# Environmental Entomology

### Editor-in-Chief

E. Alan Cameron

1.315
1.597
36 out of 94

Published on behalf of

## Host Range and Reproductive Traits of Diabrotica speciosa (Germar) and Diabrotica viridula (F.) (Coleoptera: Chrysomelidae), Two Species of South American Pest Rootworms, with Notes on Other Species of Diabroticina

G. Cabrera Walsh
276-285 First published online: 1 April 2003

## Abstract

The reproductive biology, and larval and adult host range of Diabrotica speciosa (Germar), Diabrotica viridula (F.), Acalymma spp. (Coleoptera: Chrysomelidae: Galerucinae), and other Diabroticina are described. These Diabroticina are pests of several crops in South and Central America. The adult feeding hosts were compared, among species, and within species in different seasons. Laboratory oviposition and larval development tests on several hosts, provided the basis to construct a table of putative hosts, and general reproductive traits related to two species groups of Diabrotica (virgifera and fucata). Eggs of D. speciosa and D. viridula, were exposed to low temperatures to detect the ability to be dormant. Multivoltinism and lack of egg diapause was demonstrated for the three species, and field data suggest other South American species present the same traits. Diabrotica speciosa (fucata group) larvae developed well on maize (Zea mays L.), peanuts (Arachis hypogaea L.), and soybeans (Glycine max (L.) Merrill) roots, and not so well on pumpkin (Cucurbita maxima Duchesne and Cucurbita andreana Naudin), beans (Phaseolus spp.), and potato (Solanum tuberosum L.) roots. Oviposition preferences roughly paralleled larval suitability, but there was a clear preference for cucurbits as adult food, when available; pigweed (Amaranthus quitensis Kunth), sunflower (Helianthus annuus L.), and alfalfa (Medicago sativa L.) were in second place. Diabrotica viridula (virgifera group), preferred maize as adult and larval food, and for oviposition. Acalymma spp., were associated in every respect to cucurbits. Other species showed varying degrees of preference for oviposition and feeding, but in general, cucurbits were the preferred adult feeding hosts, followed by several wild plants, and maize the preferred oviposition host. Whereas cucurbits were consistently visited by the adults of every species, the virgifera group species oviposited and developed exclusively on Monocotyledonae. However, D. speciosa, as expected for a fucata group species, oviposited and developed on a wide range of hosts. This new knowledge on South American Diabroticina is discussed in the context of the current knowledge on North American Diabroticina. Differences and similarities are discussed in connection with their pestiferous status, and their potential for adaptation to new hosts.

• Diabrotica speciosa
• Diabrotica viridula
• Acalymma spp
• host range
• host shifts

## Introduction

THE SUBTRIBE DIABROTICINA (Coleoptera: Chrysomelidae) has many polyphagous and oligophagous species that feed on many food, ornamental, and wild plants. Of these, the genus Diabrotica Chevrolat includes the greatest number of pest species, including some of the most important row crop and vegetable pests of the Americas, be it the foliage, fruit- or flower-feeding adults, or the root-feeding larvae. In South America, the most common and problematic species is Di-abrotica speciosa (Germar). The adults cause important damage in maize (Zea mays L.), cucurbits, heavy defoliation in soybeans (Glycine max (L.) Merrill), and damage to the tender parts of almost every crop (Christensen 1943, Link and Costa 1978). They may also transmit bacterial wilt in cucurbits (unpublished data). Although the effect of the larvae on the crops they feed on has not been evaluated rigorously, there is sound evidence that they seriously damage maize, potatoes (Solanum tuberosum L.) and peanuts (Arachis hypogaea L.) (Sarasola et al. 1980, Gassen 1984). Diabrotica speciosa belongs to the fucata group of Diabrotica, to which the North American banded, spotted (or southern corn rootworm), and western spotted cucumber beetles belong (Diabrotica balteata LeConte, Diabrotica undecimpunctata howardi Barber, and Diabrotica u. undecimpunctata Mannerheim, respectively). Another group within the genus, the virgifera group, includes the northern (Diabrotica barberi Smith and Lawrence), western (Diabrotica virgifera virgifera LeConte), and Mexican (Diabrotica v. zeae Krysan and Smith) corn rootworms, the worst pests of maize in North America, as regards the annual losses in yields and control costs (Metcalf 1986). The sole South American species of this group considered a pest is Diabrotica viridula (F.), reported to cause significant damage on maize (Harries 1975, Olalquiaga 1980, Krysan 1986), and to transmit maize chlorotic mottle virus in Peru (Reyes and Castillo 1988).

