W. Eberhard, R. Achoy, M. C. Marin, and J. Ugalde.
Natural history and behavior of two species of Macrohaltica (Coleoptera: Chrysomelidae).
Psyche 100:93-120, 1993.

Full text (searchable PDF, 2924K)
Durable link: http://psyche.entclub.org/100/100-093.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

NATURAL HISTORY AND BEHAVIOR OF TWO SPECIES OF MACROHALTICA (COLEOPTERA: CHRYSOMELIDAE) BY W. EBERHARD"~, R. ACHOY~, M. C. MAR IN^, AND J. UGALDE~'~ The genus Macrohaltica and the allied genus Altica have approximately 30 described species in Central America and per- haps 200 or more worldwide (R. White pers. comm.). Larvae and adults often feed on onagraceous and ericaceous plants (Scherer 1969, Phillips 1977a), though other plants are also utilized (Woods 1918, Balsbaugh and Hays 1972, Phillips 1977a, Barstow and Git- tins 1973, LeSage 1990). Some species of Macrohaltica are con- sidered secondary pests of crops in Costa Rica (King and Saunders 1984). Some species are difficult to distinguish morphologically (Phillips 1977b), and both parthenogenesis and interspecific hybridization are known (Phillips 1977a, 1979). This paper describes aspects of the natural history and behavior of a pair of species in the central highlands of Costa Rica, a metal- lic blue species M. jamaicensis (hereafter M. j.), and an apparently undescribed metallic purple species (hereafter M. sp.). The beetles frequently attract attention because they aggregate, sometimes in masses of up to several thousand individuals. A second striking trait is that some aggregations include individuals of both species, and cross-specific male-female pairs are common in such aggrega- tions.
MATERIALS AND METHODS
Observations were made over a span of nine years, during which time we saw many tens of thousands of beetles in the field, and raised thousands in captivity. The behavior of beetles in four aggregations was observed in detail:
Aggregation I - both species on Ludwigia octovalvis (Ona- graceae) plants near San Antonio de Escazu, el. 1300 m, 27 ' Smithsonian Tropical Research Institute Escuela de Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica Institute Nacional de Biodiversidad, Santo Domingo de Heredia, Costa Rica Manuscript received 11 January 1993.




================================================================================

94 Psyche [vo~. 100
June-13 July 1983). In this and other field observations, adults of M. sp. were identified by their purple color; since Fl hybrids between the species are also purple (R. Achoy unpub. thesis), these identifications are somewhat uncertain. Nevertheless hybrid indi- viduals (distinguished by genitalic characters) are uncommon in the field (Eberhard and Ugalde in prep.); Aggregation I1 - M. j. on Cuphea sp. (Litraceae) plants near the edge of a lawn near San Vito de Java, el. 1000 m, 23-25 Jan. 1987; Aggregation I11 - M. sp. on low grass in a pasture near Mon- teverde, el. 1300 m, 14-15 Feb. 1985; and Aggregation IV - M. sp. on various plants along approximately 50 m of a roadside above San Antonio de Escazu, el. 1400 m, Feb-June, 1992. These and other sites mentioned in the text are shown in Fig. 1.
All beetles in Aggregation I were counted on 12 nights. Each night all unmarked pairs of males and females were marked on the dorsal surface with airplane enamel (two spots for males, one for females). Unmarked solitary beetles were not marked. Beetles were raised in captivity in San Jose with ambient tem- peratures and light cycles. Larvae were fed freshly cut leaves in petri dishes, and adults were fed either the same or kept in plastic bags on branches of potted food plants. When larvae ceased feed- ing and began wandering, they were placed in containers with soft soil where they could pupate.
Experiments on preferences for pupation substrates were per- formed as follows. An approximately cylindrical plastic cup (8.8 cm in diameter at the mouth) was placed in the center of an approximately cylindrical, larger plastic container (16.0 cm in diameter at the mouth). The cup was filled to the brim with one substrate and set in the center of the larger container, which was then filled to the brim of the cup with the other substrate. Larvae placed on the surface thus had access to both types of substrate (60.8 cm2 of the substrate in the cup, 140.3 cm2 of that in the larger container). Neither type of soil was compacted. After the larvae (30-40 per repetition) had disappeared underground, the smaller container was removed from the larger, and the numbers of adults reared from each were recorded. In replica experiments sub- strates in the larger and smaller containers were reversed. Voucher specimens of adult beetles are deposited in the U. S. National Museum, the British Museum (Natural History), and the



