Beekeeping improves shea pollination and fruit set in West African Agroforestry parklands

Shea (Vitellaria paradoxa C.F Gaertn) is a fruit tree of domestic and industrial importance in arid and semi-arid regions of Sub-Saharan Africa. Fruit set is largely dependent on insect pollination but recent studies have revealed a pollen deficit. Introduction of managed bees into orchards is an effective approach for enhanced pollination services in temperate climates. However, there is limited information to guide the adoption of this practice in shea agroforestry parklands.

This study investigated the effect of managed honey bee colonies (Apis mellifera) on fruit yield in six shea parklands across three regions of Ghana.

Tree proximity to the apiary had a detectable effect on fruit set within a 500 m range of the apiary.  Proximity of shea trees to apiary was significantly related to number of immature fruit set but not number of mature fruits. Fruit weight and size were not significantly related to distance from apiary nor pollination treatment.

This implies that the introduction of beekeeping has the potential to address shea pollination deficit at least within a 500 m range of the apiary.  Further studies are needed to determine the optimal hive density per acreage of shea parkland to maximise pollination services.

Methods
Field layout and tree selection

The study was conducted in shea parklands of six selected sites in Upper West, Upper East and North East regions of Ghana. In each region, two replicate sites were selected ensuring a minimum distance of 4 km and a maximum of 25 km to minimize variability in climatic conditions between sites (Figure 1).  Experimental sites were located in young fallows (shea parklands that were uncultivated for 2-5 years) that had established apiaries. The selected parklands had apiaries of ten empty beehives each installed in July 2017.   Sites that had six to eight hives colonised by Apis mellifera adansonii as of October 2020 were selected for the study. Apis mellifera is native to the Guinea savanna and known to be a primary pollinator of shea (Kwapong 2014).

In the shea parkland, four distances;100, 500, 1000 and 4000m (control) away from the apiary were established. The control was located at 4000m away from the apiary considering the design of Delaney et al. (2020). This was to limit the influence of external factors (semi- natural sites, feral bee colonies, human settlement, and climatic factors etc.) on pollinator abundance.

Five focal trees of a predetermined size class (DBH; 25-30 cm, Height ≥ 10 m and crown area ≥ 20 m2) were randomly selected at each distance range and the control parkland. These DBH and height classes were selected because they contain the highest mature shea tree abundance in the Guinea savanna zone (Tom-Dery et al., 2015). This also standardized the samples to minimise tree effect on yield. Similarly in the control shea parkland, a predetermined point was identified and five focal trees of the pre-determined size class were selected within a 100 m radius.

Experimental Treatments

Three pollination treatments were applied to inflorescences of each focal tree and replicated three times in a tree.  The pollination treatments were:

Open pollination (“open”): Tagged inflorescences that were left unmanipulated and exposed to all agents of pollination throughout the flowering period (Figure 2.a)

Insect exclusion (“bagged”): inflorescences bagged at bud stage with a muslin net (net mesh size diameter of approximately 1mm) to prevent all insect pollinators from accessing the flowers throughout the flowering season.  A binding wire (diameter ~ 0.61mm) was erected over the inflorescences before the net was wrapped to prevent wind from pushing net in close contact with flowers (Figure 2.b).

Supplemental Hand pollination (hand):  Pollen was manually applied to protruding stigmas directly from the anthers of other neighbouring shea trees.  Pollen was supplemented twice at an interval of two weeks during peak flowering (Figure 2.c).

In the UW region, all treatments were applied in December/January 2020/21 whilst that of NE and UE were applied in February/March 2021. The timing of treatment application differed due to variation in timing of peak flowering.  At the fructification stage, the number of fruit set per focal inflorescence was counted every fortnight till fruit maturity. Mature fruits were harvested and weighed prior to fruit fall. Trees were climbed with the help of a ladder.  Fruit length and width were measured with a digital calliper, after which the fruit pulp was removed before nut weight and size were measured (Figure 3).