Mobility of honey bees
(Apidae, Apis mellifera L) during foraging in avocado orchards
Apidologie 29 (1998):209-219
G Ish-Am 1*,
D Eisikowitch 2
1
The Hebrew University of Jerusalem, Department of Horticulture, P O Box 12,
Rehovot 76100, Israel
2 University of Tel-Aviv,
Department of Botany, P O Box 39040, Tel Aviv 69978, Israel
Running title - Mobility and cross pollination of honey bees
* Corresponding author. The
Hebrew University of Jerusalem, Dept of Horticulture, P O Box 12, Rehovot
76100, Israel. Tel (972)-8-9481089, Fax (972)-8-9468263
Summary - The mobility of the honey bee (Apis mellifera L) during foraging has a great influence on the effectiveness of the bees as cross-pollinators. In this work, honey-bee mobility was measured in avocado orchards, between neighboring trees and up to a distance of 15 rows. The average number of bees crossing between adjacent rows in a 10-min period was linearly correlated to bee density, and the corresponding percentage increased with the increase in wind velocity, from 30% in a light wind (4 km/hr) to up to 65% in a strong wind (45 km/hr). The bees tended to travel upwind, and this tendency increased with increasing wind velocity. Consequently, under strong-wind conditions, up to 100% of the bees traveled to the adjacent upwind row in a 10-min period. The percentage of cross-pollinating bees decreased with increasing distance from the pollen source, following a hyperbolic curve, and reached 1%-2% of the bees at a distance of 10-15 rows.
Key words - honey bee / mobility / cross-pollination / avocado / wind velocity
INTRODUCTION
Pollinator mobility is essential mainly to plants that require, or are benefited by cross-pollination (i e, pollen transfer between plants of different cultivars within the same species). This need occurs in dioecious crop species (kiwi, some papaya cultivars), in self-incompatible monoecious or hermaphroditic species (almond, cherry), and in crops which are partially self-incompatible (some apple and most avocado cultivars) (McGregor, 1976). In some of these species, such as apple and almond, cross-pollination is performed by honey bees both directly, during the foraging flight, and indirectly, by means of pollen transfer through body contact of bees inside the hive (Free and Durrant, 1966; DeGrandi-Hoffman et al, 1984, 1992). However, no efficient avocado pollen-transfer occurs among the honey bees within hive, and avocado cross-pollination is totally dependent on bee mobility during foraging (Ish-Am and Eisikowitch, 1993; Ish-Am, 1994).
Honey-bee mobility during foraging is limited: a single bee usually forages within an area of no more than two or three neighboring trees (Clark, 1923; Butler et al, 1943; Bateman, 1947; Free, 1960b, Free and Spencer-Booth, 1964a, 1964b; Bergh, 1967; Davenport, 1986; Free, 1993). Nevertheless, foragers moving tens, or even hundreds of meters, have been observed. These bees stopped foraging before their loads of nectar or pollen were full, flew above the canopy, and landed elsewhere for continued foraging (Stout, 1933; Butler, 1945). This ‘orientation flight’ (Free, 1960a), which was later called ‘sampling behavior’ (Free, 1966a, 1966b) or ‘monitoring flight’ (Free, 1993), was assumed to be made by the ‘scout bees’, which are the only honey bees that perform long-distance cross-pollination during foraging (Free, 1993).
The purpose of the current research was to measure honey-bee mobility during foraging in an avocado orchard, between adjacent trees and in a larger range, and to correlate this mobility to bee density on the trees, to wind speed and direction, and to the distance from the pollen source.
MATERIALS AND METHODS
Observations and measurements were conducted in avocado plots in the western Galilee of Israel, using medium size trees (5-8 m high) of the cultivars ‘Fuerte’, ‘Ettinger’, ‘Hass’, ‘Nabal’, and ‘Reed’. They took place in the 1982-1984 seasons in an orchard near kibbutz Eilon, and in the 1989-1993 seasons in orchards of Eilon, Rosh-Hanikra and Kabri.
