number of stromata with fly larvae ranged from 6 to 98%, stromata development was uniform. Up to 10 larvae were present on a stroma, and these consumed >90% of the perithecia. Comparisons of pupal weights indicated that the fungal resource was not limiting, even at high larval densities. An exclusion study in a D. glomerata field also indicated that E. typhina fertilization occurred without the fly. In Oregon, the fly clearly benefits from the association with the fungus, but there is no evidence of benefit to the fungus. Thus if obligatory mutualism in the fungus–fly relationship described from the midwestern USA is the norm, our studies suggest a shift in the interaction to one of simple foraging on the fungus by fly larvae.     

The Botanophila–Epichloë association in cultivated Festuca in Oregon: evidence of simple fungivory

We investigated the Botanophila (Diptera: Anthomyiidae)–Epichloë (Ascomycetes: Clavicipitaceae) interaction in cultivated Festuca spp. (fine fescue) in Oregon in western USA. Epichloë spp. are endophytic fungi of grasses in the subfamily Pooideae. They develop a felt‐like stroma on the surface of grass culms and a dense mycelium within the culms that typically prevents seed head emergence. As a result, seed yields are suppressed, and hence the disease is known as choke. Studies of Epichloë spp. on wild grasses indicate that the fly–fungus interaction is an obligatory mutualism. During oviposition, Botanophila transfers Epichloë spermatia between stromata of opposite mating types, and the perithecia that develop after fungal fertilization serve as food for Botanophila larvae. In the current study, we surveyed 19 cultivated fields of Festuca spp. in Oregon, and observed choke caused by Epichloë festucae Leuchtmann, Schardl and Siegl in 10 of these. However, perithecia were observed in only four fields, and on only 1.0–2.6% of stromata. Perithecial development was also low, and rarely covered 50% of the stroma surface. Despite the absence or low frequency of fertilized stromata, Botanophila lobata Collin larvae were present in all choke‐infested fields. Infestation levels ranged from 2.5 to 70.7%, based on an examination of 175–200 stromata from each field. Only eight (= 2%) of the 450 stromata with B. lobata had perithecia, and the greater majority of B. lobata larvae completed development and exited from unfertilized stromata. This is the first report of the B. lobata–E. festucae association in the USA, and of B. lobata larvae developing successfully on unfertilized Epichloë stromata. The average pupal weight (0.0032 g) did not differ significantly from pupae (0.0030 g) originating from larvae that had developed on fertilized stromata of E. typhina on Dactylis glomerata in a neighboring field. This result indicates that in cultivated fine fescue fields in Oregon, B. lobata forages on E. festucae, but fly development is not dependent on the fertilized stromata of Epichloë.     

Daily activity rhythm and use of shelter in the slugs Deroceras reticulation and Arion distinctus under laboratory conditions

Daily activity of slugs, Deroceras reticulatum and Arion distinctus, was studied in the laboratory in relation to the presence of shelters so as to define optimal conditions for using traps to forecast crop damage. Under constant temperature and humidity, activity of slugs began with the lights off. Its arrest occurred after a period of activity with the lights on. Slugs rested most frequently under a shelter on bare earth and rarely used the same shelter for more than two consecutive days. When food was placed under the shelter, the number of slugs found under the shelter was more constant throughout the 24 h period, but lower during diurnal rest. The presence of 4% methiocarb pellets under the shelter led to an increased occupation by slugs, due to the poisoning of some of them. Acts of aggression were more numerous in A. distinctus than in D. reticulatum. In D. reticulatum the smallest slug was the least aggressive and suffered more attacks from the other slugs. In A. distinctus the numbers of attacks suffered or provoked varied little with the size of the slug, but the smallest slug showed less interactions with other slugs.     

