Activity cyde of the ponen wasp , Trimeria howardi ( Hymenoptera : Vespidae ) in Southeastern Brasil

Among the wasps, the Masarínae are a peculiar subfamily. It is the only group of wasps Ihat provi­ sions brood cells with pollen and nectar. The studied species Trimeria howardi Bertoni, 1911, was until recent­ Iy abundant in Ribeirao Preto, Sao Paulo, southeastern Brazil. This paper deaIs with plant species visited by T. howardi, in relation to the species' annuaI cycle of activity, based on periods of nest foundation, flight activity, and development stages of immatures present in the cells. During the study period (five years) T. howardi visit­ ed four species of plants in two botanical families: Bidens pilosa L.(Asteraceae); Xanthium spinosum L. (Asteraceae); Stylosanthes guianensis (Aulb). Sw. (Leguminosae), and Zornia diphylla (L.) Pers. (Leguminosae). Based on the number of visited families, T. howardi can be considered a narrow polylectic species. In general, the species annual activity cycle consists of three pha�es: Active phase: January to July; Inactive phase: August September; Active phase: October to December. The annuaI cyc1e is bivoltine or mul­ tivoltine and the diapause is facultative, because it occurs in a single generation during the year.

While the tribe Gayelliní is restricted to the Neotropical Regíon, the tribe Masarini is mueh more widespread, being represented in the Neartie, Neotropieal, Palearetie, Afrotropi cal, and Australian Regions (Carpenter 1993, Gess 1996).The genus Trimeria belongs to the Masarini and its distribution is restrieted to the Neotropical Region.
Masarinae, within the Vespoidea, are a peculiar subfamily, because they wasps feeds the immatures with pollen and neetar, as bees do; while the remaining wasps nourish their brood with animal protein.Besides, Masari nae wasps transport their provisions in the erop, a eharaeter of searce oecurence among bees.Indeed, only Colletidae bees (Hylaeinae and Euryglossinae) and Lestrimelitta (Api dae: Meliponinae) are exclusive erop pollen transporters (Miehener 1974).Equally nota ble in the group are the morphological modi fieations assoeiated with neetar colleetion: a prolonged and retraetile proboseis (Carpenter 1997).The pollen wasps, like the long-ton gued bees, therefore have the potential to ob tain neetar from a wider range of flower forms than do short-tongued wasps and bees.However, the Masarinae are associated with a relatively small range of flower families visi ted by aculeata Hymenoptera (Gess 1996).
Pollen-wasps nest in cavities built in the soil, in pre-existing cavities, or build cell groups fastened to rocks or branches (Malys hev 1968, Gess 1996).
Trimeria howardi nests in the soil, in bu rrows dug by a female solitarily.The preferen ces are for areas where the vegetation is spar se, that is, quite sunny in well-drained soil (Mechi, in press).
This paper reports on plant species visited by T. howardi, and describes the species' an nual life cyc1e, based on the periods of nest foundation, flight activity, and the develop mental stages oí the immatures found in the cells, throughout the year.