Honey-bee density (BPT = bees per tree) was assessed by counting the bees on a whole tree (five trees per cultivar), while circling it at a distance of about 1 m from the canopy for 1 min, using a manual counter (Free and Spencer-Booth, 1963; Mayer et al, 1986). Calibration counts were performed at the beginning of each day, until the discrepancy among the workers was reduced to less than 10%. Wind velocity (WindVel) and wind direction were estimated in an open field beside the orchard. Beaufort scale for field estimation of wind velocity was used, to be later converted to km/hr.
Honey-bee mobility among adjacent trees was measured on four trees in full bloom: two neighboring trees in a row and two trees next to them in the adjacent row. At the beginning of each measurement period, wind speed and direction, and bee density were recorded. Later, all the bees that crossed among the observation trees were counted, simultaneously in all directions, by an observer, who sat with his back to the sun, and used a multi-station counter with a marked station for each direction. Using these data, the average number of bees that had crossed in a 10-min period from one tree to the adjacent row (NoMob) was calculated. The percentage of bees which crossed between adjacent rows in 10 min (PrMob) was calculated as (Equation 1):
PrMob = (NoMob/BPT)*100
The ratio of
wind-direction preference of the bees (UWMP = upwind mobility
preference) was calculated for each measurement as
UWMP = No. of bees traveling upwind / No. of bees traveling downwind
Correlations of these variables with both bee density and wind velocity were examined.
The monitoring of honey-bee mobility in a range greater than one row was enabled due to the exclusive flowering phenology of avocado. During 4-6 hr in the morning, under normal temperature conditions (Ish-Am and Eisikowitch, 1992), type-A avocado cultivars bear only female-stage flowers, while type-B cultivars carry male flowers exclusively, with a reversal of roles in the afternoon (Stout, 1923). Therefore, bees that visit flowers of a cultivar in male-stage bloom ‘mark’ themselves by building pollen loads, which render them identifiable and subject to follow-up while crossing to a cultivar in female-stage bloom (Ish-Am, 1994). Measurements were taken in avocado plots with adjoining blocks of opposite flowering-type cultivars which bloom during the same season. Plots of ‘Hass’ contiguous to ‘Ettinger’ were selected in the orchards of Eilon, Rosh-Hanikra and Kabri, and a ‘Reed’ block bordering with a block of ‘Nabal’ was selected in Rosh-Hanikra orchard. Measurements were performed during the female bloom of type-A cultivars (‘Hass’ or ‘Reed’), and the pollen-releasing stage of type-B cultivars (‘Ettinger’ or ‘Nabal’ , respectively), which occur in the morning on warm days and at midday on cool days (Ish-Am and Eisikowitch, 1992). At the onset of each measurement, wind direction and velocity and bee density of each cultivar were recorded. Later, pollen-load carrying bees and nonloaded bees were simultaneously counted in several rows by several observers. Each observer walked along one row for 15-20 min and used two counters. Counts were carried out in the two rows of the male-blooming cultivar facing the female bloom, and in several rows of the female-blooming cultivar (No. 1, 2, 4, 5, 7, and 15 away from the male bloom). For each row we performed two to five counts, pooled them and calculated the average percentage of pollen-load carriers, which was marked:
PPM(0) = percentage of pollen-load carriers on male-bloom trees
PPF(d) = percentage of pollen-load carriers on female-bloom trees, in ‘d’ rows away from the male bloom
We assumed that all the pollen-load-carrying bees on the female blooms had arrived from the adjacent male bloom during foraging flight. Also, we assumed that the mobility of pollen-load carriers is similar to that of the nonloaded bees, i e that the proportion of the pollen-load carriers among the bees visiting the male bloom is similar to that among the cross-pollinating bees on the female bloom. Therefore, the percentage of cross-pollinating bees in row ‘d’ of the female bloom, which was designated PCrBee(d), was calculated as (Equation 3):
PCrBee(d) = (PPF(d) / PPM(0)) *100
Take-off and landing directions were observed on three trees in full bloom: two flowering branches (branches which bear inflorescences in each leaf node) per tree were marked, one facing an adjoining tree within the row, and the other facing a parallel tree in the adjacent row. At the beginning of each observation, wind direction and velocity and bee density were measured. Later, all bees visiting the marked branches in a 10-min period were recorded, noting their landing and take-off directions. For improved accuracy, bee-movement directions were recorded by the eight major and secondary compass points. The secondary point counts were later proportionally divided between the adjacent major points. Observations were made by one observer under light wind conditions (1 Beaufort=3.7 km/hr), and by three observers under mild wind conditions (4 Beaufort=24.1 km/hr). Each observer performed three observation periods of 10 min.
