Insects are the only invertebrates known to use fruits and seeds of Prosopis as a food resource (Kingsolver et al. 1977). Feeding by insects may result in total destruction of fruits (e.g. Lepidoptera) but the most common result is that fruits or seeds are aborted or that seeds are destroyed. I refer to seed-feeding insects as predators because they destroy an entire organism (i.e. a seed) but the insects could also be described as parasites when their feeding does not interfere with the ability of a seed to germinate.
For convenience, I follow Kingsolver et al (1977) by dividing the insects that feed on fruits and seeds of Prosopis into two groups, those that feed principally from the outside and those that feed from the inside. Those that feed from the outside are adults and nymphs of Hemiptera and larvae of Lepidoptera. Internal feeders include larvae of Lepidoptera and Coleoptera of the families Curculionidae, Cerambycidae and Bruchidae. Of the beetles, bruchids are by far the most important pests. To some extent the age of the fruit influences the type of feeding by insects. Hemiptera have piercing-sucking mouthparts so soft immature Prosopis fruits are readily fed upon, but dry, woody pods are neither attractive to them nor readily punctured. Larvae of Lepidoptera that feed externally also utilize soft, immature pods. The age of the pod, on the other hand, is not as important for most of the internal feeders because most of them feed on seeds and the seeds are often soft. Gaining entry into the pod and seed is the most crucial step for internal feeders, at least for those that have been most studied, the bruchid beetles. All internal feeding known is by larval stages. Adults of internal feeders are thought to feed on nectar and pollen, although some bruchid beetles do not feed as adults.
True bugs (Hemiptera) are important native pests of mesquite. According to Ueckert (1973) the leaf-footed bug, Mozena obtusa Uhler (Coreidae), destroys large quantities of pods of Prosopis glandulosa in North America. In Texas, these insects become active in early spring and can remain active until August. Damage to mesquite in Texas was estimated at 50% of the potential fruit production by Swenson (1969). This is the most frequently found hemipteran on young fruits of Prosopis velutina near Tucson, Arizona (Werner and Butler 1958). Ueckert (1973) found that abortion of immature pods of P. glandulosa varied from 33% to 89% when fed upon by these bugs. In addition, feeding of the bugs reduced the dry weight of pods from 23% to 59% and the seed germination rate was reduced to 4.3% from a normal rate of 65%. Since mesquite reproduces only by seeds, the action of this natural enemy may be significant in reducing the reproduction of the plant. Another bug, conchuela (Chlorochroa ligata (Say)) (Pentatomidae), destroys mesquite seeds in Texas by sucking juices from seeds, leaving only dry seed coats in the pod (Smith and Ueckert 1974). Pods in control cages produced seeds with a much higher germination percentage (76%) than those produced in cages with conchuela (0 to 4% germinable). In a study of Prosopis flexuosa in Argentina only 40 of 145 incipient fruits, and in P. chilensis only 4 out of an initial 70 fruits reached maturity (Solbrig and Cantino 1975). According to Kingsolver et al. (1977), it is possible that insect damage similar to that caused by the leaf-footed bug in North America, was responsible for much of this fruit abortion in Argentina. It is also possible that other factors such as water stress and self-abortion may have also contributed to the mortality.