Several species of the genus Acalymma Barber, akin to the North American striped cucumber beetle (Acalymma vittatum (F.)), are also very common in subtropical and tropical South America. Species such as Acalymma albidovittata (Baly), Acalymma bivittula (Kirsch), and Acalymma bruchii (Bowditch), often reach very high populations on cucurbit cultures, causing significant damage to flowers, foliage, and young fruit (unpublished data). The species studied overwinter as adults, and are almost exclusively associated to Cucurbitaceae (Bosq 1943, Christensen 1943, D’Araujo e Silva et al. 1968, Radin and Drummond 1994, Eben et al. 1997, Cabrera 1999, Eben 1999a).

Reproductive traits, natural larval hosts, adult feeding and oviposition hosts, and overwintering mechanisms of the different species of Diabroticites, present certain generalizations that can be extracted from former studies: the virgifera species larvae studied up to now show a strict association to both cultivated and wild Monocotyledonae, and although generalist feed ers as adults, they are still more abundant on cereals than cucurbits, forbs, or vegetables (Branson and Ortman 1967, 1970, 1971). The species most studied are generally associated to temperate climates with harsh winters, climates with seasonal droughts, or eleva tions, and present adaptations to unfavorable conditions, and/or ephemeral hosts through egg diapause mechanisms, and univoltinism (Krysan and Smith 1987). However, at least one species in the group, D. scutellata Jacoby, of tropical and subtropical distribution, is known to overwinter as an adult, and is most probably multivoltine (Eben 1999b).

Species of the fucata group, however, are generally multivoltine, polyphagous, and are not known to have species with diapausing eggs, but rather overwintering adults (Krysan and Smith 1987). They are more commonly associated with cucurbits, allegedly the ancestral host of the Diabroticina, and, at the same time, more generalists both during the adult and immature stages. They are distributed mostly in climates with more constant conditions (tropical and/or temperate, lacking temperature or drought extremes). D. speciosa is known to overwinter as an adult (Christensen 1943, Krysan and Smith 1987), and authors who have reared the species in the laboratory never came upon diapause symptoms (Pecchioni 1988, Silva-Werneck et al.1995, Avila et al. 2000, Cabrera Walsh 2001).

Another relatively unexplored area of the behavior of Diabroticites is the relationship between oviposition and feeding preferences. Although Diabroticites will oviposit on several moist substrates, such as sand, soil, cotton wool, and paper, the presence of fresh food, and especially the larval host, will greatly stimulate oviposition (Branson et al. 1975, unpublished data), suggesting Diabroticina females normally choose an adequate host for the larvae. Furthermore, the larvae seem to be unable to choose a suitable host other than by taste (Branson and Krysan 1981, Bernclau and Bjostad 1998). This unspecific behavior, would be compensated by the narrow oviposition host range observed so far in the species of the virgifera group (Branson and Krysan 1981), a trait of unquestionable adaptive utility in a natural environment, where adequate hosts may grow disseminated.

However, all these studies deal with North American Diabroticina, but their extension to South American species is yet to be confirmed. The objectives of this work are to report field observations and results of experiments performed at the ARS-USDA South American Biological Control Laboratory (SABCL), Hurlingham, Argentina. The host preferences and ecology of D. speciosa, D. viridula, and Acalymma spp. were studied in the laboratory. Relevant related data on other Diabroticina collected in the field, and reared in the laboratory for natural enemies and biology studies, are also presented. These include Paranapiacaba significata (Gahan), Diabrotica limitata (Sahlberg), Diabrotica tripunctata (F.), Diabrotica panchroma Bechyne´, and Diabrotica emorsitans Baly. Special emphasis was placed on the adult, larval, and oviposition host ranges, and egg and adult overwintering strategies.