Data analysis was performed according to Sokal and Rohlf (1981). The statistical procedures are listed further below.
RESULTS
Mobility between adjacent rows
The number of bees that crossed to the adjacent row in 10 min (NoMob) increased significantly with the increase in BPT, which explained 50% (p<0.0001) of NoMob variance (Fig 1), and increased only slightly (not significantly) with the increase in wind speed (Fig 2). NoMob related to BPT and to WindVel by the function (Equation 4, multivariable regression):
NoMob = 0.418*BPT + 0.277*WindVel - 4.02
which explained 51.1% (p=0.0001) of NoMob variability.
The percentage of bees crossing to an adjacent row in 10 min (PrMob) was found to be hardly influenced by bee density, and was varied around an average of 41% of the bees (Fig 1). However, it rose with the increase of wind velocity (Fig 2), from 30% during a light wind to 60%-70% during a strong wind of 44.5 km/hr (6 Beaufort, the strongest wind during which we observed bee activity), and wind velocity explained 13% (p=0.042) of PrMob variance. PrMob related to BPT and to WindVel by the function (Equation 5, multivariable regression):
PrMob = 0.0476*BPT + 0.897*WindVel + 26.2
which explained 13.5% (p=0.073) of PrMob variability. A parametric transformation of PrMob [TPrMob = arcsin(PrMob/100)½] (Sokal and Rohlf, 1981) related similarly to these variables (r²=0.13, p=0.081).
Preference for the upwind direction
In all measurements of honey-bee mobility between adjacent trees, more bees traveled upwind than downwind (Fig 3). The preference for the upwind direction (UWMP) was found to increase with the increase in WindVel, which explained 43% (p=0.0007) of UWMP variability (Fig 3).
Mobility to a range greater than one row
The percentage of pollen-load carriers was maximal during pollen shedding, and dropped gradually thereafter. It reached 40-60% on the male-blooming cultivars, and was much lower on the female bloom: during pollen shedding we recorded up to 8% pollen-load carriers in rows adjacent to the male bloom, but in more remote rows, and at a later period, only 0-5% of them were observed (Fig 4). The best-fit function of regression of the averages of percentage of cross-pollinating bees on female bloom vs distance (‘d’ in rows) from the male bloom was (Equation 6):
PCrBee = 12.3/d + 0.323
where ‘d’ explains 85.9% (p=0.0079) of PCrBee variability (Fig 4).
Take-off and landing directions
Under light wind conditions, no differences were found among the directions of either take-offs (p=0.985) or landings (p=0.521) (Table I). However, at a mild wind velocity (about 24 km/hr), a clear difference was recorded among take-off directions (p=0.028), with an even greater difference among landing directions (p=0.0009). Landings were mostly directed upwind (p=0.0038), whereas most of the take-offs were directed perpendicularly to the wind (p=0.0030).
DISCUSSION
Two foraging strategies can be distinguished among the social bees: fixed-course foraging (‘traplining’), where the bee gathers food along a route which it returns to in subsequent flights and on subsequent days; and fixed-site-and-species foraging (‘species and site constancy’), where the bee collects the food from one species in a small patch (1-3 trees or a similar field area), to which it returns in subsequent flights and on subsequent days (Frisch, 1967; Heinrich, 1983).
Fixed-site-and-species foraging strategy is typical of the honey bees, which sophisticatedly transfer information about food sources among the individual bees (Frisch, 1967; Heinrich, 1983). According to this information, ‘work groups’ are formed in the hive, where all the bees visit the same food source. These workers show a high level of constancy to the food location and to the flower species, and hence their mobility during foraging is limited and their efficiency as cross-pollinators is low. The information gathering is effected by ‘scouts’, who visit several locations and flower species during a flight (Frisch, 1967; Heinrich, 1983), and perform efficient cross-pollination. The proportion of scouts among the foraging bees sets the cross-pollination efficiency of the colony (Free, 1993).
