Food collection behavior of Apis mellifera and Tetragonisca angustula bees in Brassica napus L. in response to different environmental covariates

The objective was to evaluate the behavior of Apis mellifera and Tetragonisca angustula bees in pollination tests in Brassica napus at different times of the day, temperature and relative humidity. The experimental design was completely with eight treatments and two pollination tests, repeated in five randomized blocks during seven days of observations for two years, totaling 560 records. During the visits, the following parameters were recorded: collected resources, nectar collection site, time spent on flower, number of flowers visited in one minute, pollen load in the pollen basket and bee contact with anthers and stigma. Data were analyzed using generalized linear models. The number of Africanized and T. angustula bees collecting nectar increased with the passage of time throughout the day and with the decrease of relative humidity. The same was observed for nectar collection in both nectaries. The proportion of bees collecting pollen was higher in the morning hours, as well as when there was an increase in temperature for the two species of bees. Foraging behavior of A. mellifera in B. napus crop favored its pollination, indifferent of which collected floral resource as they came into contact with anthers and stigma. T. angustula bees performed pollination only during pollen collection. Pollination of B. napus was more effective in the warmer hours of the morning, when more of both species of bees carried out pollen collection. Due to its foraging behavior, A. mellifera had greater efficiency for pollination of B. napus; however, the association with T. angustula may potentiate the benefits generated for the crop by cross-pollination.

nutritional resources [8]. B. napus flowers have an average nectar production ranging 66 from 0.02 to 0.75 µL per flower [9] and produce on average 9 kg/ha of pollen per day 67 [10]. 68 To an efficient pollination process, it is necessary to the floral visitor to present 69 specific morphological and behavioral characteristics for each type of flower [1]. This is disappearance of the main pollinator [13]. In canola cultivation, species of Halictus sp. 76 [8], Bombus lapidarius [14], and Colletes lacunatus [11] bees were identified as 77 alternative pollinators. In Brazil, the stingless bee species Trigona spinipes was identified 78 with a potential pollinator for canola [15]. 79 Considering the importance of native bees in pollination of agricultural crops [16],  The pollination cages were made following the model proposed by [20], with a 2 118 x 2 mm mesh nylon screen, supported by ¾ inch PVC tubes, 4 m wide, 6 m long, and 2 119 m high at the top, totaling an area of 24 m² (Fig 1). They were assembled before flowering,   Deviance analysis was used to adjust the GLM from a model represented by 143 systematic portion: in which: T j is the effect of time (j = 1, 2, 3, and 4); 150 β 1 is the regression coefficient of g(µ) about X; 151 X ijk is the observation of the covariable "temperature" in each plot; 152 is the general mean temperature; X 153 β 2 is the regression coefficient of g (μ) about W; 154 W ijk is the observation of the covariable "relative humidity" in each plot; 155 is the general average relative humidity; and W 156 Ɛ ijk is the random error of the plots.

157
In the occurrence of a higher probability of significance for a given factor, it was  The block effect, pollination test, time, temperature, and relative humidity on the

283
For the proportion of bees collecting pollen, it was shown that the estimate was 284 higher with the increase of temperature and relative humidity and in the morning hours, 285 decreasing with the hours of the day for both species (Fig 6). It was observed that 286 estimation of bees collecting nectar and pollen increased with increasing relative 287 humidity, especially above 65%, for both A. mellifera and T. angustula (Fig 7). For both species the number of flowers visited in one minute was higher when the 320 relative humidity was lower and in the morning period, decreased with the passage of the 321 day (Fig 10). Estimates of the number of A. mellifera and T. angustula bees that came into 329 contact with anthers and stigma and contact with anthers, was higher as the relative 330 humidity increased (Fig 11 and 12, respectively). The estimated average peak of 331 Africanized honeybees that came in contact with anthers and stigma occurred when

381
It was observed that both species of bees preferred to collect more concentrated 382 nectar. Nectar has a higher sugar concentration as the relative humidity decreases [26], 383 that is, the low relative humidity provides greater evaporation of water from the nectar.

384
Possibly, bees prefer more concentrated nectar because to turn it into honey they spend

391
The nectar produced mainly by the nectaries that are most exposed tends to reach a 392 concentration in equilibrium with the relative humidity of the air; that is, high relative 393 humidity further dilutes the nectars so that, consequently, there is an increase in its   adjust their resources to ensure that pollinators move between as many flowers as possible 429 [1]. Considering that the highest number of bees collecting pollen was in the morning 430 period and that during this period there were more flowers being visited by both species 431 studied, these characteristics showed the pollinator potential of A. mellifera and T. 432 angustula for the canola crop.

433
With the increase of relative humidity, there was an increase in the number of A.