================================================================================

19931
Eberhurd, Achoy, Murin & Ugalde
Pic0 Blanco
Fig. 1. Map of study area, showing sites mentioned in the text. Museo Nacional de Costa Rica housed at the Institute Nacional de Biodiversidad. Vouchers of the plants are in collections of the Mis- souri Botanical Garden, St. Louis, MO and the Escuela de Biologia of the Universidad de Costa Rica.
Unless noted otherwise, all statistical tests were two-tailed Chi squared tests. Averages are followed by one standard deviation. Altitudinal Range
The altitudinal ranges of the two species differed. The highest and lowest sites at which we found M. sp. were approximately 2300 m on Cerro de la Muerte, and about 1000 m near San Rafael de Escazu; M. j. was absent at higher elevations, occurring from



================================================================================

96 Psyche [vo~. 100
San Josecito de San Isidro de Heredia and above San Antonio de Escazu (both about 1400 m) to San Rafael de Escazu (1000 m). Life Cycle
Both species were raised from egg to adult in captivity. Approx- imate durations of immature life stages were 5 days for the egg, 21 for the larva, and 14 for the pupa, but durations of the larval stage varied somewhat for both species on different host plants (Table 1). The longest adult life in captivity was 94 days for a female M. j. (a Fl hybrid female lived 99 days). Eggs were about 1.2 mm long, oval in shape, and yellowish in color. They did not have the fecal streak found on the eggs of some Altica spp. (Woods 1917, Phillips 1977a). Each egg was covered with a highly reticulate mesh of a white, somewhat extensible, adhesive material. When the egg was dry this covering gave it a white color. When a group of 50 newly- laid eggs (<24 hr old) was tasted, they produced a burning sensa- tion similar to that of the larvae (below) but weaker. Oviposition behavior of M. j. was observed in captivity. The female made repeated brief, apparently exploratory extensions of her genitalia against the substrate before ovipositing. The pair of coxites, which bear setae which are probably chemosensory and/or mechanosensory in function (Phillips 1978), projected rearward beyond the tip of the ovipositor with each extension. Table 1. Larval weight at maturity, and average development time for larvae raised on two different food plants. (/. = M. jamaicensis; sp. = M. sp.). Numbers in parentheses are total number of larvae weighed, and number of groups of larvae weighed. Larvae from all crosses were significantly lighter when raised on G. insig- nis (p<0.01 with Kruskal-Wallis using groups of larvae). Larvae of j. X j. cross raised on G. insignis were significantly lighter (Pc0.005 with Kruskal Wallis) than those of all other species.
Ludwigia octovalvis Gunnera insignis
- - -- - -
larval wt. development larval wt. development Cross at maturity time at maturity time
female male (mg) (days) (mg) (days)
j. X
j. 45.3å±2.2(533,8
14.3 10.610.2(3,2) 40.3
sp. X
sp. 42.3å±3.8(326,13 21.6
30.2(144,7) 34.7
j. X sp. 42.9å±2.2(131,5
17.5
36.651.9(111,5) 29.3
sp. X
j. 45.9å±2.0(64,4 16.4 34.7å±0.9(78,4
29.4




================================================================================

19931 Eberhard, Achoy, Marin & Ugalde 97 Eggs were found on the leaves of all three genera of host plants. Those of M. sp. on the large G. insignis leaves, which were up to more than 1 m in diameter, were usually solitary, each alongside a vein of the leaf; in a sample of 728 eggs, there was one group of 9, 3 of 3, 20 of 2, while all the other 670 eggs were not within .5 mm of another egg (average group size = 1.05k0.37). In contrast, the eggs of M. j. were much more often grouped (av. group size = 10.9k10.9, range 1-59, n=1341) on the thinner and much smaller leaves of Ludwigia octovalvis (about 1 cm wide). Eggs of M. j. on L. octovalvis were nearly always on the underside of the leaf (70 of 74 groups). In captivity M. j. also laid eggs in tight spaces such as the crack between the top and bottom of a petri dish. Eggs of M. j. generally failed to hatch unless they were quite humid. Thus eggs laid on leaves of potted L. octovalvis plants in the lab (where neither rain nor dew wet the eggs) late in the wet season (Oct. 1991) often failed to produce larvae (only 21.9% of 2061 eggs in 54 groups hatched); at least one egg hatched in only 16 of 54 groups laid by 5 females (branches with eggs were enclosed in small plastic sandwich bags with many small holes). In the early dry season (Nov. 1991) hatching success was even lower (16 of 128 groups laid by 12 females, p<0.01). In contrast, when leaves on which eggs had recently been laid were placed on filter paper or paper napkins saturated with water in small, culture dishes, hatching success was much higher (79% of 506 eggs; 39 of 39 groups had at least one egg hatch) (p<0.001 compared with groups of eggs on leaves).
Survivorship of eggs and larvae on plants where adults were aggregated was probably reduced. Aggregations of adults some- times completely stripped G. insignis leaves and entire Cuphea and Ludwigia plants (Fig. 2), so eggs could be consumed along with the plant. We observed captive adults eating eggs laid on the walls of petri dishes. Larvae which eclosed on plants with aggregations of adults also ran the risk of not having adequate food. Females of at least M. sp. tended to oviposit on leaves where adults were not aggregated. Inspection of 16 pairs of G. insignis leaves, one with an aggregation of adults and the other the nearest relatively uneaten leaf of the same plant, showed that even though M. sp. beetles were seldom seen away from aggregations, the densities of eggs (eggs/cm2) were lower on the occupied leaves: average densi- ties were 4.28å±7.01/cm (range 0-26.4) vs. 6.71k6.96/cm2 (range