Our data show that the proportion of honey-bee mobility to an adjacent row is not influenced by bee density (Fig 1), which may be a result of zero energetic gain to the food collectors from increasing mobility of this type. Therefore, we assume that the reports about increase in bee mobility with increased bee density (Clark, 1923; Stout, 1933; Butler, 1945; Free, 1960b; Jaycox, 1964; Frankie and Baker, 1974; McGregor, 1976; Vithanage, 1990) refer to growth in the number of bees crossing between adjacent trees, and not in their percentage. However, The percentage of short-range bee mobility is seems to increase with an increase in wind velocity (equation 5, Fig 2), as has been qualitatively observed previously (Free, 1960b; Friesen, 1973; Woodell, 1978). This effect may result from the increase in upwind-directed movement of the bees with the increase in wind velocity (Fig 3), and perhaps also from an expansion of their foraging area.
The “tendency to forage upwind” (Dag and Eisikowitch, 1995), which we found to grow with increased wind velocity (Fig 3), has been explained as the product of aerodynamic restrictions, since it is easier to take-off and land against the wind (Woodell, 1978), or as the result of olfactory search for food (Friesen, 1973; Woodell, 1978). We found that the bees tend to land against the wind (Table I), but they may change their flight direction beforehand. Nevertheless, strong evidence for the olfactory explanation has been provided (Wenner and Wells, 1990), and it appears that during foraging of either a food collector or a scout bee, food source location is aided by the sense of smell, and thus motion in the upwind direction is preferred. Another explanation is that the bees choose to forage in an upwind direction while loading, and to fly downwind back to the hive when they are full and heavy (D. Roubik, personal communication).
We found an average of 41% of the bees crossing between adjacent rows of avocado medium-size trees in a 10-min period. Because under good avocado-pasture conditions honey bees fill their crop in about 10 min (Ish-Am, 1994), we used this figure to estimate the percentage of bees crossing between adjacent rows during foraging. Both bee-mobility rate and the upwind-direction preference grew with increasing wind velocity (Figs 2, 3), and under a wind speed of about 45 km/hr the upwind mobility to the adjacent row reached 100% of the bees. This short-range mobility is higher than has been described before (Clark, 1923; Butler et al, 1943; Bateman, 1947; Free, 1960b; Free and Spencer-Booth, 1964b; Bergh, 1967; Davenport, 1986; Free, 1993), and may explain the high rate of both cross-pollination and fruits of cross-pollination that was found in rows adjacent to the pollen donor, and also the growth in production in these rows, compared to rows further away (Free, 1962; Free and Spencer-Booth, 1964a; Bergh, 1967, 1969; Gil et al, 1986; Degani et al, 1989; Ish-Am, 1994).
Our work indicates that the percentage of cross-pollinating bees stabilizes at about 1%-2% of the workers, at a distance of 10 to 15 rows away from the male bloom. This percentage conforms with the early assumption, that this value is about 1% (Free, 1966b). This low percentage of cross-pollinating honey bees should explain the reports of cross-pollination, and cross-pollinated fruit set, at long distances from the pollen source (Vrecenar-Gadus and Ellstrand, 1985; Degani et al, 1989; Ish-Am, 1994). Nevertheless, a rise in the percentage of scouts among the foraging bees with an increase in competition (Free, 1993) requires further study. The division of labor between food and information gatherers may be flexible, and, under increased-competition conditions, a food collecting honey bees may switch to information gathering. Under such circumstances, an increase in cross-pollination efficiency of the honey-bee population is expected.
An inconsistency was encountered with regard to the short-range bee mobility. Direct counts of bees resulted in an average of 40% of the bees crossing to an adjacent row (Fig 1), whereas counts of pollen-load carriers led to only 14% cross-pollinating bees in the first female-blooming row (Fig 4). This difference may stem from a low tendency of the bees that collect pollen and nectar on the male bloom, to cross to the female bloom, which carries nectar only.
