Life History of Abedus herberti in Central Arizona (Hemiptera: Belostomatidae)

Torre Bueno in I9O6 noted the meagerness o information on the biology and immature stages oi: aquatic Hemiptera. Although this condition has changed somewhat in the more than 60 subsequent years, complete lie histories or representatives o only two o the five genera belonging to the subfamily Belostomatinae (Lauck and Menke 96 are available. Belostoma is represented by B. flumineum Say (Torre Bueno 9o6), B. oxyurum Dufour B. bi[oveolatum Spinola (Schnack 97 ), and B. malkini Lauck (Cullen I969). The Old World genus Limnogeton is represented by L. fieberi Mayr (Voelker x968). Menke (96o) was successful in bringing a single individual Abedus dilatatus (Say) to its fifth instar, at which time it died. Life histories for representatives of the genus Lethocerus include: L. americanus (Leidy) (Rankin 935), L. cordofanus Mayr (nee. L. niloticus Stl) (Tawfik 969), L. mazzai De Carlo (De Carlo 962), and L. maximus De Carlo (Cullen I969). Nothing is known of the immature stages of Horvathinia (subfamily Horvathiniinae, Lauck and Menke 96I). The taxonomy and distribution of dbedus herberti, as well as a behavioral and ecological sketch of the genus, is provided by Menke (I96O).


INTRODUCTION
Torre Bueno in I9O6 noted the meagerness o information on the biology and immature stages oi: aquatic Hemiptera. Although this condition has changed somewhat in the more than 60 subsequent years, complete lie histories or representatives o only two o the five genera belonging to the subfamily Belostomatinae (Lauck and Menke 96 are available. Belostoma is represented by B. flumineum Say (Torre Bueno 9o6), B. oxyurum Dufour B. bi-[oveolatum Spinola (Schnack 97 ), and B. malkini Lauck (Cullen I969). The Old World genus Limnogeton is represented by L. fieberi Mayr (Voelker x968). Menke (96o) was successful in bringing a single individual Abedus dilatatus (Say) to its fifth instar, at which time it died. Life histories for representatives of the genus Lethocerus include: L. americanus (Leidy) (Rankin 935), L. cordofanus Mayr (nee. L. niloticus Stl) (Tawfik 969), L. mazzai De Carlo (De Carlo 962), and L. maximus De Carlo (Cullen I969). Nothing is known of the immature stages of Horvathinia (subfamily Horvathiniinae, Lauck and Menke 96I). The taxonomy and distribution of d bedus herberti, as well as a behavioral and ecological sketch of the genus, is provided by Menke (I96O).
METHODS AND MATERIALS I collected adult bugs, including males encumbered with eggs, with the aid of a gasoline lantern at night from the following central  (lO.5 by 4.5 cm), provided with aquarium gravel and deionized water to a depth of about 4 cm, received hatchlings obtained from eggs ot: field-captured encumbered males. Gravid females, paired with unencumbered males, produced fresh eggs for incubation studies. First and second instar nymphs were fed on cultured vestigial winged Drosophila which were floated on the water's surface.
Third and fourth instars received legless crickets (Acheta domestica), and fifth instars ate intact crickets. I renewed food and water every other day, and recorded molts which had occurred each day. Cast skins were preserved in seventy percent alcohol, one percent glycerin for study and measurement. Open water temperature in the laboratory was a reasonably constant 18 C (+/-0.5 C).
I observed hatching and used still and motion picture photography to record this event for analysis. I measured eggs, first, second, and third instar nymphs and the legs of fourth and fifth instars using a stereoscope ocular micrometer, and the bodies of fourth and fifth instars with a dial caliper.