## Materials and Methods

### Field Collections and Observations

The beetles used for observations, and to establish laboratory colonies were collected with sweep nets, funnels, and aspirators on at least 20 cultivated and wild host plants: mainly cucurbits, maize with fresh pollen and immature ears, and sunflower. Together with these methods, polyester cloths sprayed with root or berry extracts of the wild, perennial cucurbits, tayuya (Cayaponia spp.), were deployed at every stop as baits, and shaken into sweep nets (Cabrera Walsh 2001).

Also, data on adult feeding hosts, and host range shifts related to the season, were gathered through direct observation in the field. The favorite adult hosts of D. speciosa, D. viridula, D. tripunctata, D. limitata, D. panchroma, D. emorsitans (the last two species were the other South American virgifera group species studied herein), A. albidovittata, A. bivittula, and P. significata, were determined in terms of number of beetles of each species collected per site, on each available host per hour. The proportion of beetles of each species per host was calculated, and averaged for every site per season. The average proportion of beetles collected on each host per season was taken as a measure of host preference.

### Laboratory Rearing and Handling

Adults, both field and laboratory reared beetles, were kept in cages: up to 150 beetles in 1.5-liters cages, and up to 1,500 beetles in 19-liters cages. Diabrotica speciosa, D. viridula, D. limitata, and D. tripunctata adults accepted a diet based on Campbell and Jackson (1987), to which antimicrobials had to be added (Cabrera Walsh 2001). Young beetles were provided thin slices of raw squash as fresh food to increase survival. The other species collected were also offered a meridic diet, but they survived better on squash slices and seedlings (unpublished data). Water was provided from 45-ml plastic cups with cotton wicks through the lids.

### Oviposition Preference

Oviposition for all the species reared in the laboratory was obtained in 45-ml plastic cups with moist folded squares of black cotton/ polyester cloth, with seedlings rooted amid the folds. Tests on oviposition preference were performed using these cups with maize, cucurbits (butternut squash (Cucurbita pepo L.), and pumpkin (Cucurbita maxima Duchesne)), beans (PhaseolusvulgarisL., and Phaseolus coccineus L.), soybeans, peanuts, and pigweed (Amaranthus quitensis Kunth) seedlings. Potato plants were tested in 460-ml containers, to accommodate the germinated tubers. Maize was considered the control host for Diabrotica spp., and squash was considered the control host for Acalymma spp., because of previous experiments and publications indicating they were the favorite larval hosts for one and another genus respectively (Reed et al. 1984, Cabrera Walsh 2001). Moist cloth squares alone were also offered, and considered second controls to compare with the rejected, or nonpreferred, putative hosts. In preliminary multiple-choice experiments, where several putative hosts were offered simultaneously, oviposition was greatly reduced, and results were inconclusive (unpublished data). Consequently, the oviposition cups with different seedlings were offered in pairs (maize + another putative host for Diabrotica spp. and P. significata, and squash + another putative host for Acalymma spp.) in five cages with cohorts of newly emerged laboratory reared D. speciosa, A. bivittula, A. albidovittata, and D. viridula, and field D. emorsitans, D. panchroma, D. limitata, and P. significata.

The total number of eggs laid in each cup throughout the colonies’ life span, was calculated to an average daily number to determine an order of favorite oviposition hosts. When possible (i.e., regular oviposition was obtained), the average number of eggs on each host per harvest were compared with a two sample t-test. The eggs were harvested by rinsing them off the seedlings and cloths into a beaker, and counted under a dissection microscope when few eggs were harvested (≤300), or estimated according to volume in a graduated pipette with large collections (Cabrera Walsh 2001).

### Larval Host Range

Wild and cultured cucurbits, grain crops, pasture crops, and wild species that are common hosts of adult Diabroticites, were sampled from the year 1995 to the year 2000, to verify larval hosts in the wild. The plants were pulled or dug up, and the roots, and the soil surrounding the roots were shaken onto a white canvas, and visually inspected. When any chrysomelid larvae were found, they were placed in a maize seedling culture, and transported to the laboratory for identification.