ACKNOWLEDGMENTS
The authors thank the farmers from Eilon, Rosh-Hanikra and Kabri, who helped in the execution of this research; R Ticher, who performed the statistical analysis; and R Hofshi, C Vainstein and T Steinitz, who assisted with the English edition. We also thank the anonymous reviewers for their enlightening comments. The research was funded by the Israeli Fruit Growers Organization and the Beekeeper Association, by the Committee for New Cultivation of the Ministry of Agriculture and by the David Ben-Gurion Fund of the Histadrut.
Detailed summary - Honey-bee mobility during foraging greatly influences the effectiveness of bees as cross-pollinators. This is especially important for crops such as avocado, which require, or are benefited by, cross-pollination, and the pollen of which is not transferred by the bees through the hive. We measured honey-bee mobility in avocado orchards, between neighboring trees and to a distance of 15 rows, as a function of bee density, and of both wind velocity and direction. Honey-bee density (bees per tree) was counted on a whole tree, and averaged for five trees. Bee mobility among adjacent trees was measured by counting bee movements in all directions among four adjacent trees. Honey-bee mobility to a distance greater than one row was calculated from counts of bees with and without pollen loads, simultaneously in avocado rows bearing pollen-releasing male flowers and in adjacent rows bearing only female flowers, at several distances from the male bloom. Honey-bee take-off and landing directions were recorded on avocado flowering branches, and related to wind velocity and direction. The average percentage of bees that crossed in a 10-min period between adjacent rows was 41%. This percentage was not influenced by bee density, i e the number of bees that crossed correlated linearly to their density (Fig 1). However, this percentage increased with the increase in wind velocity, from 30% in a light wind to up to 65% in a strong wind (Fig 2). The bees traveled mostly toward upwind, a tendency which also grew with increased wind velocity (Fig 3). Thus, during a strong wind, close to 100% of the bees traveled in an upwind direction from one row to the adjacent one in a 10-min period. Under light wind conditions no differences were found among the directions of either take-offs or landings of the bees, but at a wind velocity of 24 km/hr the landings were mostly directed upwind, while most of the take-off movements were perpendicular to wind direction (Table I). Bee mobility for a distance greater than one row was effected above the canopy, apparently in search of alternative food sources. The percentage of cross-pollinating bees derived from this mobility decreased following a hyperbolic curve with increasing distance from the pollen source, approaching 1%-2% at a distance of 10-15 rows (Fig 4). Most of the field-working honey bees show a high level of constancy to both flower species and food location, and hence their mobility during foraging is limited, and their efficiency as cross-pollinators low. Only the scout bees, who gather both food and information, exhibit high mobility during foraging, and perform cross-pollination efficiently. However, since they consist only 1%-2% of the field-working honey bees, the efficiency of the honey bees as cross pollinators is low.