================================================================================

98 Psyche [VOI. 100
Fig. 2. Adults of M. sp. on small (above) and large (below) !eaves of Gunnera imgnis.




================================================================================

19931 Eberhard, Achoy, Marin & Ugalde 99 0-28.2) (in 13 of 16 pairs the density on the unoccupied leaf was higher, p<0.01).
Larvae remained on the host plant day and night. Although lar- val coloration was not aposematic (pale brownish yellow or orange with black head capsule, legs, and bristles), some larvae of both species had a strongly disagreeable taste. Recently moulted third instar larvae (mostly black in color) had little or no taste, but fully grown larvae (mostly orange in color) of M. sp. on G. insignis plants produced a delayed but powerful burning sensation. When one entire larva was tasted, the burning began about 30 sec later, and lasted for more than 15 minutes. A larva of M. j. on a L. octo- valvis plant and an adult M. sp. on a G. insignis plant had a similar, though less intense taste (larvae were identified by association with nearby monospecific aggregations of adults). First instar larvae generally began feeding in a group near eggs from which they hatched, but older larvae did not aggregate as they fed. The average number of M. j. larvae per occupied leaf on two L. octovalvis plants was 1.8 å 1.8 (N = 57 leaves with larvae). Early instar larvae of both species usually fed on only the lower layers of the leaf, leaving the upper side intact. Later instars ate entire leaves of Ludwigia spp., but usually left one (usually the upper) side of the much thicker G. insignis leaves intact. Larval damage to G. insignis leaves was thus easily distinguished from that produced by adults, as in some Altica (Woods 1917, 191 8). Larvae on Ludwigia plants which had been stripped of leaves also fed on the cuticle of stems and fruits. Although there was substan- tial variation, it appeared that M. j. larvae tended to feed on leaves lower down on L. octovalvis plants.
Larvae left host plants to search for appropriate pupation sites. In nature mature larvae of M. sp. (identified as such by heavily damaged Gunnera plants and the absence of Ludwigia or Cuphea plants nearby) were seen crawling on the ground 5 m from the nearest plant with larval feeding damage; mature M. j. larvae (identified by raising them to adults) were found to up to >1 m from the base of the nearest L. octovalvis. Active selection of pupation sites was demonstrated experimen- tally in M. j. When offered either rich humus or sandy soil, larvae tended to pupate in humus: 74 vs. 1 adults were raised in two repli- cas with sandy soil in the central portion, 57 vs. 13 in two replicas with sandy soil on the periphery (ratio of area in center: periphery