### Egg Dormancy and Diapause Tests

The ability of D. speciosa and D. viridula eggs to survive cold through dormancy was tested for eggs laid by laboratory adults, and field adults collected in autumn and winter. For this, five 1-d old cohorts of around 1,000 D. speciosa and D. viridula eggs were held in large Petri dishes with moist blotting paper, refrigerated at 5 ± 1°C. Every 3–5 d, a batch of around 100 eggs was washed off the paper into a beaker, pipetted on to a new piece of blotting paper, and incubated in Petri dishes at 25 ± 1°C to count eclosion success.

The production of diapausing eggs was also tested for D. speciosa and D. viridula. Eggs were obtained from a minimum of five colonies, according to current availability, of ≥100 beetles each, of the following characteristics:

1. 20 colonies of D. speciosa, and 15 colonies of D. viridula, of laboratory adults reared throughout their life span in a walk-in rearing chamber, at 25 ± 2°C, 14:10 L:D photoperiod, 60% RH; to test for an obligated diapause as observed for the North American virgifera group Diabrotica (Howe and George 1966, Branson et al. 1975).

2. Six colonies of D. speciosa and five colonies of D. viridula, of laboratory adults reared since egg stage in garden conditions during fall, and transferred to the same walk-in rearing chamber described above, two wk after emergence; to test for diapause mechanisms that could be associated with exposure to fall/winter temperatures or photoperiods during development.

3. Ten colonies of D. speciosa and eight colonies of D. viridula, of laboratory adults reared in the rearing chamber, and transferred to garden conditions before mating; to test for diapause mechanisms elicited in adults by exposure to fall/winter conditions prior or after mating.

4. Five colonies of D. speciosa and five colonies of D. viridula field adults collected in autumn and winter, and incubated in the rearing chamber, to test for evidence of egg diapause in field specimens.

The eggs obtained from these colonies (≥500) were collected from the oviposition cups and incubated in Petri dishes on moist blotting paper at 25 ± 1°C to record percentage of viable eggs, and hatching time.

## Results

### Field Collections and Observations

Adult hosts of the most common southern South American Diabroticina are presented in Tables 1, 2, and 3. At least 116 species in 24 families were examined in southern South America, and found to host feeding adult beetles at least once (Table 1). The results in Table 2, indicate the favorite hosts of D. speciosa in terms of average proportion of beetles per host, in different seasons. However, because not all the hosts were present at each collection site, the average proportions do not necessarily sum 1. The best collections of D. speciosa during spring and summer, were on pumpkin flowers (C. maxima), pigweed, and sunflower, followed by alfalfa, maize silks, peppers (Capsicum spp.), peanuts and potatoes. This pattern held during fall, although alfalfa and soybeans were regularly found to host a large proportion of beetles as well. In winter, however, a drastic shift was observed, being pigweed, maize silks, leaf vegetables, especially spinach (SpinaciaoleraceaL.) and Chinese cabbage (Brassica chinensis L.), and flowers of the wild plants Datura arborea L., and sticky nightshade (Solanum sisymbri-ifolium Lamarck), some of the few plants found hosting any beetles.

View this table:
View this table:
View this table:

The results in Table 3 show the favorite hosts of the other common Diabroticina, in terms of average proportion of beetles per host, in different seasons. Again, however, not all the hosts were present at each collection site, so the average proportions do not necessarily sum one for all the hosts for each species. Diabrotica viridula, of tropical and subtropical distribution, was found more on maize (mainly silks, but also tassels) than cucurbits, as compared with D. speciosa. In winter it was only found on maize silks. The collection data of this species (Table 3), show a polyphagous species, but more closely associated to maize than D. speciosa in its feeding preferences. All the species of Acalymma sampled were found exclusively on cucurbits. Diabrotica limitata, was consistently found on pumpkin and maize silks, and only maize silks in winter. Diabrotica emorsitans was consistently found on pumpkin and tayuya, and in D. arborea blossoms in winter, but never on maize, even in locations where young maize with new silks and tassels were present, and frequented by other Diabroticina. P. significata’s most common hosts were pumpkin, sticky nightshade flowers, and pigweed foliage.