REFERENCES
Bateman AJ (1947)
Contamination in seed crops. III. Relation with isolation distance. Heredity 1, 303-336
Bergh BO (1967) Reasons for low yields of avocados. Calif
Avocado Soc Yearbook 51, 161-172
Bergh BO (1969) Avocado, Persea americana
Miller. In: Outlines of Perenial Crop Breeding in the Tropics (FP
Ferwerda, F Wit, eds), Veerman & Zonen, Wageningen, Netherlands
Butler CG (1945) The influence of various physical
and biological factors of the environment on the honeybee activity. An
examination of the relationship between activity and nectar concentration and
abundance. J Exp Biol 21, 5-12
Butler CG, Jeffree EP, Kalmus H (1943) The behaviour
of a population of honeybees on an artificial and on a natural crop. J Exp
Biol 20, 65-73
Clark IO (1923) Avocado pollination and bees. Calif
Avocado Assoc Ann Rep 8, 57-62
Dag A, Eisikowitch D (1995) The influence of hive
location on honeybee foraging activity and fruit set in melons grown in plastic
greenhouses. Apidologie 26, 511-519
Davenport DL (1986) Avocado flowering. Hort Rev
8, 257-289
Degani C, Goldring A, Gazit S (1989) Pollen parent
effect on outcrossing rate in 'Hass' and 'Fuerte' avocado plots during fruit
development. J Amer Soc Hort Sci 114, 106-111
DeGrandi-Hoffman G, Hoopingarner R, Baker KK (1984)
Pollen transfer in apple orchards: tree-to-tree or bee-to-bee? Bee World
65, 126-133
DeGrandi-Hoffman G, Thorp R, Loper G, Eisikowitch D
(1992) Identification and distribution of cross-pollinating honey-bees on
almonds. Journal of Applied Ecology 29, 238-246
Frankie GW, Baker HG (1974) The importance of
pollinator behavior in the reproduction biology of tropical trees. An Inst
Univ Nal Auton Mexico, Ser Botanica 45, 1-10
Free JB (1960a) The behaviour of honeybees visiting
flowers of fruit trees. J Anim Ecol 29, 385-395
Free JB (1960b) The pollination of fruit trees. Bee
World 41, 141-151, 169-186
Free JB (1962) The effect of distance from
pollinizer varieties on the fruit set on trees in plum and apple orchards. J
Hort Sci 37, 262-271
Free JB (1966a) The foraging areas of honeybees in
an orchard of standard apple trees. J App Ecol 3, 261-268
Free JB (1966b) The foraging behaviour of bees and
its effect on the isolation and speciation of plants. In: Reproductive
Biology and Taxonomy of Vascular Plants. B.S.B.I. Conference Report No.9.
Botanical Society of the British Iles, Pergamon Press, Oxford
Free JB (1993) Insect Pollination of Crops.
2nd ed. Academic Press, London
Free JB, Spencer-Booth Y (1963) The foraging areas
of honey-bee colonies in fruit orchards. J Hort Sci 38, 129-137
Free JB, Spencer-Booth Y (1964a) The effect of
distance from pollinizer varieties on the fruit set of apple, pear and
sweet-cherry trees. J Hort Sci 39, 54-60
Free JB, Spencer-Booth Y (1964b) The foraging
behaviour of honey-bees in an orchard of dwarf apple trees. J Hort Sci
39, 78-83
Free JB, Durrant AJ (1966) The transport of pollen
by honey-bees from one foraging trip to the next. J Hort Sci 41, 87-89
Friesen LJ (1973) The search dynamics of recruited
honey bees, Apis mellifera ligustica spinola. Biol Bull 144,
107-131
Frisch K von (1967) The Dance Language and
Orientation of Bees. Harvard University Press, Cambridge, Massachusetts
Gil I, Harat T, Moran I (1986) Fertilization of
‘Hass’ avocado cultivar (In Hebrew). Alon Hanotea 40, 443-455
Heinrich B (1983) Insects foraging energetics. In:
Handbook of Experimental Biology (CE Jones, RJ Little, eds), Scientific and
Academic Editions, New York
Ish-Am G (1994) Interrelationship between avocado
flowering and honey bees and its implication on the avocado fruitfulness in
Israel (In Hebrew, English Abstr). Ph D Thesis, Tel Aviv University, Tel
Aviv, Israel
Ish-Am G, Eisikowitch D (1992) New insight into
avocado flowering in relation to its pollination. Calif Avocado Soc Yearbook
75, 125-137
Ish-Am G, Eisikowitch D (1993) The behaviour of
honey bees (Apis mellifera) visiting avocado (Persea americana)
flowers and their contribution to its pollination. J Apic Res 32,
175-186
Jaycox ER (1964) Pollination of fruit trees by honey
bees. Amer Bee J 104, 338-339
Mayer DF, Johansen CA, Lunden JD (1986) Bee
pollination of tree fruits. Washington State Univ PNW Coop Ext Bull
02820
McGregor SE (1976) Insect Pollination of
Cultivated Crop-Plants. U S D A
Agriculture Handbook No. 496, Washington D C
Sokal RR, Rohlf FJ (1981) Biometry: The
Principles and Practices of Statistics in Biological Research. State
University of New York, Stony Brook, New York
Stout AB (1923) A study in cross-pollination of
avocado in southern California. Calif Avocado Assoc Ann Rep 7, 29-45
Stout AB (1933) The pollination of avocados. Fla
Agricl Exp Stat Bull 257, 1-44
Vithanage HINV (1990) The role of the European
honeybee (Apis mellifera L.) in avocado pollination. J Hort Sci
65, 81-86
Vrecenar-Gadus M, Ellstrand LC (1985) The effect of
planting design on out-crossing rate and yield in the 'Hass' avocado. Sci
Hort 27, 215-221
Wenner AM, Wells PH (1990) Anatomy of
Controversy: the Question of a "Language" Among Bees. Columbia
University Press, New York
Woodell SRJ (1978) Directionality in bumblebees in
relation to environmental factors. In: The Pollination of Flowers by Insects
(AJ Richards, ed), Academic Press, London
Legends to figures and
tables
Table I:
Percentages of honey-bees taking-off to, and landing from various directions,
at two wind speeds
Differences among directions for each row were tested using a C² test of homogeneity (see Results).
