Can climate shape flight activity patterns of Plebeia remota ( Hymenoptera , Apidae ) ?

Flight activity of foragers of four colonies of Plebeia remota (Holmberg, 1903) was registered from December 1998 to December 1999, using an automated system (photocells and PLC system). The colonies originated from two different regions: Cunha, state of São Paulo, and Prudentópolis, state of Paraná, Brazil. Flight activity was influenced by different climatic factors in each season. In the summer, the intensity of the correlations between flight activity and climatic factors was smaller than in the other seasons. During the autumn and winter, solar radiation was the factor that most influenced flight activity, while in the spring, this activity was influenced mainly by temperature. Except in the summer, the various climatic factors similarly influenced flight activity of all of the colonies. Flight activity was not affected by geographic origin of the colonies. Information concerning seasonal differences in flight activity of P. remota will be useful for prediction of geographic distribution scenarios under climatic changes.

Currently, as climate changes have already become evident and studies on their effects focus on plant-pollinator interactions (Hegland et al., 2009), the geographic distribution of these plants/pollinators (Opdam & WascHer, 2004), food production, and the different future scenarios based on predicted climate changes, we find that we know very little about the effects of abiotic conditions on flight activity of native stingless bees.Relatively, little is still known how living beings adapt to temperature variation, whether through gene expression, control of growth and development, or changes in the organism as a whole (clarke, 2003).The use of bees as ecosystem services providers has been the focus of attention for many crops (for a review, see klein et al., 2007).
Study of the influence of abiotic factors on flight activity consists of the counting of bees on flowers or registration of the number of bees that leave or enter colonies, with or without visible collected material, during a certain time interval.nOgueira-netO et al. (1959) were pioneers in this type of study in stingless bees; they recorded how climate variables affected the number of foragers on coffee flowers.
Studies of stingless bees as pollinators are just beginning (for a review, see slaa et al., 2006).It is known that stingless bees use a higher diversity of food sources than do honey bees and have a greater diversity of foraging behaviors, pollinating both native and cultivated plants (ramalHO et al., 1990;Wilms & WiecHers, 1997;kleinert et al., 2009).
As mentioned by micHener (1974), generally factors that could affect flight activity are analyzed independently, without determining if there are relations between these factors; univariate analyses are used, instead of multivariate analyses, which would be more informative.However, it is rare that rainfall distribution follows a normal distribution, which means that nonparametric analyses normally need to be implemented.Another difficulty in analysis is that flight direction and flight times have a circular distribution.This is why we previously analyzed the effects of rainfall and wind on Plebeia remota (Holmberg, 1903) flight in a totally different manner and independently (HiláriO et al., 2007a,b).
We analyzed the influence of climate factors and their seasonal variation on the flight activity of colonies of the stingless bee, P. remota, which originated from two regions with different climates, using automatic counting devices.Molecular biology data indicates that the bees of this species from these two regions constitute independent evolutionary units (FranciscO & arias, 2010).Thus, we also want to know if there are differences in foraging behavior of bees from these two regions.

MATERIAL AND METHODS
There are 38 species of Plebeia Schwarz, 1938, distributed from Chihuahua, Mexico, to Montevideo, Uruguay.In Brazil, P. remota is found in the states of Espírito Santo, Minas Gerais, São Paulo, Paraná, Santa Catarina and Rio Grande do Sul (camargO & pedrO, 2007).
The flight activity of the bees that entered and left the colonies was registered with photocells installed in plastic tubes; the bees had access to the fields outside of the laboratory through these tubes.The photocells were connected to a computer through a programmable logic controller (PLC).Hourly data on flight activity (sum of entries and exits) during 281 days was registered from December 23, 1998to December 22, 1999; this data was grouped by season of the year.Sunrise and sunset times were obtained from the Naval Observatory of the United States of America (http://aa.usno.navy.mil/).This allowed us to delimit the time intervals during which foraging by P. remota was possible.Flight activity during 3,829 hours (Summer,935 h;Fall,956 (zar, 1999) to examine the relationship between the monthly means of some of these climatic factors for 1999 and the respective historical month means from 1961-1990 (Instituto Nacional de Meteorologia).We also made a correlation analysis to measure how much each climate variable affected flight activity during the different seasons.The correlation coefficients were then ranked for each season, from the highest to the lowest value."Top-down" correlation analysis (zar, 1999) was used to determine if the climate factors had different degrees of influence on flight activity within each season.We were interested in determining if the colonies that had come from Prudentópolis (Pr1 and Pr2) could be separated from those from Cunha (Pr3 and Pr4), presuming that their flight activity was differentially affected by abiotic factors.The same type of analysis was used to determine if the effect of the different climate factors on flight activity varied with season.A Bonferroni adjustment for the level of significance was applied to control for Type I errors when repeated tests are used (critical p = 0.0083).