### Oviposition Host Range

Clear preferences were observed in the number and regularity of eggs laid on the different hosts. Diabrotica speciosa, D. viridula, D. emorsitans, and P. significata consistently laid more eggs on maize, whereas Acalymma spp., andD. limitata preferred pumpkin seedlings (Table 4). Because of the lack of sufficient field beetles, not every host chosen could be tested on all the species, only the ones reared at the laboratory. Diabrotica speciosa significantly preferred maize over C. maxima seedlings (t = 2.15; df =22; P <0.05), and bean seedlings (t =3.61; df =9; P < 0.003), but exhibited no preference for peanuts to maize (t =1.94; df =10; P > 0.08). Potatoes, and pigweed were not favored oviposition hosts, and were not significantly chosen over the moist cloth controls. These results reflect clear differences in feeding and oviposition host choices for some species: D. speciosa and P. significata selected maize and peanuts for laying eggs, although they are not among their favorite feeding hosts. At the same time, pigweed and cucurbits were heavily fed on, but not chosen for oviposition, or even rejected (<control) in the case of the pigweed (Table 4). An extreme case of different feeding to oviposition host was observed in D. emorsitans, that although never collected feeding on maize, only oviposited on maize. However, other species, as D. viridula, D. limitata, and A. bivittula, maintained a correspondence between their respective feeding and ovipositing hosts.

View this table:

### Larval Host Range

Results indicate thatD. speciosa is more polyphagous, both as an adult and larva, than either D. viridula or A. bivittula. Diabrotica speciosa has been reared, with greater or lesser success on the roots of maize, wheat, squash, potatoes, beans, soybeans, and peanuts (Table 5). The most suitable hosts, as regards survival from egg to adult, were maize, wheat, and peanuts (32, 18.3, and 25%, respectively), whereas potato and pumpkin were the least suitable of the successful hosts (4.9 and 6.3%, respectively) (Table 5). On cotton, C. bonariensis, pigweed, and sweet potatoes the first instars died after 2–5 d without feeding.

View this table:

Diabrotica viridula only developed in the laboratory on maize and wheat (mean ±SD survival from egg to adult: 25.0% ± 11.5; and 11% ± 7.4, respectively) (Cabrera Walsh 2001). No larval survival was observed on the Dicotyledonae tested. A. bivittula and A. albidovittata developed on pumpkin (C. maxima) (mean ± SD survival from egg to adult: 15.6% ± 8.3), as well as several cultivated and a wild cucurbit: Cucurbita andreana Naudin, melon (C. melo L.), Cucur-bitella asperata (Gillies) Walpers, and cucumber (C. sativus L.), but not on C. bonariensis, or any of the non cucurbitaceous plants tested.

No additional larval hosts to the ones cited in the literature or tried in the laboratory have been found in the field.

### Egg Diapause

The eggs of D. speciosa refrigerated at 5 ± 1°C were viable for a maximum of 60 d with a steady decline in viability from the tenth d, but high eclosion rates lasted around 20 d. Initial eclosion success at 27 ± 2°C averaged 92.6%, and took 8 ± 1 d. In the same laboratory conditions, D. viridula eggs hatched in 9 d ($$mathtex$${\rm{\bar x}}$$mathtex$$ = 91%), and chilled eggs were viable for a maximum of 40 d (Fig. 1). The eggs of either species that survived refrigeration hatched within the normal period at 25°C.

Fig. 1.

Survival of eggs of D. speciosa and D. viridula to chilling at 5 ± 1°C.

The three groups of adults tested for diapausing mechanisms (regular laboratory beetles, laboratory beetles incubated/reared in the open in winter, and winter collected field beetles) invariably produced eggs that hatched within the normal time when incubated at 25°C, indicating winter conditions did not elicit production of diapausing eggs. Also, 6–7 generations a year of D. viridula could be obtained in the laboratory at constant temperature, and up to three generations in garden conditions in the temperate Buenos Aires climate, indicating D. viridula is very likely a multivoltine species in its native subtropical environments.

The other virgifera group species studied, D. emorsitans and D. panchroma, have been collected in winter in Southern Brazil and northeastern Argentina (average temperature. 17.5°C and 18.4°C, average min temperature 13°C and 15.8°C, respectively), and in the area of the SABCL (average temperature 16.5°C, average min temperature 11.1°C) in April and May (late fall), indicating they can overwinter as adults. Although D. emorsitans could not be laboratory reared, eggs from winter specimens (n =27) hatched promptly, showing no signs of diapause.