Fig 1: Honey-bee mobility to the adjacent row
vs bee density
Number of honey bees crossing between
parallel trees of neighboring rows in a 10-min period (NoMob), and the
corresponding percent (PrMob), as a function of bee density (BPT = bees per
tree).
Linear regressions are:
NoMob = 0.412*BPT - 0.146, r = 0.70, p < 0.0001,
n = 26
PrMob = 0.028*BPT + 38.8, r = 0.04, n s, n = 26
Fig 2: Honey-bee mobility to the adjacent row
vs wind velocity
Number of honey bees crossing between
parallel trees of neighboring rows in a 10-min period (NoMob), and the
corresponding percent (PrMob), as a function of wind velocity (WindVel, in
km/hr).
Linear regressions are:
NoMob = 0.147*WindVel + 21.6, r = 0.07, n s, n = 26
PrMob = 0.882* WindVel + 29.1, r = 0.36, p = 0.042, n =
26
Fig 3: Honey-bee preference of upwind
direction vs wind velocity
Honey-bee preference of upwind direction
(UWMP = upwind mobility preference), which is the ratio between the number of
honey bees crossing to the nearest tree in the upwind direction and the
corresponding number in the downwind direction, as a function of wind velocity
(WindVel, in km/hr).
The free linear regression is:
UWMP = 0.052*WindVel + 1.08, r = 0.65, p = 0.0007, n
= 23,
and the forced to [0,1] linear regression is:
UWMP = 0.056*WindVel + 1.00, r = 0.65, p = 0.0006, n
= 23,
Fig 4: Percentage of cross-pollinating bees vs
distance from the pollen source
Percentage of cross-pollinating bees on
female-stage blooming trees (PCrBee), as a function of distance (D, in rows)
from the male-stage trees. Honey-bees carrying pollen loads and nonloaded bees
were simultaneously counted on trees in male-stage and on trees in female-stage
bloom. The counts of each row were pooled to calculate the average PCrBee per
row (see Materials and Methods).
The best-fit function of regression of the
averages to D is:
PCrBee = 12.3/D + 0.323, r = 0.93, p = 0.0079, n
= 6,
Table I:
Percentages of honey-bee taking-off to, and landing from various directions, at
two wind speeds
|
Date |
Wind |
Wind |
Honeybee |
n |
Direction of honey-bee movement |
|||
|
|
speed |
direction |
activity |
|
N |
E |
S |
W |
|
27.4.90 |
3.7 km/hr |
East |
take-off |
91 |
25.7 |
23.2 |
24.2 |
26.9 |
|
|
|
|
landing |
114 |
17.8 |
31.5 |
24.8 |
25.9 |
|
28.4.90 |
24 km/hr |
West |
take-off |
150 |
12.8 |
25.9 |
36.4 |
24.9 |
|
|
|
|
landing |
180 |
16.9 |
20.9 |
16.4 |
45.7 |
Differences among the directions for each row were tested using a C² test of homogeneity (see in Results).
Fig 1
Fig
2
Fig
3
Fig
4