Most larvae of the Lepidoptera that are external feeders consume the flowers, immature fruits and ripening pods of Prosopis (Kingsolver et al. 1977). A number of species of Lepidoptera feed inside the pods of Prosopis in North America. Members of the Lycaenidae (Strymon leda), Olethreutidae (Ofatulena spp.), Pyralidae (Paramyelois spp.), Blastodacnidae (Chaetocampa spp.), Notodontidae (Didigua argentilinea) and Cochylidae (Phalonia leguminana) consume both the insides of the pods and the seeds. The Indian-meal moth (Pyralidae: Plodia interpunctella (Hubner)) has been reported to breed in considerable numbers in mesquite pods in storage in Hawaii (Bridwell 1920b). Eggs of this moth are laid on bags in which pods are stored or on pods, and then caterpillars enter the pods. They preferably enter broken pods where they feed on the sugary pulp. When the caterpillars are mature, they leave the pods and seek a sheltered area where they pupate. According to Koch and Campos (1978) Cryptophlebia carpophagoides Clarke (Lepidoptera: Olethreutidae) feeds in pods of Enterolobium contortisiliquum Vell. in Argentina, in Prosopis tamarugo and P. juliflora in Chile. This species feeds on the inside of pods on tender seeds and tissues. They frequently attach their excrement to the outside of the fruit. These authors attribute as much as 30% loss of pods of P. tamarugo to this moth. Habit et al (1981) described the life cycle and structure of this moth in Chile and reported that it damaged 17.4% of the fruits of P. tamarugo. Koch and Campos (1978) and Habit et al. (1981) reported that Leptotes trigemmatus Butler (Lepidoptera) also fed upon the fruits of P. tamarugo in Chile. This species prefers to feed upon fruits in their primary stages of development, which eventually causes the fruit to fall.
In Arizona three species of Curculionidae (Apion subornatum, A. ventricosum, Microtychius sp.) feed in seeds of Prosopis velutina (Kingsolver et al. 1977). In Argentina, the weevils Sibinia asulcifera and S. mastuerzo probably feed in seeds of Prosopis (Clark 1979). Beetles of the genus Lophopoeum (Cerambycidae) are known to develop inside the pods of Prosopis in Argentina and to consume all of the seeds within a pod while developing (Kingsolver et al. 1977).
Seed beetles (Bruchidae) are by far the most numerous (in both numbers of species and individuals) and best known insects that feed in fruits of Prosopis in the New World (Kingsolver et al. 1977). The larvae of burchids feed in the seeds of a reported 32 plant families (Johnson 1981a) but most species feed in one of the families, the Leguminosae. Bruchids are closely related to the families Chrysomelidae and Cerambycidae. Although the other two families feed in a variety of plant parts, bruchids are known to feed only in seeds. At present the family Bruchidae consists of about 1 300 species, grouped into 56 genera in the subfamilies Amblycerinae, Bruchinae, Eubaptinae, Kytorhininae, Pachymerinae, and Rhaebinae (Johnson 1981a). Most (80%) species of bruchids are presently assigned to the subfamily Bruchinae.
A typical life history of a bruchid is illustrated in Figure 1 but seed beetles attack seeds in a variety of ways (see below). Their life history is usually that the adult female lays eggs on a seed or pod (Figure 1), the first stage larva chews through the egg shell, pod wall and/or seed coat and then into a seed. The first stage larva (Figure 1) is highly modified to enter seeds and has many spines, hairs, etc. for this purpose (Pfaffenberger and Johnson 1976). Shortly after entering a seed it molts into a leg-less grub that is very different from the first stage larva and is modified for feeding inside seeds. The larva usually feeds inside one seed, or in some bruchids, two to several seeds, molts usually three more times as it continues to feed and increase in size. It usually then pupates inside a single seed, although some species leave the seed and pupate in a cocoon, while other species glue several seeds together as a pupal chamber.
After pupation the adult completes a typical round exit hole (Figure 1) that was almost completed by the larva and leaves the seed to begin a new life cycle. The duration of the life cycle varies but it usually is about 30 days. Adult bruchids probably feed on nectar and pollen and are not known to feed on, or in seeds, except incidentally, such as when emerging from a seed or when a female chews a hole in a pod and then lays eggs in it. Some species of bruchids, especially those of economic importance, survive for many generations in containers of seeds in the laboratory or in storage without the adults feeding.