## Discussion

The indirect density measure used to determine the favorite adult hosts of the different Diabroticites studied is far from precise, because different hosts allow different possibilities of finding and capturing the beetles (e.g., it is easier to aspirate beetles from the bell shaped blossoms of a pumpkin plant, than to examine a whole alfalfa plant in a row). Also, the mobility of the beetles makes counting very imprecise because, for instance, they can fly away or hide more easily in the foliage of a soybean row than in the large, relatively isolated sunflower head. Finally, the different collection method required (sweep nets in a row crop and wild plants, aspirators in vegetables and cucurbits, funnels in sunflower), can hardly be compared. However, drastic beetle density differences normally observed when two or more of the hosts were found in the same site compensate for these imprecisions, in that rarely were the densities observed close enough as to pose any doubts as to which plant species had more beetles feeding on it. Furthermore, the favorite hosts consistently had more beetles than the concurrent second choice hosts. So it gives a rough but representative estimation of the favorite, mediumly favored, and little visited hosts for each common Diabroticites species in different seasons.

Biological traits universal for the North American species of the virgifera group were not found in the species of that group which we studied, D. viridula, D. panchroma, and D. emorsitans. The North American species of this group are apparently without exception uni- or semivoltine, oligophagous, and possess drought or cold resistant diapausing eggs (Krysan 1982, Branson et al. 1982, Krysan 1986). However, except for a narrower host range, and a closer association to maize for ovipositing than the fucata group representatives described herein, D. viridula, D. emorsitans and D. panchroma did not show in the laboratory experiments, or field observations, any of the virgifera group traits described above. Their eggs did not diapause, nor did overwintering adults produce eggs in any way more resistant to cold. Unexpectedly, D. viridula proved if anything, less capable of adapting to winter conditions than D. speciosa, as far as egg resistance to low temperatures is concerned. This, however, would agree with the more northerly known distribution of the species. Although multivoltinism and lack of egg diapause was expected for D. speciosa, and Acalymma spp., this is the first demonstration of lack of egg diapause in a virgifera group species. Also, D. viridula, D. panchroma and D. emorsitans, together with D. scutelatta (Eben 1999b), are the only virgifera group species known thus far to be multivoltine, although Krysan (1986) had suggested this trait for D. viridula.

In their work on the northern hemisphere species of Diabrotica belonging to the virgifera group, Branson and Krysan (1981) suggested that univoltinism and diapausing eggs responded to their adaptation to a few annual grasses, taking specialization as the way to not have to deal with many different toxins, as the generalist fucata group larvae would. In this context, the South American virgifera group Diabrotica may have found a more benign climate, where suitable hosts were available most of the year, and overwintering as adults was possible, thus losing these traits, or never developing them.

In the subtropical environments of D. speciosa, D. viridula, and Acalymmaspp., the occurrence of adults in the field is merely a matter of benign weather, because suitable hosts are available all year round. In the southernmost homelands of D. speciosa (central Argentina, and southern Uruguay), where D. viridula and Acalymma spp. are not present, or rare, the insect is hard to find throughout the winter months, although it is always possible to collect a few on winter crops and wild plants. However, the sudden appearance of suitable crops in late winter/early spring, especially the first pumpkin crops, and pigweeds, is immediately heavily infested by both overwintering and, a few weeks later, newly emerged adults (Table 3). Radin and Drummond (1994), reported that no adjustment could be found to a degree-day model for the striped cucumber beetle in its temperate distribution. Only an association to the first warm days was observed (>12°C). They also reported mass attacks on squash seedlings early in the growing season. This is likely the situation in Argentina with D. speciosa, and other species: olfactory cues from the early crops congregate the extant overwintering adult population of the area as soon as the temperature enables foraging.

The apparent host shift observed in winter for D. speciosa, could respond to the fact that none of its favorite hosts were present in winter (noticeably cucurbits). Adult Diabroticites of most of the species studied are primarily pollen feeders, and pharmacophagous on bitter cucurbits (Nishida and Fukami 1989), and they seem to follow down a series of hosts as the favorite ones go missing in the field: starting with Cucurbita spp. blossoms, to pollen rich flowers, to tender foliage of the second choice hosts, ending in the low nourishment old growth of many plants (Table 1). However, actual feeding damage was not always observed on every plant where the overwintering beetles congregated. In this sense, cool weather may be regarded more as a preservation benefit until suitable hosts appear, than a hazard.