RESULTS
Except for relative humidity (r s = 0.489; p = 0.104), there were significant positive correlations between the monthly means of 1999 and the historic means for all of the climate parameters, including temperature (r s = 0.951; p = 0.001), barometric pressure (r s = 0.888, p = 0.003) and rainfall (r s = 0.720; p = 0.016).Thus, we assumed that the climatic factors analyzed in our study were similar to the ones of historic means.It means that, for instance, the summer in our study was similar to the typical summer.
Though the flight activity varied among seasons for all colonies, flight activity was positively correlated with solar radiation and negatively correlated with rainfall (Tab.I).Significant correlations were found for all colonies and during all seasons (except colony Pr2, during summer) between flight activity and temperature (positive) and between flight activity and relative humidity (negative, Tab.I).These correlations were highest during spring for all colonies.During winter and spring seasons, flight activity and barometric pressure were significantly and negatively correlated.During Can climate shape flight activity patterns of Plebeia remota... (Holmberg, 1903) fall season, significant positive correlations were found between flight activity and wind speed for all colonies (Tab.I).

Tab. II. "Top-Down" (r T ) correlation coefficients between pairs of colonies of Plebeia remota
The degree of correlation between climate factors and flight activity was generally weakest during summer.The climate factors that most influenced flight activity varied among colonies.Solar radiation influenced flight activity in the colonies Pr1 and Pr3 and rainfall influenced the colonies Pr2 and Pr4.During fall and winter, solar radiation was the factor that most affected flight activity for all colonies, followed by relative humidity and temperature.During spring, flight activity was mainly affected by temperature and by solar radiation, though relative humidity and barometric pressure also affected colony flight activity.
When pairs of colonies were compared during the different seasons (Tab.II), we verified that the various climate factors equally affected the flight activity of all of the P. remota colonies.However, during summer, the factors that most influenced flight activity of colonies Pr1 and Pr3 were different from those that most affected the flight activity of colonies Pr2 and Pr4.
When we compared seasons for the different colonies (Tab.III), we found that the same climate factors affected P. remota flight activity for all colonies in fall and winter.These same climate factors also affected the flight activity of colonies Pr1 and Pr3 during summer.
The variation in flight activity patterns of the four colonies of P. remota as affected by time of day, temperature, relative humidity, solar radiation and barometric pressure differed with season (Figs 1-4).During summer (Fig. 1) and spring (Fig. 4), two plateaus were often seen, with a relative decline in flight activity approximately 12:00 -13:00 h (Fig. 1) and 11:00 h (Fig. 4), among colonies Pr2, Pr3 and Pr4.The decrease in flight activity appears to be related to an interaction between high degrees of solar radiation and high temperatures.During the other seasons (fall, Fig. 2; winter, Fig. 3), there was a plateau in flight activity in all of the colonies.Standard deviations were not included in these figures because their inclusion altered the scales, making comparisons more difficult.
Correlations between flight activity and climate factors can also be visualized in these figures.During summer (Fig. 1), the influence of climate factors on flight activity was weak, principally for colonies Pr2 and Pr4.In colony Pr1, the flight activity pattern was more closely related to the solar radiation and temperature cycles and was negatively correlated with relative humidity.Colony Pr3 had somewhat different flight activity patterns, though the relation with climate factors was similar to that of colony Pr1.During the other seasons, it was clear that flight activity of all of the colonies mainly accompanied solar radiation (Figs 2, 3) and temperature (Fig. 4).Generally speaking (Figs 1 -4), the greatest flight activity of P. remota was correlated with strong solar radiation, high temperature, low relative humidity and low barometric pressure.