EGGS
The eggs of Abedus are always laid on the back of the male ( Fig. IA). Oviposition, which continues for 12-48 hours in the laboratory, alternates with copulation. Encumbered males are ound throughout the year in Arizona. A detailed study of courtship and male brooding behavior in A. herberti will be presented elsewhere (Smith, in prep.).
A mucinous glue is secreted by the female prior to the deposition of each egg starting at the apex of the male's hemelytra and proceeding forward, uniformly covering the male's back up to and sometimes including the pronotum. It is usually the case that males carry the eggs of only one female, but polygynous matings apparently do occur (Menke 196o). I inferred rom data on hatching, that of seven encumbered field captured males, only one individual had consorted with two emales to acquire his load of eggs. In this instance, 72 eggs hatched over a five day period, this followed by three days during which no eggs hatched. The remaining 15 eggs hatched over a 24 hour period. In all other cases, hatching occurred more or less continuously over a three to five day period.
Freshly laid eggs begin swelling immediately. Swelling has the 8mithMLife History of Abedus herberti 275 ultimate effect of compacting the mass, enhancing its structural integrity and conformity to the male's back. The resulting gelatinous plate with its embedded eggs is strong, flexible and adheres tenaciously to the male's hemelytra so long as it is kept moist. If exposed to air or 6o minutes or longer, the gelatinous pad becomes brittle, loses its adhesive properties and on prolonged exposure will eventually t:all off.
Eggs increase in size from mature oviducal (3.o64 -4-.o22 mm by 1.732 ---+ .oi7 mm, n 5o) to those containing nymphs ready to hatch (4.983 .o36 mm by 2.OOl .Ol 7 mm, n = 5o). This represents a 59 percent mean increase.in length and a 15.5 percent mean increase in width. The initial swelling of the egg (during the first 48 hours) may be due to imbibition of water, however later increase in size seems to be due to the development of the embryo. (925)  Mature oviducal and freshly laid eggs are variable in shape, usually imperfect oval with one nearly straight side, yellow in color and irregularly hexagonal microreticulate. Soon after occlusion, eggs darken to tan. The cap (upper ) differentially darkens and remains several shades darker until just prior to eclosion of the nymph at which time it becomes grayish white in color. As development proceeds the shape of the egg comes to resemble an elongate "printed comma"; that is broad and rounded at the free end, narrow and pointed at the attached end, with one convex and one concave side. The concave side seems always to face the posterior of the mass. Cobben (1968, Fig. 232) shows the head of the embryo (of Sphaerodema Diplonychus) facing the attached end, but embryonic rotation must occur prior to hatching. I observed one instance of egg failure due to nonrotation of the embryo. INCUBATION Male brooding of eggs, a complex of behaviors, is necessary or achievement of the greater than 95 percent survival observed or eggs of field-captured encumbered males. I never observed any type of egg parasitism. Egg failure, in every case but the one, seemed due to insemination failure; usually a few eggs on each plate did not develop. Open spaces in the laying pattern in Figure A are undeveloped eggs. Harvey (19o6)  coris macronyx (--Abedus indentatus Haldeman) drop off and are replaced (presumably by the original or another female) by others as late as the sixth or eighth day of incubation, but in every instance I observed, nonviable eggs were retained in the nidus through incubation and hatching of the viable eggs. Incubation periods from laying of the first egg to hatching of the. first ranged from 2 to 23 days in the laboratory. Voelker (968) has shown developmental time for eggs of Limnogeton fieberi to. be directly correlated with temperature, the optimal temperature being 32-33 C. He reports an Psyche [June of eggs for A bedus could be substantially accelerated by increasing the temperature of the incubation water. Open water temperature in my laboratory (18 C) fell within the range of spring, fall and winter stream temperatures in Arizona. HATCHING Twenty-four hours prior to hatching, the free end of the egg swells dorsally. Pressure from within eventually causes a rupture in the chorion around the cephalic cap (Fig. B). Invariably, nymphs hatch facing the brooding male's posterior. I can't explain this constant orientation of hatching nymphs, but its effect is that adjacent eggs can hatch simultaneously with a minimum of interaction between eclosing nymphs. The cap is lifted by the head as the nymphal thorax emerges, but remains hinged to the egg on the side of the latter which faces the ventral surface of the emerging nymph. The nymph's head slips from under the operculum as peristaltic contractions free one half of the body. When the nrmph is two thirds removed, it rests in a position perpendicular to the plane ot? the egg plate (Fig. C). Moments later it arches backward and stretches its legs which are employed in final extraction when the nymph again leans t?orward. A newly emerged bug generally lingers on the back of the male for several minutes to inflate its laterally constricted abdomen (Fig. D). This accomplished, it moves off in search of cover among aquatic plants or debris where it assumes the predatory stance characteristic of nymphs and adults alike. Freshly emerged nymphs are light honey yellow in color and translucent, but darken within one hour.