Many host plants are now known for bruchids (Johnson 1981a). This does not mean that an enormous amount is known about bruchid hosts. In fact, recent field studies lead the author of this Handbook to believe that bruchids feed in seeds of many more plants, especially Neotropical Leguminosae, than are now recorded.
 | |
 |  |
 |  |
 | |
Figure 1. Composite life cycle of bruchids on pods of Prosopis. a. Eggs glued to pod surface or laid in cracks in pod or in emergence holes of an adult bruchid (round holes). b. Entry holes of first stage larvae that have burrowed through pod wall and first stage larva enlarged to show hairs, spines and legs which are modifications for entering seeds. c. Cross section of pod and seed showing the burrow made by the entering first stage larva. d. Later stage larva inside cavity chewed in seed. e. Pupa inside larval feeding chamber. f. Adult emerging through hole prepared by last stage larva.
Five genera of Bruchidae are known to feed only in the seeds of species of Prosopis (Table 1, page 33) but five other genera feed in Prosopis and seeds of plants found in close association with Prosopis such as Acacia, Cerecidium and Parkinsonia. In the New World, about 32 species of bruchids feed in Prosopis fruits. Of these species, 28 are obligately restricted to Prosopis, but individual species of bruchids may feed on several species of Prosopis. The bruchids and their host plants are listed in Table 2 (page 34) and the Prosopis species and the bruchids that feed in their seeds are in Table 3 (page 38).
Species in the genera Algarobius, Neltumius and Mimosestes are the principal Prosopis seed-infesting bruchids from the United States to Venezuela (Table 1). All species in these three genera that feed in Prosopis are obligately restricted to Prosopis except for Mimosestes amicus and M. insularis. The other hosts of the latter two species are listed in Table 2.
The life histories and damage potential to seeds of Prosopis have been studied most extensively for Algarobius prosopis, Mimosestes amicus, M. protractus and Neltumius arizonensis (Swier 1974, Kingsolver et al. 1977, Conway 1980), so available data are reviewed for these four species below.
This species is by far the most abundant in seeds of North American Prosopis and thus has great potential for damage to mesquite seeds. According to Swier (1974), this species probably overwinters in seeds as immatures or adults in Prosopis velutina in Arizona. Adult females emerge in the spring and lay eggs on the remnants of the previous year's seed crop. When small, immature pods become available on trees in late spring (May, June), females lay eggs on them and continue to lay eggs on maturing and mature pods until late fall. A. prosopis does not glue eggs to pods or seeds as many bruchids do, but the female inserts them into cracks and crevices in the pod (Bridwell 1920a, Swier 1974). Often emergence holes of adults are utilized as egg-laying sites by this species. Larvae of A. prosopis are very motile, possessing well-developed legs and sensory setae (Pfaffenberger and Johnson 1976), which correlate well with strong locomotory behaviour. Inside very immature pods the larvae probably feed on the syrup in the pods until the cotyledons develop (Bridwell 1920b). On more mature pods the larvae either enter the first seed they encounter or crawl through the pod pulp to enter another seed. In storage larvae may bore through the seed coat from the surface of the pod. After feeding and molting about three times inside a single seed, the larvae line the chamber they have produced with wastes and then pupate.
Under natural conditions, Algarobius prosopis has three or more generations per year. According to Bridwell (1920b), this species will not breed in dry, stored mesquite pods but will continue to breed if the pods are moist. Swier (1974) has observed larvae of this species to enter dry pods. More research needs to be done using both dry and moist pods to determine the potential for A. prosopis to destroy Prosopis seeds in storage.
Under natural conditions, for a given crop from one tree this species may destroy from 8% to 75% of the seeds (Swier 1974, Glendening and Paulsen 1955, Kingsolver et al. 1977). Swier found that this species accounted for 93% of the bruchid predation of seeds of Prosopis velutina in Arizona.
No ecological studies of Algarobius bottimeri Kingsolver have been made, but it is assumed that it has ecological characteristics that are very similar to A. prosopis. According to Kingsolver et al. A. bottimeri has been introduced into the Hawaiian Islands where it feeds on Prosopis pallida. Conceivably, the ecological studies of “Bruchus prosopis” by Bridwell (1918, 1920 a, b) actually were made with A. bottimeri.
Applications of insecticides to control A. prosopis should probably begin when pods are immature and the presence of adults of this species can be verified.
