Respecto de los derechos comunicativos

4.4.1 Microclimate, physiology and activity

Microclimatic parameters, and temperature especially, through their influence on physiology w ill determine activity patterns of insects through time. The power available from muscles w ill be an important determinant in the behavioural options open to an insect. Terrestrial locomotion is possible when muscles can only produce limited power, though the resulting movement may be relatively

slow. For flight, however, there is a threshold for power which is necessary for an individual to remain airborne.

Muscle performance of insects is strongly temperature-dependent (Josephson 1981) and the power generated during endothermie warm-up of bees (Stone 1993) and during flight is proportional to thoracic temperature. The ability of a particular individual to fly w ill then depend upon its body temperature (determined by size and ambient conditions, section 4.5.4) and so the proportion of a given population of insects that can fly is a function of the proportion that have thoracic temperatures above the threshold requirement.

The influence of various microclimatic factors upon insect activity has been widely studied (see section 1.1). Several authors (e.g. Juillet 1964; Burrill & Dietz 1981; Schone & Tengo 1992) have measured the variation in the abundance of flying insects as a direct function of Tg and L (and sometimes RH and Wv); and others have additionally looked at the importance of the availability of floral rewards in these relationships (e.g. W illmer 1983; Stone 1994). The abundance of H . rubicundus females flying at the nest-site is presumed to be a good indicator of the level of foraging taking place at Invergowrie.

4.4.2 Abundance changes on a single day

The changes in relative abundance and microclimatic parameters through time on a typical warm and clear day at Invergowrie are given in Figure 4.12. Bee activity commences at about 08:30 and has completely finished by 16:30, and this is one of the longest recorded activity periods for H. rubicundus at this site. On cooler days activity started later and finished earlier with overall abundances being much lower; and when Tg failed to reach 14 °C bees never left their burrows. Once females begin foraging their numbers increase steadily until the middle of the day.

after which they decrease into the late afternoon (Figure 4.12a), and this unimodal pattern is unchanged throughout the season (subject to changes in ambient conditions). In the absence of microclimatic constraints female abundance coincides well with the peak pollen availability of B. napus and other floral resources (see section 3.5.1). Male abundance is also unimodal and closely follows the levels of female abundance (paired t-test: T = 0.03, d.f. = 28, p = 0.970); presumably this is a result of males coinciding their activity with females as they seek mating opportunities.

The variations in Tg, Tg, Tn and L through the day (Figure 4.12b and c) were discussed in section 4.3.1. RH w ill tend to be negatively related to Tg and L, and so w ill decrease through the day as the ground warms up (Figure 4.12d). The increase later in the day is probably a result of a drop in solar radiation added to the effect of moister air being brought in as Wv picks up after midday (4.12e). Activity dropped to virtually nothing around 30 minutes before the onset of rain and remained at this level even when other microclimatic conditions were suitable for flight. H . rubicundus females must be sensitive to some cue(s) in the environment (such as RH and/or pressure) that allows them to respond in anticipation to oncoming unfavourable weather (far more effective than local Meteorological office predictions).

As can be seen from Figure 4.13a, relative abundances for both sexes are predicted well by a quadratic function of Tg (females: y = -0.262x2 = 11.835x - 121.422, r2 = 0.717, d.f. = 13 and males: y = -0.237x2 = 10.240x - 96.948, r2 = 0.827, d.f. = 13) and neither deviated significantly from the model (p = 0.100 and p = 0.916 respectively). However L was not such a good predictor of abundance; and only when L was less than « 650 W m"2 was there any sort of proportional relationship between the two variables (Figure 4.13b).

4.4.3 Abundance changes on combined days through the season

In order for any model of abundance to be of more general use, it needs to combine data for a series of days across the season with different weather conditions. If abundance is plotted against Tg for several days (Figure 4.14a), the relationship is almost identical to that for a single day (females: y = -0.154x2 = 6.899X - 67.600, r2 = 0.224, d.f. = 47 and males: y = -0.225x2 = 9.412x - 85.428, r2 = 0.146, d.f. = 47) and neither deviated significantly from the quadratic model (p = 0.129 and p = 0.051 respectively). For both sexes, the amount of variation in abundance explained by Tg, for a single day, was more than 70 % but less than 25 % over a series of days. This decrease may be due to variation in other microclimatic parameters, such as cloud cover and wind, that remained relatively constant on 14.8.92, but may have caused rapid short term changes in abundance on other days.