Tab. III.Top-Down" (r T ) correlation coefficients between pairs of seasons, for all of the colonies of Plebeia remota (Holmberg, 1903)
In P. remota, we observed that all of the climatic factors that we analyzed influenced flight activity in some way.There was a slight tendency towards increased flight activity in all colonies with increased wind speed in the fall, while it decreased with increased rainfall during all seasons.Flight activity increased with decreasing barometric pressure, during winter and spring, though this tendency was stronger during spring.Barometric pressure has a direct effect on the flight dynamics of insect wings.Increased barometric pressure increases air viscosity, making flight more difficult (spedding, 1992).Generally speaking, solar radiation, temperature and relative humidity, in this order, were the climate factors that most affected the flight activity of P. remota.
kleinert-giOvannini & imperatriz-FOnseca (1986) noted that the effect of temperature on flight activity of Melipona marginata obscurior Moure, 1971 was maintained during two different periods.However, relative humidity had an opposite effect: between January-July 1982, flight activity was negatively correlated with relative humidity, while during October 1982-January 1983, the highest peak of flight activity occurred at higher humidities.The same tendency was found for flight activity of Plebeia sp.nov.-early referred as Plebeia pugnax Moure in litt.(HiláriO & imperatriz-FOnseca, 2002).Consequently, despite the evident correlation between temperature and relative humidity (kleinert-giOvannini, 1982), we analyzed these factors separately.The greatest flight activity of  colonies of P. remota was associated with periods when there was low relative humidity.Nevertheless, during summer the ideal relative humidity range for promoting flight activity was similar to that reported by imperatriz-FOnseca et al. (1985), in a study made in the spring and summer of P. remota colonies from Rio d'Oeste, state of Santa Catarina (60 -84%).
Temperature is one of the climate factors that most influences bee flight activity.Low temperatures limit P. remota flight, as can be seen in fall and winter; a reduction in temperature from 6:00 to 7:00 h, reduced flight activity.During the coldest seasons, the flight activity was more compact (concentrated in few hours) than during the other seasons.A similar pattern was found by Heard & Hendrickz (1993) for the flight activity of T. carbonaria, during a typical winter day.The ideal temperature range for flight activity of P. remota was 21 -32°C during all seasons, similar to what was reported by imperatriz-FOnseca et al. (1985), who indicated 22 -29°C.knOll & imperatriz-FOnseca (1993) noted that P. remota foragers were not found at flowers during fall and the beginning of winter.They postulated that because of the low temperatures, there were fewer flowers and bees foraged only a few hours a day.Also, during these seasons, P. remota stops building brood cells, a process denominated reproductive diapause (riBeirO et al., 2003).Nevertheless, the flight activity of P. remota was not interrupted during this reproductive diapause.nunessilva et al. (2010) observed that the nectar collection in P. remota was similar in the reproductive phase and in the reproductive diapause.On the other hand, the pollen collection in P. remota was lower in the diapause than in the reproductive phase.A similar result was found by pick & BlOcHtein (2002a) with P. saiqui.
When we analyze the raw data (not the means given in figures), we find that solar radiation above 600 W/m 2 decreased flight activity of P. remota colonies during all seasons.However, up to 600 W/m 2 , this was the climate factor that most stimulated flight activity.This was also found for the flight activity of T. carbonaria during a typical summer day (Heard & Hendrickz, 1993).Plebeia remota foragers should avoid high solar radiation levels in order to reduce water loss.Some strategies can be used to avoid or reduce this loss; water can be taken directly from the air, when relative humidity increases (gilles & delpire, 1997), or from dilute nectar collected from flowers (Willmer et al., 2000).This extra water can be used for evaporative cooling, as a consequence of increased ventilation (prange, 1996), or by exudation (rOuBik, 1989).However, evaporative cooling seems only to be possible for bee with body mass above 0.1g (prange, 1996), while P. remota foragers have a much smaller body mass, approximately 8.4 mg (HiláriO & imperatriz-FOnseca, 2002).
During summer, colonies Pr1 and Pr3 were different from colonies Pr2 and Pr4, in terms of the climatic factors that influence their flight activity.During the other seasons, the flight activity was influenced in a similar way by the climate factors.Consequently, based on the analysis of the effects of the climatic factors on flight activity, it was not possible to distinguish the colonies from Cunha, SP from those from Prudentópolis, PR. kleinert-giOvannini & imperatriz-FOnseca (1986) found that peak flight activity of Melipona marginata marginata Lepeletier, 1836 and M. m. obscurior occurred at different temperatures, which coincided with the type of climate of the distinct regions inhabited by these subspecies.Nowadays, M. m. obscurior have a new status as Melipona obscurior Moure, 1971.
How much climate factors influenced flight activity of colonies of P. remota varied among seasons.The seasonal cycles mold the phenology of insects through extreme temperatures, resource availability and competition, affecting population structure, mating, growth and development, among other factors (underWOOd et al., 1997).Even in subtropical regions, seasonal climate changes influence the foraging behavior of bees.During the coldest periods, environmental factors probably limit the foragers physiologically, as well as reducing the availability of floral resources.During summer, these conditions are changed, and flowering is highly related to sunshine (FOWler, 1979).During part of fall and winter, the P. remota colonies were in reproductive diapause and flight activity was mainly influenced by solar radiation.During spring, the factor that most affected flight activity was temperature.pick & BlOcHtein (2002b) compared the flight activity of P. saiqui during two colony phases, during and after reproductive diapause.They indicated that solar radiation and temperature influenced flight activity during both periods, though after reproductive diapause, relative humidity also contributed.