In a given crop under unique conditions, this species may feed in a significant number of Prosopis seeds. According to Swier (1974) and Conway (1980), however, this species causes far less damage than A. prosopis. This species is similar in its life cycle to A. prosopis except that M. amicus lays eggs in late spring (June) only on immature pods with well developed cotyledons and on mature pods, it cements eggs randomly to the pod, often superimposing eggs on top of each other, and the larvae enter the pod directly through the bottom of the egg. Swier (1974) found that this species will destroy up to 3% of the seeds of Prosopis velutina in Arizona. This species feeds in seeds other than Prosopis (Tables 2, 3).
As with M. amicus, this species may be locally abundant and destroy significant numbers of seeds of Prosopis. The life cycle of M. protractus is similar to A. prosopis except that it probably over-winters as an adult (Swier 1974), it ovipsits only on immature pods and is the first bruchid species to utilize pods in the late spring and the first to leave pods in late summer. M. protractus preferably cements eggs near the pod suture, the larvae are less motile and enter the seed pod directly through the bottom of the egg and, since it only utilizes immature pods, only one generation per year is produced by this species in P. velutina in Arizona (Swier 1974) Although M. protractus is a somewhat rare bruchid, Conway (1980) found that it accounted for 61% of bruchid damage to P. velutina seeds near Black Canyon City, Arizona, while M. amicus and A prosopis accounted for 22% and 14% of damage respectively. Thus, under certain ecological conditions, uncommon bruchid beetles may breed in abundance.
This uncommon species has not been reported to destroy large numbers of Prosopis seeds, even under local conditions. Its life cycle is similar to A. prosopis except that it does not lay eggs on extremely immature pods but on those that have developed cotyledons in the seeds and it cements eggs randomly on the pod and the larvae bore through the bottom of the egg directly into the pod. N. arizonensis will breed continuously in seeds and probably has about three generations or more per year (Swier 1974). In a given tree this species may destroy up to 3% of a seed crop but usually it feeds on less than 1% of the seed crop of P. velutina in Arizona (Swier 1974). Conway (1980) found that the numbers of N. arizonensis were insignificant compared to the other three species of mesquite burchids in his studies in Arizona.
Species in the genera Acanthoscelides, Pectinibruchus, Rhipibruchus and Scutobruchus are the principal Prosopis seed-infesting bruchids in Argentina, Chile and Peru. Species in the latter three genera are obligately restricted to feeding in seeds of Prosopis (Tables 1, 2, 3).
The life histories and damage potential have not been studied nearly to the extent as those in North America. According to Kingsolver et al. (1977) there are behavioral differences among species of these genera. Species of Scutobruchus lay eggs individually on immature and mature pods and on pods both on the tree and on the ground. One species of Rhipibruchus lays its eggs in partially overlapping stacks similar to those of Mimosestes amicus in Arizona (Swier 1974). Adults of Rhipibruchus emerge from pods of Prosopis chilensis before those of Scutobruchus and the females emerge before the males. Species of both genera overwinter as adults that hide in sheltered places such as old pods or fissures in bark. Unlike species of North American bruchids, adults of these two genera do not visit flowers.
Reyes and Hermosilla (1974) studied the life cycle of Scutobruchus gastoi Kingsolver in the laboratory in fruits of Prosopis tamarugo. They described the egg, larva and pupa and found that in the laboratory the life cycle varies from 80 days to four months. Habit et al. (1981) report that in the field this species overwinters in pods on the ground and emerges from the pods when flowering begins again in P. tamarugo.
According to Kingsolver et al. (1977) many of the same types of interactions observed among species in seeds of P. velutina also occur in bruchids infesting algarrobos in Argentina. Detailed sampling of nine trees of P. flexuosa in Argentina showed that the total seed destruction rose logarithmically over time from the ripening of the first pod to about 75 days later. Counts of fruit predation in North America showed a levelling off of predation through time. The proportional reduction in predation was due to fruits becoming more difficult to locate as the season progresses. At the study area in Argentina, a levelling off of predation did not occur because domesticated sheep and goats removed all of the pods that fell to the ground. If not for the sheep and goats, a levelling off of predation by bruchids could be expected as is the case in North America.