Abundance varies proportionally with L (Figure 4.14b), although this is not initially apparent from the data due to the same sort of 'noise' described for Tg above. Consequently the amount of variation in abundance accounted for by variation in L is not high, especially for males (females: y = 0.013x - 1.426, r2 = 0.255, d.f. = 48, p < 0.001 and males: y = 0.012x + 1.824, r2 = 0.078, d.f. = 48, p = 0.024). High abundances are associated with low humidity levels (Figure 4.15a), owing to the approximately inverse relationship between RH and L. A Wv greater than 3m s"^ was never recorded when bees were active at Invergowrie; for the majority of the time Wv was less than 1.5 m s"^ during bee activity (Figure 4.15b). There is no observable direct association between abundance and Wv, but W v w ill have an indirect effect through Tg (see section 4.3.3).

Putting together the influences of light and temperature on abundance, a model can be drawn up. Although abundance and Tg have a curvilinear relationship, the

majority of abundance measurements were made at TgS less than 23 °C (which are the normal conditions at Invergowrie). For this part of the curve (15 to 23 °C) the increase in abundance is roughly proportional to Tg. It may therefore be acceptable to enter this variable, along with L and t into a multiple regression as predictors for the response of abundance (Table 4.5). The results show a clear difference between the sexes. The abundance of males is clearly a positive function of L (p = 0.001) and Tg (p = 0.012), but independent of f (p = 0.282); and 75 % of the variation in abundance is explained by the first two variables. In contrast, female abundance is simply a function of L (p < 0.001) and not Tg (p = 0.472) or f (p = 0.623); and only 20.1 % of abundance variation is accounted for by L. Clearly the model would be improved by entering Tg as a quadratic term. The fact that for males Tg comes out as significant w ill be largely due to the influence of the seven relatively high abundances recorded below 23 °C, which w ill increase the linearity of the association at lower TgS.

The overall relationship between abundance, L and Tg is presented in the contour graphs in Figure 4.16. The linear change in abundance above 700 W m-2 is clear, as is the peak of the curvilinear Tg term at 21 °C. The similarity in the overall contour shape for males and females is striking. It is not, however, possible to determine whether this is due to both sexes' flight capabilities responding identically to microclimate characteristics, or whether male abundance is determined by female abundance per se. The former would seem unlikely as there is a large size difference between the sexes which inevitably results in very different body temperatures under a given set of microclimatic conditions (see section 4.5.4). As an approximate predictor of female abundance Tso can be used (Figure 4.17). As only 14.9 % of abundance variation is accounted for by Tgo variation, others factors must also be important. The minimum temperature necessary for flight is indicated by a Tso of « 20 °C (see section 4.5.4E). There are few records of high

abundances below this critical Tso, and these are accounted for by observations made on a single day (22.7.92) when females had been prevented from foraging on the previous two days by heavy rain. The following clear day then presented the first opportunity with Tgo > 20 °C, which presumably led to increased pressure to forage and consequently high abundances. The equivalent Tg for a 20 °C Tgo is 16 °C on a clear day (4.10a)., and this Tgo is represented by the area of 15 -18 °C Tg and 300 - 700 W m-2 on Figure 4.11a.

If figures 4.11b and 4.16b are compared visually it can seen that the peak values are in different positions: maximum Tg for Tso and maximum L for female abundance. Overall, L alone is a better predictor of female abundance for this data set (Figure 4.14b). However the data collected here do not cover the entire range of microclimatic conditions that occur at Invergowrie; most notably Tg values of 22 to 28 °C which are associated with L values greater than 800 W m-2. It might be predicted that this area of the graphs would encompass peak Tgo values and also peak female abundance. If this were the case then Tgo and abundance would have a much stronger relationship, and consequently Tgo would be of a much greater predictive value.