================================================================================

100 Psyche [vo~. 100
= 1:2.31, p<0.001). Although their food plants generally grow in sites with waterlogged soil, larvae preferred to pupate in drier soil. When offered either humus soil with water barely standing (approximately 43% water content by weight) or drier humus soil (approximately 32% water content by weight), the larvae tended to pupate in the drier soil (78 vs. 0 adults in two replicas with the drier soil on the periphery (pc0.001).
Underground, larvae constructed small, ovoid pupation cham- bers. The depth of 62 chambers in captivity averaged 1.7k1.2 cm. The long axis of 56 of 59 chambers was more nearly vertical than horizontal. The chambers were closed on all sides, and the inner walls were relatively smooth, perhaps due to addition of a secre- tion as in Altica spp. (Woods 19 17, 19 18). The length of 1 1 cham- bers (average 1.05k0.10 cm) was substantially greater than that of the pupae in these chambers (average 0.68L0.02 cm). Of 59 pupae of M. j., 56 faced more nearly upward than hori- zontal, 3 were more nearly horizontal than upward or downward, and 0 were more nearly downward. As in Altica spp. (Woods 19 17, 1918, Barstow and Gittins 1973), pupae were covered with stiff bristles, presumably to keep the body out of contact with the walls of the chamber. The especially large bristles at the tip of the abdomen might seem designed to support the pupa's weight, but similar enlarged bristles occur in Altica bimarginata, which lies on its dorsum as a pupa (Woods 1917).
Young pupae were orange with black setae, while the legs and heads of M. j. pupae about to moult to adults were blue, and the abdomen was swollen and extended directly rearward instead of curling ventrally. The pupal cuticle ripped near the anterior end, and the adult pulled its antennae and legs free as movements of the abdomen pulled the cuticle posteriorly. After freeing itself from the cuticle, the adult beetle extended its hind wings, projecting them posteriorly and ventrally from under the clear elytra. Individ- uals in captivity (on moist paper towels) braced their middle and hind legs laterally during this stage. Presumably beetles in pupal chambers braced themselves in the upper end of the chamber, thus giving the hind wings room to expand in the lower portion of the chamber. This would explain why pupation chambers were longer than the pupae. Young adults remained underground at least briefly before emerging on the surface. Sex ratios were close to 50:50.



================================================================================

19931 Eberhard, Achoy, Marin & Ugalde 101 48% of 914 M. sp. and 42% of 488 M. j. raised to adults were males.
Males did not emerge before females, as in some other chrysomelids (Waloff and Richards 1958). Adult females may emerge from pupation with more reserves than males. Thus of 27 survivors of a group of 88 adult M. j. which emerged in a container with moist earth and were kept with no food for about 7 days, 3 were males and 24 females (p<0.001 compared with overall sex ratio).
Host Plants
As in many other alticines (Jolivet 1988), both species were oligophagous. They fed on different but overlapping sets of host plants. Adult and larval M. sp. occurred on Gunnera insignis (Gun- neraceae) (approximately 50-150 aggregations of adults seen, totalling probably 10-25,000 beetles near Rio Hondura, Zurqui, Cerro de la Muerte, Tapanti, and about 40 km NNW of San Jose on the road to Puerto Viejo de Sarapiqui), but M. j. was never seen on this plant. On the other hand M. j. but not M. sp. adults were com- mon on Ludwigia hexapetala (Hook and Arn.) (Onagraceae) (more than 30 aggregations totalling probably >10,000 beetles seen at Rio Reventado, and Rio Barquero near Cartago). Adults and larvae of both species occurred at least occasionally on L. octovalvis (Jacq.) (estimated 100-200 aggregations of adults) and L. peru- viana (L.) (estimated 10-20 aggregations of adults) in and near the Universidad de Costa Rica in San Pedro de Montes de Oca, San Antonio de Escazu, Escazu, San Isidro de Heredia, and Cartago; M. j. was consistently more common than M. sp. on L. octovalvis (Table 2). Both species were also found on Cuphea sp. plants (Litraceae), but in monospecific aggregations (M. j. on C. calo- phylla Cham. and Schlechtd near Alajuelita about 5 KM E of Escazu, and on C. sp. near San Vito de Java, San Antonio de Escazu, and the Universidad de Costa Rica; M. sp. on C. sp. in Zurqui). Scattered adults of M. sp. were seen feeding on Tabachina longifolia (Melostomataceae) adjacent to large aggregations on G. insignis near Zurqui, and near San Antonio de Escazu. The four families of host plant are not closely related, but all are early secondary, weedy species. Ludwigia and Cuphea grow in humid soil, often with slowly moving or standing water, and are