DEVELOPMENT
Of 31 nymphs reared in the laboratory, 26 survived to adulthood. Individual developmental records are illustrated in Fig. 2 and mean developmental periods are presented in Table I. These results compare favorably with those of Voelker (1968) for Limnogeton and probably reflect critical minimum periods required for morphogenesis between molts under optimal conditions. That availability of food is an important factor related to intermolt time for these sedentary hunters was demonstrated in a simple experiment. Ten of 2o sibling first instars, ed Drosophila daily, all molted in under eight days while the other ten, fed every fourth day, did not begin to molt until the nineteenth day. Thirty additional sibling first instars were di- vided into three groups of ten individuals. Each group was kept at a different temperature (8 C, 22 C, and 3t C) and all nymphs received Drosophila daily in numbers exceeding that which they could consume. All molted at approximately the same time (on the sixth or seventh day), suggesting that temperature is relatively unimportant. These experiment's were not intended to be conclusive and it is probable that a food-temperature interaction could be detected with more sophisticated experimentation.

NYMPHS
There are five nymphal instars for A. herberti. Dorsal views of the five are figured (Fig. 3) and mean morphometric data for fifteen characters of each instar are presented in Table 2. The wide (.oo) confidence limits around each mean should encompass interpopulation variation for central Arizona. Mean size of bugs seems to be stream specific, perhaps due to low dispersal of these animals and reduced gene flow between their populations. A. h. herberti intergrades with Psyche +1 +l +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +l +1 +1 The trst through the ourth instars all have two segmented antennae and one segmented pro-, and two segmented meso-and metatarsi. The beak is our segmented from the first instar to imago.
One antennal segment is added in the fi(th instar and usually another in the adult. All tarsi add one segment in the adult molt. The most noteworthy changes through instars are the allometric development o the wing pads (Fig. 3) and o the protarsal claws.
Torre Bueno (9o6)  probably Belostoma as well, only the magnitude of the size difference between the external and internal protarsal claws increases with each molt due to nondevelopment of the external claw and continued growth ot: the internal. In the (irst instar relative length of the external claw is 50 percent o the internal. This is reduced to about 38 percent in the second, 30 percent in the third, 23 percent in the 1:ourth, and I9 percent in the fifth instar. Curiously, in the adult the external is proportionally slightly longer than in the (ifth instar and the internal claw is 'shorter and broader in the adult. This results in an increase in the relative length o the external claw to about 28 percent that of the internal in the adult bu. LONGEVITY A. herberti is a relatively large animal and is a sedentary hunter, both of which characteristics suggest that it might be long lived. This seems to be the case; several specimens have lived in the laboratory or up to one year and one in particular, a emale collected on October 4, 972, lived until November 3o, 973 during which time she produced our sets o eggs totaling 344. Nymphs rom the. rst through the third instars are probably subject to predation by other aquatic organisms. I ha.ve ound rst instar nymphs in the stomachs o trout taken at three orks o the Black River, Apache County, Arizona. Nymphal instars seem to play an important role in the zSz Psyche [June food economy of the species during periods of low stream productivity; I commonly find nymphs and adults feeding on smaller nymphs of their own species. After the second instar, cannibalism seems to be far the most common cause of fatality among nymphs. Ecdysal ailure and physical catastrophes (flood and drought) also regulate populations. Lethocerus adults, which are seasonally common but never abundant in Sonoran streams, may regularly take adult Abedus but I have observed only one instance of this. In general, A. herberti adults are at the pinnacle of the aquatic food chain, and, having reached adulthood, stand an excellent chance of surviving to produce or brood several clutches of eggs.

DISCUSSION
The life history of Abedus herberti seems to follow closely the pattern reported for other genera belonging to the subfamily Belostomatinae. Allometric development of the protarsal claws closely relates A bedus to Belostoma, and suggests their common ancestry with the two clawed genera. Five instars is the number characteristic of acquatic Hemiptera (Hungerford I92O).