Predation by the carabid beetle Harpalus rufipes on the pest slug Deroceras reticulatum in the laboratory

The Harpalini species Harpalus rufipes, as many other generalist carabids, consume a wide variety of prey and it is known to feed on pest slugs such as the grey field slug Deroceras reticulatum, but quantitative data about the predatory activity of H. rufipes on slugs are very scarce. In laboratory experiments, we assessed the capability of male H. rufipes to kill eggs and different‐sized slugs of the pest species D. reticulatum in either the absence or the presence of alternative live prey (dipteran larvae and aphids). We also investigated the preference of H. rufipes for eggs and hatchlings of D. reticulatum in a choice experiment. H. rufipes killed considerable amounts of eggs and small juveniles (≤5.0 mg) of D. reticulatum, both in no‐choice and in choice situations. Medium‐sized juvenile slugs (10–20 mg) were seldom killed only in no‐choice situations, and no large juveniles (50–60 mg) were killed. Dipteran larvae and aphids were killed also in no‐choice and in choice situations. The type of alternative prey presented with slug eggs affected the survival of the eggs to H. rufipes predation. The presence of dipteran larvae as alternative prey did not affect the survival of juvenile slugs. When eggs and small juvenile slugs were offered together, the survivals of both items were similar. The obtained results under laboratory conditions suggest that the generalist predator H. rufipes might realise an important contribution to the control of pest slugs.     

Predation by the Carabid Beetles Pterostichus melanarius and Poecilus cupreus on Slugs and Slug Eggs

Predation on slugs, Arion lusitanicus (Mabille) and Deroceras reticulatum (Müller), and their eggs by the carabid beetles Pterostichus melanarius (Illiger) and Poecilus cupreus (L.) was investigated in the laboratory. Slugs of different size and slug eggs were offered to the beetles in petri dishes with and without several alternative prey. Pterostichus melanarius destroyed eggs of D. reticulatum and A. lusitanicus, with a clear preference for D. reticulatum eggs. The availability of some alternative prey types adversely affected feeding on D. reticulatum eggs by P. cupreus and feeding on A. lusitanicus eggs by P. melanarius. Yet consumption of D. reticulatum eggs by P. melanarius was not significantly influenced by any alternative prey. Only P. melanarius killed small D. reticulatum in the presence of alternative prey. This suggests that P. melanarius and P. cupreus may have the potential to reduce slug populations in the field by destroying slug eggs and, in the case of P. melanarius, killing freshly hatched slugs.     

Size-specific susceptibility of the pest slugs Deroceras reticulatum and Arion lusitanicus to the nematode biocontrol agent Phasmarhabditis hermaphrodita

The nematode Phasmarhabditis hermaphrodita is a commercially available biocontrol agent against slugs. This product is especially interesting for use in organic farming, where products containing metaldehyde or carbamates cannot be used for controlling pest slugs. We investigated the potential of P. hermaphrodita for the control of the pest slugs Deroceras reticulatum and Arion lusitanicus. These two species are the most harmful slug pests in Switzerland. At different times of the year, we collected slug specimens of different weight and assessed their susceptibility to P. hermaphrodita in the laboratory. Batches of five slugs were subjected to five different doses of nematodes plus an untreated control and replicated three times. During six weeks, feeding and survival of the slugs were recorded. D. reticulatum was strongly affected by increasing nematode doses, irrespective of the slugs' body weight. In small specimens of A. lusitanicus, feeding and survival were strongly affected by the nematodes, while larger specimens remained almost unaffected. Because A. lusitanicus has an asynchronous development in Switzerland, it seems difficult to control the entire population with a single nematode application. To what extent nematodes will be used in practice for slug control depends on their effectivity against the pest slugs of major importance, on the longevity of the molluscicidal effect and on the price of nematodes.     

Effects of metaldehyde and methiocarb on feeding preferences and survival of a slug predator (Pterostichus melanarius (F.): Carabidae,Pterostichini)

The potential for reducing slug populations in crops through predation by generalist carabid beetles is well documented. However, few studies have considered interactions between biological and chemical control agents of slugs. Laboratory assays supported previous findings that the consumption of metaldehyde by slugs (Deroceras reticulatum) leads to increased duration of feeding bouts by carabid beetles (such as Pterostichini) on sub‐lethally affected individuals. However, a similar effect was not found for Pterostichus melanarius exposed to slugs fed on the other widely applied pelleted molluscicide formulation (methiocarb). Examination of beetle survival after consumption of slugs containing molluscicides demonstrated the strong biocidal properties of methiocarb, whereas metaldehyde consumption (ingested through slug predation) did not differ from control slugs killed by freezing. Beetle avoidance of slugs containing a more toxic molluscicide compound and the interaction between slug mucus production and beetle attack rates are discussed.     