Scutobruchus ceratioborus destroyed 26% of the seeds of Prosopis flexuosa and 90% of those of P. chilensis in Argentina (Kingsolver et al. 1977), and it seems to be the South American ecological equivalent of Algarobius prosopis in North America.
The amount of damage by bruchids varies considerably from tree to tree both in Arizona and Argentina (Kingsolver et al. 1977). In nine trees of P. flexuosa in Argentina, seed damage varied from 0.7% in the fallen pods of one tree to 53% in those of another. Similar variation in seed damage by bruchids from tree to tree was reported by Swier (1974) in P. velutina in Arizona. Swier (1974) found the difference in the amount of seed damage can be ascribed in part to differences between individual trees. Other factors that contributed to variation in the amount of seed destruction were the site of the mesquite population, the amount of seed and fruit production by a given tree, the direction of the source of bruchids from other trees, and the relative height of pods in the canopy. However, the significance of the amount of the contribution of each variable differs between each species of bruchid. For example, Swier (1974) found there was a significant tendency for most species except Neltumius arizonensis to prefer fruits in the upper parts of Prosopis plants.
Caryedon serratus feeds in seeds of Prosopis in Hawaii and Israel (Bridwell 1920b, Belinsky and Kugler 1978). This Old World species feeds primarily in seeds of tamarind (Tamarindus indica) and thus has been introduced into most tropical areas along with tamarind. The bruchid has been introduced into the New World and is widely distributed in tamarind in Mexico. C. serratus is a serious pest of peanuts (Arachis hypogaea) in Africa. Because peanuts were introduced into Africa from the New World, this species has the potential to switch hosts and become a serious economic pest. Potentially, then, it could become a problem in New World Prosopis. Belinsky and Kugler (1978) studied its life history in Prosopis in Israel as did Bridwell (1918, 1920b) in Hawaii. Belinsky and Kugler (1978) found that the subspecies they studied, Caryedon serratus palaestinicus Southgate, preferred its natural host, Prosopis farcta, to peanuts under laboratory conditions, and reached the conclusion that this subspecies was not a threat to peanuts in Israel.
The genus Acanthoscelides contains more species (about 300) than any other genus of New World bruchids (Johnson 1981b). Because the important pest of beans, Acanthoscelides obtectus (Say), is in the genus, significant amounts of data have accumulated about the ecology of A. obtectus and other, non-economic species of Acanthoscelides (Johnson 1970, 1981b). Because of its many species and varied host plants (nine plant families) little can be predicted about the behaviour and pest potential of Acanthoscelides that feed in Prosopis seeds.
The genus Amblycerus is similar to Acanthoscelides in that it is large and its species feed in a variety of plant families. Because three species of Amblycerus are reported from Prosopis, these species pose a threat to seeds of other species of Prosopis.
It has been suggested that most of the specificity and competitive advantages of bruchids on host seeds is due to precise chemical relationships (Kingsolver et al. 1977). This does not seem to be the case with bruchids that feed in Prosopis. Competitive interactions seem to limit the numbers of bruchids on a given species of Prosopis. It may be argued that there is a lack of chemical specificity because the seeds are not known to be toxic, that there are numerous examples of shifts to new hosts by bruchids when the Prosopis flora is lacking, and there are several species of the bruchid genus Stator that breed freely in the seeds of Prosopis velutina under experimental conditions in the laboratory (Johnson 1981d). In the last hundred years Prosopis pallida, a native of Peru, has been introduced into the Hawaiian Islands as a source of food for livestock (Fosberg 1966). Several species of bruchids have also been introduced inadvertently. These include Mimosestes nubigens (M. sallaei), a species normally found in seeds of six species of Acacia in North America (Kingsolver and Johnson 1978), Caryedon serratus, and Algarobius bottimeri that feeds only on P. glandulosa in Texas (Kingsolver et al. 1977). In the Hawaiian Islands where these bruchids have been introduced, all use the pods of P. pallida. All of these data indicate that the fruits of Prosopis are vulnerable to consumption by bruchids and that up to four species may coexist in the same locality. This coexistence may be due in part to their attacking a host at different times during the season (Figure 2) or by laying eggs on fruits at different heights in the tree, or by differential ability to use other hosts.