MATERIALS AND METHODS
The research was done for five years in the University of Sao Paulo Campus at Ribeirao Preto, Sao Paulo State, southeast Brazil (21°11'S 47°48'W, 500 to 700 m alt.and 581.53 ha).The c1imate is savanna tropical dry winter (AW, according to Koeppen's c1assifica tion), which is characterized by a very dry win ter, and a hot and h umid summer.The original vegetation was Cerrado (Brazilian savanna) and Semi-deciduous Mesophic Forest.Present1y, the Campus area is covered by gardens, allied to reforestation with native plants, "capoeira" and discontinuous patches of very disturbed primiti ve semideciduous Mesophic Forest.
Plants visited by Trimeria howardi were censured by sweeping the plants with flowers according to Sakagami & Matsumura (1967).
The flowering intensity was rated as: 1rare; 2-sorne; 3-quite, and 4-large number of plants with flowers.
The frequency of nests of T. howardi was estimated by a monthly census in five nesting sites.For this, quadrats were set as follows: analysis 40 quadrats of 1 600 m 2 (site A), 16 of 400 m 2 (site B), 66 of 400 m 2 (site C), 25 of 25 m 2 (site D) and 20 of 25 m 2 (site E).
Each nest found was coded, mapped, enu merated with a plastic plate and afterwards fi xed near the nest.In the initial phase of this in vestigation the old nests (commonly scars) we re marked with a nail c10se to the nests to re-o cord probable reactivations.
All new nests recorded after each census were considered new foundations and"active nests" were those with adults (males or fema les).In this species, the adults lack diapause in all periods of the year.
The seasonal variations of the stages of development of the immature presents in the nests, as well as " of the flight of the wasps we re registered through daily observations and periodic excavations of the nests.
To verify if there were differences among the proportions of founded nests according to the rainfall (or temperature) and the year under consideration was calculated: X 2 , Pearson's con tingency coefficient (C), and Goodman's test.
Months c1assification was based on monthly mean temperatures, and in the Index of effectiveness of the rainfall.
The monthly activity index was calcula ted by dividing daily flights frequency, to the frequency of monthly observations.Observa tions were organized fortnightly (during win ter) and weekly (others seasons).Every daily observation period involve-d twelve hours (6:00 AM to 18:00 PM).
The area maps containning the nesting si tes, and nests location can be seen in Mechi (1981), and Mechi (in press).
The specimens of the T. howardi was de posited by Soichi F. Sakagami at Hyogo Uni versity Agriculture Museum of Kobe Univer sity, Japan.In general, a relatively large number of active nests, foundations, immature in seve ral developmental stages, and high flight ac tivity were observed during the spring, sum mer and autumn.Conversely, low frequency or absence of foundations, and flight activity was verified in the winter (Figs. 2 to 8).At winter beginning, only immatures in the pre pupa stage, were found in cell.At the end of this period, retaking on the development of the immature, starts.Effectively, immature in the pre-pupa stage, pupa and imago appear (Fig. 8).

Visited
Trying to verify relationships between pe riod of abundance of foundations and environ mental factors, statistical tests, were used.The analysis of the association test between year and rainfall in relation to founded nests, showed that association exists between year, rainfall and fre quency of founded nests (X 2 calculated = 385.72X 2 critical = 15.51), but the association degree was little pronounced (C = 0.36).On the other hand, analysis of the associa lÍon between year and classified temperature, in relation to founded nests, it showed that as sociation exists between year, temperature, and founded nests (X2 calculated = 178.0852X2 cri tical = 9.49), but the association degree was liUle pronounced (e = 0.2209).