Exotic Slugs Pose a Previously Unrecognized Threat to the Herbaceous Layer in a Midwestern Woodland

Developing effective restoration strategies requires first identifying the underlying factors limiting native plant recovery. The slug Deroceras reticulatum is an important herbivore in Europe, a global agricultural pest, and is introduced and abundant throughout eastern North America, but little information is available on the effect of this exotic herbivore on the forest herbaceous layer. Here, we test the palatability of 12 forest herbs to the introduced slug D. reticulatum and use field surveys to determine the degree to which slugs are damaging plants in the field. In laboratory feeding trials, slugs readily consumed most plants, but avoided the grass Elymus virginicus, the invasive forb Alliaria petiolata (garlic mustard), and thicker leaved plants. In the field, we documented significant slug damage, with close to 50% or more of plant leaves damaged by slugs on five of the six native species tested. Slug damage in the field was predicted by laboratory‐determined acceptability, but was significantly greater on short‐statured rosette species than on erect plants for a given acceptability value. Our results identify introduced slugs as an important, but overlooked obstacle to forest herb restoration and potential drivers of larger scale understory compositional change. The relaxed herbivore pressure on A. petiolata, relative to native competitors, suggests that invasive plant removal alone may not result in the recovery of native flora. Rather, restoration of unpalatable native species should accompany invasive plant control in slug invaded areas. Erect forbs, thick‐leaved plants, and graminoids should have the greatest success where introduced slugs are abundant.     

Notonomus gravis (Chaudoir) (Coleoptera: Carabidae) predation of Deroceras reticulatum Müller (Gastropoda: Agriolimacidae), an example of fortuitous biological control

The increasing use of pesticides in broad-acre cropping in South eastern Australia is suspected to have reduced native carabid beetle populations which fortuitously control potential pest populations. Slugs are increasingly becoming an establishment pest of canola, which is often attributed to stubble retention introduced to arable farming systems. Exclusion enclosures were employed to test the effect of the native carabid Notonomus gravis on the exotic pest slug Deroceras reticulatum. The native predatory species limited D. reticulatum populations and this was further supported by a negative field association between the predator and slug numbers. However, while N. gravis contributed to control of slug populations, enclosure experiments suggest that slug damage was not reduced below economic thresholds by this predator alone. Although N. gravis provides a “lying in wait” pest control option for slugs, multiple predators and environmental interactions need to be considered in developing robust integrated pest management guidelines.     

Impact of the field slug Deroceras reticulatum on establishment of ryegrass and white clover in mixed swards

The relationship between numbers of Deroceras reticulatum and the establishment of ryegrass cv. Melle and white clover cv. Milkanova was studied on a mini-plot experiment. Assessments 12 and 33 days after sowing showed that Deroceras reticulatum had a proportionally greater impact on survival and growth of clover than on ryegrass. Slugs had no apparent effect on numbers or growth of three species of broad-leaved weeds (Senecio vulgaris, Cardamine hirsuta and Epilobium sp.) which occurred in the field experiment. In pot experiments where seedlings of ryegrass together with one of three clover cultivars, with or without chickweed (Stellaria media), were exposed to Deroceras reticulatum, seedlings of white clover cv. Aran were less affected by slug grazing than cvs S184 or Milkanova. The presence of different clover cultivars did not significantly influence the impact of slugs on growth of ryegrass or chickweed, nor did the presence or absence of chickweed influence the effect of slugs on ryegrass and clover seedlings. Slugs had less impact on the growth of chickweed than on ryegrass or clover. Thus slug grazing would make clover less able to compete with chickweed as well as the other weeds found in the mini-plot experiment.     

Three ascomycetes on leaves of evergreen Ilex trees from Japan: Rhytisma ilicis-integrae sp. nov., R. ilicis-latifoliae, and R. ilicis-pedunculosae sp. nov.