During spring, and principally during summer, two plateaus/peaks in flight activity of P. remota were common.rOuBik (1989) attributed these peaks to the gathering of different resources, these being pollen in the morning and nectar in the afternoon.Similarly, the foraging pattern of Melipona scutellaris Latreille, 1811 also presented two peaks during the months of October, November and December in 2003 (pierrOt & scHlindWein, 2003).The flight activity plateaus found for P. remota could be due to a greater availability of resources at different times, in the morning and in the afternoon, independent of the resource that was collected.pierrOt & scHlindWein (2003) found that M. scutellaris foragers went from nectar to pollen foraging immediately after "big bang" flowering.nagamitsu & inOue (2002) also observed a similar influence of events of this type on foraging activity of four species of Trigona Jurine, 1807.
Nevertheless, between the two plateaus found in the foraging activity of P. remota, flight activity decreased around 12:00 -13:00h.This decrease could be due to a physiological limitation of P. remota, since during this interval there is an interaction between the highest intensities of solar radiation and the highest temperatures.Another possible cause is that during the hottest months, there are more bees competing for the resources, especially during this time interval.stOne et al. (1999) observed that Anthophora pauperata Walker, 1871, a desert species of solitary bee, has a bimodal foraging pattern, linked to pollen collection in the morning and in the afternoon.There is decreased activity around 12:00 h and the foraging peaks are separated by a two to three hour period; this decrease was not attributed to temperature, since the temperature stayed high during the intervals when foraging was at its peak.
In terms of competition for resources, in the Atlantic forest, P. remota has low niche overlap with Apis Linnaeus, 1758, but it competes with other stingless bee species (Wilms et al., 1996).Except for Plebeia sp.nov.(HiláriO et al., 2001), bees of the genus Plebeia have been described as unaggressive and they are evasive when confronted with other bee species (JOHnsOn & HuBBel, 1974).Consequently, their foragers must collect resources from plants that are not visited by other species, and thus a larger number of plant species are exploited by P. remota during spring, and especially during summer (ramalHO et al., 1985).Competition, the quantity and quality of resources, and especially, the need for these resources, can lead to foraging under adverse climatic conditions.In this context, during the rainy summer of 1998/1999, P. remota and Melipona bicolor bicolor Lepeletier, 1836, foragers were observed returning to their colonies with pollen in their corbiculae.The potential of P. remota as a pollinator of commercial crops is still unknown, though Plebeia spp.are known to intensively visit Coffea arabica L. (nOgueira-netO et al., 1959) and Syzygium jambos (L.) Alston (knOll & imperatriz-FOnseca, 1993).Studies made of pollen grains in pollen and honey pots of P. remota have shown a prevalence of Arecaceae and Moraceae.Visits are also made to Anacardiaceae (knOll & imperatriz-FOnseca, 1993), Cunoniaceae, Lauraceae (Wilms et al., 1996), Araliaceae, Asteraceae, Burseraceae, Clethraceae, Euphorbiaceae and Myrtaceae (ramalHO, 2004).
Understanding the influence of abiotic factors on the flight activity of stingless bees is essential for developing adequate management programs for their use in the pollination of crops, both in greenhouses and in fields.Another issue involves the adaptation of bees to the changing seasons, how they tolerate some weather events and how they forage in these conditions.All this knowledge will allow us to predict the geographical distribution for host plants and stingless bees in the face of climate change.
, for the different seasons.Probability values (p) in bold indicate significance after Bonferroni adjustment.Significant correlations indicate that the abiotic factors had a similar influence on the flight activity of the two colonies.Tab.I.Spearman correlation coefficients (r s ) between flight activity of colonies Pr1 to Pr4 of Plebeia remota(Holmberg, 1903)and climatic factors (air temperature, relative humidity, barometric pressure, solar radiation, rainfall, and wind speed), during the different seasons.Probability (p) in bold indicates a significant correlation after Bonferroni adjustment.
. A probability value (p) in bold indicates significance after Bonferroni adjustment.Significant values indicate that the abiotic factors had similar degrees of influence on flight activity of individual colonies during the different seasons.