-º
The results of the contrast among propor tions of founded nests, in relationship to rainfall, inside each of the multinomials (year), were as proceed: For 1976, the proportion of founded nests follows the following order: Very hu mid>arid>humid.In 1977, humid>very humid> arid.For the very humid 1978 year, >(humid =arid).In relation to 1979, arid«very hu mid=humid).For 1980, very humid>humid arid.
The test of the contrast among proportions of foundations in relatíon to classified tempera ture, inside each multinomials (year) presented the following results: For 1976, the proportion of founded nests was higher in the periods clas sified as tropical.In 1977, the proportions did not differ from tropical to sub-tropical.In the tropical year of 1978, > sub-tropical.Whereas, in 1978, sub-tropical>tropical. Finally, in 1989, tropical >sub-tropicaL DISCUSSION Female palIen wasps, like all non-parasi tic bees, visit f10wer to collect pollen and nec tar to provision their nests similarly to most nonmasarinae aculeate wasps and bees, both maJe and female masarine collect nectar for their own nourishment (Gess 1996).
Trimeria howardi, in the Masarinae subfa mily, is a species with narrow polylecty, be cause it collected pollen in only two famílies of plants and four species.
Analysing seasonal variations in the abun dance of foundations, in f1ight activity and in immatures stages in nests (Figs. 2 to 8), it is sugested that the annual cyele of activíty of T howardi consists of two active phases, separa ted by an inactive phase.The active phase oc curs from January to June-July, and from Oc tober to December, as wel1; while the inactive phase appears in August-September.
In the active phases, nest foundations oc cur allied to flight activity and immatures' pre sence according to various developmented sta ges.ConverseJy, in the inactive phase were no ticed absence of foundations and of f1ight acti vity (Figs. 2 to 8) were the rule.In general, ac tive nests collected in the winter contained se nile females or males, and the new nests con tained only a shaft (Mechi, 1981).
The results of the tests among the propor tions of founded nests, in reJatíon to rainfall and temperature, inside each year (multino míals), showed that larger proportions of foun ded nests happened in the periods elassified as very humid, humid and tropical.
In the periods with absence of nests, the humidity of the soil (from surface 20 cm deep) varied from 13.7 % to 17.10 % .In the other periods, the humidity in the same soil layer, oscillated from 12.59 % to 51.09 %.
After Figs. 2 t07, it can be stated that cy ele of activity of T howardi is synchronous.Studies carried out in the Ribeirao Preto re gion, in the same period, involving Polybia paulista Ihering (Sim5es & Mechi 1983), Po listes versicolor versicolor (Olivier) (Gobbi & Zucchi 1980) and Spilomena sp.(Carvalho & Zucchi 1989) showed asynchronous nest's foundation, that is, foundations occur all year round independently of environmental condi tions.Decreased foundations abundance in driest periods, were observed in those groups.
The presence of only pre-pupa immatures in the period of June-July in T howardi, sug gests interruption in the development (diapau se).Similar fact occurs in the other Masarinae.In the group, the overwintering stage is the last-instar larva, which after it has finished fee ding spin a cocoon and then enters a resting prepupal phase (Gess 1996).
Mechi (198 1) evidences diapause occu rrence, in the study of nests' cicle in T howar di.Nests founded in the period ranging from october to may presented shorter ontogenetical period, as immatures' emergence accurred after 4-16 days when the founder female had left the nest.Moreover, nests founded at the end of that period presented an enlarged ontogenesis.In deed, in sorne nests emergence of the immatu re occurred from 87-132 days after the founder had left the nest.So, in this species, the life cy cle is bivoltine (Zucchi et al 1976) or multivol tine.Diapause, however is facultative, because it happens in only a gene ratio n during the year.
The life cycle polI en wasps in general, ap pears to be univoltine.The fligth period in winter-rainfall areas was spring to early sum mer, and in areas receiving spring and autumn or summer rainfall it is early summer lo Jate summer (Gess 1996).
Several factors have been suggested as responsible for the diapause: Temperature, hu midity, photoperiod, quality and quantity of food, physiology of the female, competition for food, etc.
In temperate regions diapause is generally concemed with survival during cold winters when normal growth is not possible.In the tro pics it may facilitates survival during the dry season, which is characterísed by a lack of moisture and food (Chapman 1988).
It is important to point out, that during the period when diapause happened, the values of the relative humidity and of the temperature were lower (even so, inside of the limÍts of the development of insects), the rain-fall was 10wers (Figs. 2 to 6), food was scarce (Fig. 1), and soil dry (13.77% to 17.10%).
Besides, flutuation in the abundance of acti ve nests and of foundations, along the year, the se factors also varied from one year to another.
Several factors can be pointed as responsi ble for these variations: Abundance of nests in the precedent period, populational density of each nest and its composition, predators and pa rasites attacks, anthropic action, dispersion ra tes, environmental factors, etc.
In relation to the elimatic factors, as socia tion tests among year, temperature, and rainfall in relation to foundations, by means of Pear sonlts Contingency Coefficient showed that an association exists among these variables, but association degree among them is not very stronger.The fall in the number of foundations per year, results from interactions between biotic and abiotic factors.
Ecologically it is difficult to differentiate effect-cause, and many causes that seemingly are natural consequences of the population cy eles, can be just collateral effects of the cyclic changes of the population (Pianka 1974).
High frequency of nests that are abando ned without leaving descendants (107 foun dations in a sample of 116), a number relati vely big of nests (26 nests and 50 cells in a sample of 119) with cell wholesales for para sites and predators, and great number of nests with offspring that were only constituted for males, can be sorne indicators of the varia tions.In this species, every females is a po tential founder.Absence of T. howardi in this campus, where was abundant before, is important.Alt hough causes of such drastic changes are unk nown, removal of host plants (to plant other garden species) and excessive grass trimming in the main nesting spots may have played a role.
Monthly variatíon in lhe number of aclÍve nests and of foundations, relaled lo environmenlal factors (temperalure, relatíve humidity and rainfall), in the period from april lo december 1976. .FEB. MAR.APR.NOV DEC.

Fig. 3 .
Fig. 3. Monthly variation in the number of active nests and of foundations related lo environmental factors (temperature, relalÍve humidity and rainfall), in 1977.

Fig. 5 .Fig. 6 .Fig. 7 .
Fig.5.MonthIy variation in the number of active nests and of foundations related to environmental factors (temperature, re lative humidity and rainfall), in 1979.