Figure 2. Periods of oviposition and stages of pods of Prosopis on which eggs are laid of four species of bruchids that coexist in P. velutina pods in Arizona.
Various species of bruchids are known to have unique life histories that are modified to allow them to feed in seeds of specific hosts. For example, the adult bruchids that attack seeds of Prosopis are modified to lay eggs on the outside of or in cracks of indehiscent, woody pods. The larvae are specialized to penetrate the pod and seeds. Johnson (1981c) recently found that legume pods that are indehiscent, dehiscent and partially dehiscent are all fed upon by different guilds of bruchids. He found that the bruchid genera in the guild that lays eggs on legume pods are different from the genus that oviposits on seeds. He also found that seeds of some species of legumes are attacked by another guild of bruchids only after they have fallen to the ground. This latter guild, the “scattered seed guild”, is composed of several species of Stator that are very specific to the seeds of their respective hosts. What is important in the context of this handbook on Prosopis insects is that under experimental laboratory conditions Johnson (1981d) found that several species in the scattered seed guild develop readily in non-host seeds. One species, Stator sordidus (Horn) was able to complete its development in 20 different species of non-host seeds. What was most interesting was that four species of this guild developed in seeds of Prosopis velutina, a non-host. Of significance to future research on Prosopis bruchids is that bruchids in this guild may actually feed on Prosopis seeds in nature and in storage.
A number of recent studies have described the ecology and evolutionary interactions between bruchids and their host plants (Janzen 1969, 1971 a,b,c, 1980; Forister 1970; Center and Johnson 1974; Johnson and Slobodchikoff 1979; Johnson 1981c). According to Janzen (1969), most legumes seem to follow one of two strategies to avoid seed predation by bruchids. Some plants produce large toxic seeds that exclude all or almost all seed predators. Other plants produce many small seeds that are fed upon by bruchids but produce so many seeds that some escape the predators. Center and Johnson (1974) and Johnson (1981c) presented evidence that showed that bruchids had evolved mechanisms to counter the protective devices of the plants. For example, some bruchids are able to feed in very toxic seeds, while others feed in small seeds after they have been dispersed. Janzen (1969) discussed about 31 protective devices that plants have evolved to counter predation by bruchids. Species of Prosopis apparently have evolved the strategy that produces a very large seed crop that ensures that at least some of the seeds will escape predation by bruchids.
More specific mechanisms that legumes use to minimize the effects of predation by seed-feeding insects are that the trees that grow in relatively constant, equable environments have the option of staggering of the fruiting times of individuals so that insect populations cannot build up, the formation of fleshy seeds that dry out too quickly for a predator to use as larval food, or the production of seeds that are so small that they contain too little food to support the complete development of an insect. According to Kingsolver et al. (1977) none of these options is open to Prosopis in desert environments because fruiting of all individuals is synchronized to occur at the most advantageous time of the year for seed germination, and the seeds must be capable both of withstanding desiccation and of providing a good initial energy source for the new seedling. Further research is necessary before the hypotheses of Kingsolver et al. (1977) can be accepted or rejected.
Most species of bruchids are extremely host specific and most hosts support few bruchids in their seeds. Johnson and Slobodchikoff (1979) found that 82.5% of bruchids feeding in Cassia had three or fewer hosts and 90.5% of species of Cassia have three or fewer species of bruchids in their seeds. Johnson (1981b) reported that about 70% of species in the bruchid genus Acanthoscelides has one to two hosts and 74% of their hosts have one to two species of bruchids feeding in their seeds. Similar results were presented by Janzen (1980) when he discussed reasons for host specificity of seed-feeding beetles in Costa Rica.