Three species belonging to the genus Rhytisma causing tar spot were collected on leaves in evergreen trees of Ilex species from Japan. Rhytisma ilicis-latifoliae, the known species, is found on Ilex latifolia, and R. ilicis-integrae sp. nov. and R. ilicis-pedunculosae sp. nov. are found on I. integra and I. pedunculosa, respectively. Ascomata are formed on the abaxial part of the stromata in all the Rhytisma species studied, and spermogonia are formed on the amphigenous parts in R. ilicis-latifoliae and on the adaxial part in R. ilicis-integrae and R. ilicis-pedunculosae. Shape and size of asci, ascospores, and spermatia are distinctly different among the three species. The morphology of germination tubes from ascospores and appressoria is unique for each Rhytisma species. Yellowish spots arise on the newly developing leaves in mid-May, then abundant spermatia are produced in spermogonia in the three Rhytisma species. In the next year, ascospores are produced in ascomata from early April to late May in R. ilicis-integrae and from early April to early June in R. ilicis-latifoliae and R. ilicis-pedunculosae, and they are considered to be the inocula of disease infection.     

15N stable isotope labelling of slugs (Gastropoda: Pulmonata)

Stable isotope tracers are a promising tool for investigating the ecology of terrestrial slugs, including predator‐prey relationships, migration behaviour, nutrient turnover and dietary routing. The objective of the present feasibility study was to label two economically important slug groups, Deroceras reticulatum and keeled slugs (families Limacidae and Milacidae, respectively), with the stable isotope ¹⁵N under controlled laboratory conditions. Significant isotopic enrichment in slug tissue was detected after 4 days and persisted for at least 10 days after slugs had been fed on ¹⁵N enriched food for a period of 15 days. The time course of ¹⁵N uptake into slug tissues and its relation to food consumption were well described mathematically. Estimated mean ¹⁵N assimilation efficiencies from labelled maize mixed with unlabelled wheat bran were 30% and 38%, respectively, for the species groups studied. These findings suggest that slugs can be readily and efficiently labelled and that it is feasible to devise protocols for producing large numbers of isotopically labelled slugs for use in ecological studies. A simple method is described for the collection and analysis of cutaneous mucus from individual slugs which can be used to test uniformity of isotopic labelling.     

Biology of the Epichloë–Botanophila interaction: An intriguing association between fungi and insects

Epichloë fungi (Ascomycota: Clavicipitaceae) are endophytes of grasses that can produce epiphytic stromata on the culms of their hosts. The fungal stromata are visited by Botanophila flies for feeding and egg laying. We review research that has documented the heterothallic mating system of Epichloë, the mutualistic service of spermatization flies provide for the fungus, and host selection by flies. Flies display an active, stereotypical behavior immediately following oviposition by which spermatia are transferred endozoochorously to stromata. After eggs hatch larvae feed on developing perithecia. Several studies have focused on the cost (consumption of ascospores) to the fungus of engaging in the mutualism with its insect visitors. Generally, researchers have found benefits outweigh costs suggesting the mutualism is stable, however exceptions have been reported for Epichloë infecting some commercially cultivated grasses. Because Epichloë fungi are obligate outcrossers and flies are the major (perhaps only) vector of spermatia, the possibility exists that flies could promote reproductive isolation among Epichloë species through specific behavior. This idea has been tested using different approaches including observations of flies within screened cages containing Epichloë species and an analysis of fly gut content, which revealed no or very moderate selectivity by flies. Volatile fungal compounds are responsible for fly attraction and differ among Epichloë species. However, in a field bioassay using species-specific blends of the two predominant volatiles – a sesquiterpene alcohol (chokol K) and a methyl ester – flies showed no preference for specific blends. That is, flies do not appear able to distinguish between different fungal species based on their odor profiles. Thus, it appears that the flies' role in maintaining reproductive isolation among Epichloë species may be minor and that mechanisms of post-zygotic reproductive isolation are more important in keeping compatible species distinct. However, ethological mechanisms such as ‘stroma constancy’ favoring intraspecific mating may operate at a local scale and need to be investigated further.     