Feeding tests confirmed the field observations, as to the favorite adult feeding hosts, and in some cases, the marked difference between feeding and oviposition preferences (Tables 2–4). In an experiment where seedlings of the favorite hosts of D. speciosa were offered in cups to laboratory colonies, the favorite hosts in terms of foliar area consumed, were cucurbits, pigweed, beans and maize (unpublished data). Almost exactly the reverse of the oviposition preference. So the adult feeding hosts of D. speciosa, seem not to be in any way associated to the suitable oviposition and larval hosts. A similar situation was reported by Barbercheck et al. (1995) for D. u. howardi (fucata group) in the field, where the females were observed to leave peanut fields to feed, and return to them to oviposit. However, the Acalymma spp. and D. viridula showed a high similarity between their feeding and oviposition hosts.

There are official reports that D. speciosa is the main potato pest in southeastern Brazil, causing deep perforations in the tubers. According to our caged experiments, this species can develop on potato. However, first instars could not feed on the tubers, only on the roots, although mature larvae could. Development time is longer and more irregular, and percent emergence lower, showing a general impaired development on this crop. Furthermore, oviposition in cups with potato plants was very limited. Although these laboratory results may not reflect field conditions exactly, the reduced oviposition, prolonged larval stage, and greatly reduced larval survival (4.9% on potato, compared with 32% on maize, with an initial larval density kept well below carrying capacity in the containers) (Cabrera Walsh 2001), suggest it is not a fully adequate host.

Our current knowledge indicates we are dealing with a complex of vicariant species in South and North America, that share reproductive and feeding traits, and affect the same crops. However, the fact is that the South American pest Diabrotica pose different problems to the North American species. Both D. speciosa and D. viridula are present throughout the year, so crops are liable of being attacked at any moment. Also, the generalized benign climate within their distribution areas, allows for a wider range of crops and wild hosts to choose from throughout the year, restricting the potential for management alternatives. However, the lack of egg diapause, makes the population levels of South American pest Diabroticites of a given season wholly dependent on the presence of adult females, and the survival of the recent egg bank. In effect, the levels of both D. speciosa and D. viridula in crops has been observed to fluctuate drastically with local and temporary weather conditions (unpublished data). The diapausing eggs of the North American virgifera pest species, however, result in a virtually permanent egg bank, that makes the population levels of each season less susceptible to weather fluctuations.

A shift to soya as an ovipositing host has been observed in D. v. virgifera in the course of a few years (Levine and Oloumi-Sadeghi 1996, O’Neal et al. 1999, Spencer et al. 1999), enabling the egg bank to be renewed even in fields where maize is not present. This novelty, beyond the fact of neutralizing crop rotation as an effective control method, gives us an example of how unrelated oviposition and adult hosts, could operate adaptive advantages: a polyphagous parent relaxed its strict oviposition host behavior, and adapts to “waiting” for a suitable host. A similar mechanism may have enabled the adoption of potato as a larval host in the Brazilian populations of D. speciosa. This suggests oviposition host shifts may be a wide-spread adaptive mechanism among the Diabroticina, that could bring about the colonization of new hosts when the larvae were able to feed on a “mistaken” host.

The evidence of unspecific host selection by the larvae, sudden switch to new hosts, as D. virgifera on soybeans, and development of an economically significant pestiferous condition on a less than suitable host, as would be the case of D. speciosa on potato, is significant. It suggests that we face a taxon of insects whose adaptability to new and unfavorable conditions challenges our capacity to predict host crop and geographical range shifts, as well as our chances to provide new management techniques.

## Acknowledgments

I wish to acknowledge Mike Athanas, Jan Jackson, Ulrich Kuhlmann, Luiz Salles, Robert Schroder, and Stefan Toepfer for their support; Nora Cabrera for identifications and comments on the manuscript; and James Krysan for identifications and many painstaking reviews of the manuscript.