================================================================================

102 Psyche [vo~. 100
Table 2. Proportions of M. jamaicensis and M. sp. in mixed aggregations in the field on L. octovalvis plants (all purple individuals were classified as M. sp.; some may have been hybrids). An estimated minimum of 50-100 other aggregations of only M. j., and 10-30 mixed aggregations were also seen on this plant. (OCH = Ochomogo, just W. of Cartago; SAE = San Antonio de Escazu; SIH = San Isidro de Heredia; UCR = Universidad de Costa Rica). Date Site Proportion Total
M. sp. M. j.
30 V 84
6 1 84
1 VI 85
1 VI 85
14 VII 85
26 IX 85
23 IX 85
26 IX 84
VI-VII 83
22 I1 87
28 I11 87
30 XI1 85
1 I11 86
8 VI 85
10 VIII 85
18 1X 87
UCR
UCR
UCR
UCR
UCR
UCR
UCR
OCH
SAE
SJC
SJC
SAE
SAE
SAE
SAE
SIH
common in roadside ditches carrying effluents from houses in sub- urban and rural situations. Gunnera and Tabachina differ in being associated with well drained soils rather than swampy substrates, but grow in recently disturbed areas; Gunnera is associated with very high rainfall (Fernandez 1984).
Individual beetles can and probably sometimes do eat more than one food plant in nature. In one case, M. j. larvae (identified by association with mono-specific aggregation of adults) occurred in large numbers on plants of L. octovalvis which were nearly stripped of leaves, and in lower numbers on immediately adjacent Cuphea calophylla plants. In another case, a large aggregation of adult M. j. which was checked for several days first decimated a long row of L. octovalvis plants, then moved onto (and killed) sev- eral small Cuphea sp. plants before disbanding. Adults and larvae of M. sp. collected on G. insignis fed readily on L. hexapetala and L. octovalvis in captivity.




================================================================================

19931 Eberhard, Achoy, Marin & Ugalde 103 The nearly complete absence of M. sp. on L. hexapetala (3 of 5,000-10,000 beetles), and the complete absence of M. j. on G. insignis (0 of several 10's of thousands of beetles) is probably related to the relative inability of the larvae of these species to develop on these plants (Table 3). In contrast, Fl hybrid larvae matured successfully on both of these species (Table 3). It appeared that G. insignis was more "difficult" for both beetle species as well as for hybrid larvae, as larval weight at maturity was lower for all crosses, and development times were longer (Table 1, see also oviposition rates below). Lower larval weights did not, however, generally lead to higher pupal mortality, except in the extreme case of the few small M. j. raised on G. insignis (Table 3).
Reproductive Behavior
Females in captivity began to oviposit several days after emerg- ing above ground. Feeding on living rather than recently picked leaves, and in more humid conditions lowered the number of days between first copulation and first oviposition, and raised the rate of egg production during the first days of oviposition (Table 4). Oviposition rates in the field may be lower. Of ten field- captured female M. j. placed on L. octovalvis plants, only 50% oviposited during the first 24 hours; 30% took 48-72 hours to oviposit; and 20% had still not oviposited 17 days later. Data for 12 M. sp. collected on L. octovalvis were 17% oviposited in the first 24 hours, 33% after 24-72 hours, and 50% after 3-7 days. The average numbers of eggs laid in these clutches were not high (17.2k17.0 for M. j., 10.8k8.7 for M. sp.; compare with Fig. 3). -
Table 3. Survivorship of Macrohaltica Jamaicensis, M. sp. and cross-specific hybrids raised on three different food plants. (/'. = M. jamaicensis; sp. = M. sp.) A is the proportion of larvae born which grew to mature larvae; B is the proportion of larvae born which emerged above ground as adults. --
Ludwigia hexapetala L. octovalvis G. insignis Cross N survivorship N survivorship N survivorship female male A B A B A B
-
j. X j. 808 0.54 565 0.95 0.93 804 0.04 0.00 sp. X sp. 305 0.39 177 0.75 0.71 181 0.86 0.56 j. X sp. 410 0.57 469 0.87 0.74 327 0.77 0.63 sp. X j. 109 0.58 0.50 131 0.87 0.69




================================================================================

104 Psyche [vo~. 100
Table 4. Effect of conditions in early life on length of period before oviposition began in female M. jamaicensis, and on rate of egg production during the first four days of oviposition. Condition A: female kept on tender leaves of L. octovalvis plant from first day to adult life; Condition B: female kept in petri dish with moist paper towel and supplied with one newly picked leaf of L. octovalvis each day; Condition C: as in B but without moist towel. Males were combined with females, allowed a single copulation, then removed 1-2 hrs after copulation ended on the first day of mating. Similar matings were staged a day later, and, in conditions B and C, on each of the following two days. All pair-wise differences are highly sig- nificant (p<0.001) except number of days to first oviposition B vs. C, in which p<0.01) (Mann Whitney U Test).
--
Condition
- --
A B C
Age at first copulation (days) 7.5k0.5 9.0kO.O 8.4k1.7 No. of days from first copulation
to first oviposition 2.4k0.5 7.6k0.9 9.6k2.8 No. of eggs laid in first four days
of oviposition period 134.4k74.9 73.0k23.1 37.2k18.5 N (females) 12 2 1 13
Captive female M. sp. feeding on L. octovalvis laid many more eggs than those feeding on G. insignis. Average daily rates for the first 24 days of oviposition by 26 individuals on L. octovalvis was