Pathogenicity of Moraxella osloensis, a Bacterium Associated with the Nematode Phasmarhabditis hermaphrodita, to the Slug Deroceras reticulatum

Moraxella osloensis, a gram-negative bacterium, is associated with Phasmarhabditis hermaphrodita, a nematode parasite of slugs. This bacterium-feeding nematode has potential for the biological control of slugs, especially the grey garden slug, Deroceras reticulatum. Infective juveniles of P. hermaphrodita invade the shell cavity of the slug, develop into self-fertilizing hermaphrodites, and produce progeny, resulting in host death. However, the role of the associated bacterium in the pathogenicity of the nematode to the slug is unknown. We discovered that M. osloensis alone is pathogenic to D. reticulatum after injection into the shell cavity or hemocoel of the slug. The bacteria from 60-h cultures were more pathogenic than the bacteria from 40-h cultures, as indicated by the higher and more rapid mortality of the slugs injected with the former. Coinjection of penicillin and streptomycin with the 60-h bacterial culture reduced its pathogenicity to the slug. Further work suggested that the reduction and loss of pathogenicity of the aged infective juveniles of P. hermaphrodita to D. reticulatum result from the loss of M. osloensis from the aged nematodes. Also, axenic J1/J2 nematodes were nonpathogenic after injection into the shell cavity. Therefore, we conclude that the bacterium is the sole killing agent of D. reticulatum in the nematode-bacterium complex and that P. hermaphrodita acts only as a vector to transport the bacterium into the shell cavity of the slug. The identification of the toxic metabolites produced by M. osloensis is being pursued.     

Tracking predator density dependence and subterranean predation by carabid larvae on slugs using monoclonal antibodies

1. Subterranean carabid larvae are more numerous than surface‐active adults, yet very little is known about their ecological significance, dietary preferences or ability to regulate populations of prey species, particularly pests. Part of the reason for this is that predator–prey interactions beneath the soil are almost impossible to observe. 2. Extensive field studies have shown that adult Pterostichus melanarius (Illiger) can affect the temporal and spatial dynamics of their slug prey. However, if larvae too are feeding on slugs, this could radically affect overall predator–prey dynamics. 3. We tested the hypotheses that P. melanarius larvae would kill and consume two slug species, Deroceras reticulatum Müller and Arion intermedius Normand, under laboratory and semi‐field conditions, and that there would be no significant difference in rates of predation on these slug species. 4. A new monoclonal antibody was developed that was capable of detecting the presence of slug proteins in the guts of P. melanarius larvae. 5. Pterostichus melanarius larvae killed both A. intermedius and D. reticulatum in the laboratory, feeding to a greater extent, and growing more rapidly, on the latter. The larvae were equally effective at reducing numbers of both slug species in a crop of wheat grown in semi‐field mini plots, but predation was affected by density‐dependent intra‐specific competition amongst the beetle larvae. 6. Future modelling of the dynamic interactions between carabids and slugs will need to take into account predation by larvae.     

Influences of preceding cover crops on slug damage and biological control using Phasmarhabditis hermaphrodita

In a replicated field experiment, ryegrass, vetch and red clover were grown or the soil was kept bare over a 2–month period in summer to compare the effects of these treatments on slug damage to the following crop (Chinese cabbage) and on the efficacy of nematodes (Phasmarhabditis hermaphrodita) applied as biological control agents to the soil at planting time to protect this crop. Slug damage was significantly (c. two times) greater after red clover or vetch than after ryegrass. Damage on plots without cover crop was intermediate and not significantly different from either extreme. Slug damage was reduced by about one‐third by the nematode treatment. The preceding cover crop did not influence nematode efficacy. Numbers of slugs on harvested plants (mainly Deroceras reticulatum and Deroceras panormitanum) were influenced by an interaction between cover crop and nematode treatment. On subplots without nematodes, more slugs were recorded with than without a preceding cover crop. No such differences were found on nematode‐treated subplots. Soil samples were collected at intervals from 0–99 days after nematode treatment to monitor nematode survival and infectivity in bioassays with D. reticulatum. No significant effects of cover crops were detected in bioassays. Moreover, there were no significant effects of nematodes on slug survival. Their effects on slug food consumption were mostly insignificant and any effects were transient and not consistent. However, significantly more slug cadavers contained nematodes when slugs were exposed to nematode‐treated soil. The implications of these results are discussed.     

SLUG DAMAGE AND NUMBERS OF SLUGS IN OILSEED RAPE BORDERING ON GRASS STRIPS

In 1995, slug damage and numbers of slugs were estimated intwo grass strips and adjacent rape fields. Investigations beganas soon as rape seedlings emerged and lasted for five weeks.Slug damage to rape plants 1 m from the grass strips was significantlyhigher than at greater distances from the strips. Derocerasreticulatum was the most abundant slug species recorded in bothgrass strips and adjacent rape fields. Arion lusitanicus andArion fasciatus were much less abundant than D. reticulatum.In one field, D. reticulatum declined steadily with increasingdistance from the grass strips and therefore appeared to havecaused the majority of severe damage to rape plants close tothe strips. This finding was surprising because until now severeslug damage in oilseed rape beside semi-natural habitats hasbeen observed only where A. lusitanicus was abundant. (Received 12 November 1997; accepted 26 January 1998)     

Selective grazing of acyanogenic white clover: Variation in behaviour among populations of the slug Deroceras reticulatum

Summary Collections of the slug Deroceras reticulatum were made from grassland sites containing contrasting frequencies of the cyanogenic morph of white clover, Trifolium repens. In choice chamber experiments, slugs obtained from sites with a low frequency of cyanogenic clover showed a significantly greater degree of selective eating of acyanogenic morphs than slugs taken from a site containing a high frequency of cyanogenic clover. Differences in selectivity between populations were caused both by differences in the rate of initiation of feeding on cyanogenic morphs, and by differences in the extent of damage once feeding had been initiated. The implications of these results for the cyanogenic polymorphism of T. repens are discussed.     

Predation on the slug Deroceras reticulatum by the carabid beetles Pterostichus madidus and Nebria brevicollis in the presence of alternative prey

1 Slugs are important pests in many agricultural crops and potential biological control agents are being studied as an alternative to molluscicides. Carabid beetles may be able to reduce slug populations, but their role as control agents may be influenced by the presence of alternative prey. 2 Attacks on the pest slug Deroceras reticulatum (Müller) by the carabid beetles Pterostichus madidus (Fabricius) and Nebria brevicollis (Fabricius) were investigated in the presence of alternative prey (earthworms and Calliphora fly larvae). Consumption of slug eggs and aphids was also investigated. 3 All five prey types were consumed to varying degrees during the experiments. Both beetle species showed a significant preference for Calliphora larvae over slugs. Pterostichus madidus showed a significant preference for earthworms over slugs. No preference was shown between earthworms or Calliphora larvae by P. madidus females or N. brevicollis. However, P. madidus males showed a significant preference for Calliphora larvae over worms. Pterostichus madidus showed no preference between slug eggs and aphids; N. brevicollis showed a significant preference for aphids over slug eggs. 4 The results from this study indicate that generalist beetles will often attack other prey in preferences to adult slugs. Slugs may not be preferred because of their mucus. Other prey items occur frequently in arable soils and generalist carabids may ignore slugs altogether and may only feed on them when slug density is high or other prey are unavailable.     

AUTOTOMY IN ARIONID AND LIMACID SLUGS

The occurrence of autotomy was investigated in five terrestrialslug species. Half of the Dcroceras reticulatum and 80 per centof the Limax tenellus slugs which were stimulated mechanicallyin the tail, autotomized the tail. Anon fasciatus, Anon subfuscusand Limax cinereoniger did not show any signs of being ableto autotomize. The autotomized tail of D. rcticulatum and L.tenellus was small compared to the total body weight and alsoto the weight of many predaceous beetles. D. reticulatum andL. tenellus escaped in a few seconds after the beginning ofmechanical stimulation; the reaction time was significantlyshorter than that used by species which did not autotomize theirtail. Comparison between A. fasciatus and D. reticulatum alsoshowed that D. reticulatum sought for refuges significantlymore effectively than A. fasciatus. The exposure of D. reticulatumto the beetle Ptero-stichus mger indicated that autotomy canstop a beetle long enough to let the slug escape. It seems thatautotomy has evolved more frequently in those species whichrely to a great extent on escaping and hiding when defendingthemselves, than in species whose physiology and/or environmentdo not enable them to hide quickly after an attack. (Received 18 February 1993; accepted 19 March 1993)