The effects of meliponicultural use of Tetragonula laeviceps on other bee pollinators and pollination efficacy of lemon

Study site

This study was carried out in Cibodas, West Bandung Regency, West Java, Indonesia, at coordinates latitude 6°49′20″S and longitude 107°40′35″E, with an altitude of 1,219 meters above sea level. The study covered a total land area of 60 × 60 m² and included 200 plants of lemon (Citrus limon Eureka) aged 3 years, with an average height of 2 m and a canopy width of 2 m. The eastern, southern, and western parts of the lemon orchard are surrounded by horticultural farming including eggplants, tomatoes, and cabbage, while the northern part is bordered by teak forests. The colony carrying capacity in lemon has been calculated and the results require four colonies of stingless bee Tetragonula laeviceps, with approximately 400 to 600 adult worker bees for each colony (Bareke et al., 2020). The colonies were obtained from cultivators in Banjaran, West Java, Indonesia and acclimatized for 1 week before the observations.

The study was investigated in two periods, with the first period from March to June 2023 and the second period from July to October 2023, with four cultivated colonies of T. laeviceps in Modular Tetragonula Hives (MOTIVEs) with size 20 × 20 × 15 cm³ (Abduh et al., 2020). The MOTIVEs were placed in the middle of a lemon orchard in late March and left until October 2023. The study employed four plots placed in accordance with the compass, each measuring 15 × 15 m², with 16 lemon plants per plot. Each plant received 5 kg of organic fertilizer made from chicken manure in March and July 2023. Observations were conducted from 07:00 to 15:00, employing 15 min intervals for each plot. Observations for plots without meliponiculture were conducted weekly throughout March 2023, whereas observations for plots with meliponiculture were undertaken weekly from April to October 2023. Microclimate conditions at the study site were measured using the Data Logger HOBO U10-003.

Diversity, foraging behavior, and population of bee pollinators

Wild insect pollinators were collected by sweep net (mesh size 0.9 × 0.3 mm²) using a dried preservation technique and subsequently pinned. Furthermore, lemon flowers and bee pollinators carrying pollen were captured and inserted into a 25 mL colonial tube containing 15 mL of 70% alcohol. These specimens were sent to the Laboratory of Entomology, School of Life Sciences and Technology, Institut Teknologi Bandung, Indonesia for taxonomic identification (Gibb & Oseto, 2019; Mason et al., 2022) and number of lemon pollen and pollen load of bee pollinators analysis. After determining the taxonomic identification of pollinators, the diversity of pollinators was analyzed using the percentage of relative abundance and Eq. (1).

(1) $$Relativeabundance\left(\mathrm{\%}\right)={\displaystyle \frac{Totalnumberofeachpollinatorspecies}{Totalnumberofpollinators}\times 100\mathrm{\%}.}$$

Foraging behavior of bee pollinators was observed using a scan sampling method across four plots, with each plot observed for 15 min to determine the number of bees visiting blooming lemon flowers, while the time spent by bee pollinators per flower was recorded using a focal sampling method, which involves directly observing each pollinator from the present moment it visits the flower until it leaves (Putra, Permana & Kinasih, 2014; Nurdiansyah, Abduh & Permana, 2023). Lemon flowers were wrapped in an insect net (mesh size 36 mm) before they bloomed in the second period. When they bloomed fully, nectar was collected using micro hematocrit capillaries (length 75 mm, diameter 1.55 mm) each hour from 07:00 to 16:00. If nectar was not collected from the flowers, they were wrapped again in an insect net to prevent insects from accessing the nectar. Volume was measured with a micropipette, and the concentration was determined using a hand refractometer.

For pollen analysis, flowers or bees were centrifuged for 5 min at 3,500 rpm, and then the flowers or bees were removed. Subsequently, another centrifugation was performed for 3 min at 2,000 rpm, and the supernatant was removed. A solution of acetolysis (0.9 mL acetic anhydride + 0.1 mL sulfuric acid) was added, and the samples were heated in a water bath at 80 °C for 5 min. Afterward, 1 mL of distilled water was added. The number of pollen grains was quantified using 0.1 mL of samples at hemocytometer in four quadrants under a light microscope (eyepiece lens 10× and objective lens 10×/0.24) (Nikon Instruments, Tokyo, Japan). The number of pollens grains from flowers and bees was calculated using the total volume of solution multiplied by the number of pollen grains counted and divided into a volume of four quadrants (Alpionita, Atmowidi & Kahono, 2021).

The foraging behaviors of bee pollinators were assessed based on visitation rates, which were categorized into three intervals: morning (7:00 to 10:00), noon (10:00 to 13:00), and afternoon (13:00 to 16:00). The visitation rates of bee pollinators was calculated as the number of bee pollinators visiting flowers divided by the number of flowers available per observation (Gallagher & Campbell, 2020). The total number of bees visiting blooming flowers of each species and the total number of blooming flowers per month were utilized to analyze the monthly population of bee pollinators.

Pollination efficacy and productivity estimation of lemon

Pollination efficacy was investigated in two periods, with the first period from flower until harvest observed from March to June 2023, and the second period observed from July to October 2023. Pollination efficacy was compared between without and with T. laeviceps based on various parameters, including number of fruit sets, pollination success, fruit weight, and yield per plant in the first period, as well as with meliponiculture in the first and second period. The comparison involves 15 flowers per plant from 64 plants that were randomly tagged, and after days of blooming stages, the number of fruit sets was calculated. Percentage of fruit set was determined by calculating as the number of fruit sets divided by the number of flowers, multiplied by 100% (Nurdiansyah, Abduh & Permana, 2023). When harvesting, three lemons were selected at random from each plant and weighed using a digital analytical balance. The yield per lemon plant was determined by multiplying the average fruit weight by the number of harvested fruits. Lemon productivity was also estimated per hectare given there are 900 plants using a 3 × 3 m² spacing. The estimated lemon productivity was determined by multiplying the average fruit weight by the number of fruits harvested, the number of plants, and accounting for three harvest cycles within a single year (Nurdiansyah, Abduh & Permana, 2023).

Statistical analysis

All data were analyzed for normality and homogeneity of variance, and no data transformations were applied. Effects of T. laeviceps were analyzed using the two-sample t-test (p < 0.05) for the parameter’s relative abundance, daily foraging activity, fruit sets, fruit weight, and yield per plant. Analysis of variance (ANOVA) followed by Tukey’s multiple comparison test (p < 0.05) were performed to compare the pollen load, time spent, and visitation rates of bee pollinators. Pearson’s correlation coefficient was employed to assess the significance of the correlation between the number of pollinators visiting lemon flower with environmental factors. Additionally, monthly correlation calculations were conducted to examine the relationship between the monthly population of bee pollinators and the number of flowers. To identify the primary components influencing the foraging behavior of bee pollinators on lemon flowers, a principal component analysis (PCA) was performed. The statistical analyses were performed using the R program version 4.3.2. (R Core Development Team, 2023).

Diversity of insects visiting lemon flowers

Without meliponiculture, wild insects visiting lemon flowers were six species of bees, i.e., Apis cerana (Hymenoptera: Apidae), Lasioglossum albescens (Hymenoptera: Halictidae), Megachile laticeps (Hymenoptera: Megachilidae), Xylocopa confusa (Hymenoptera: Apidae), Xylocopa latipes (Hymenoptera: Apidae), and Xylocopa caerulea (Hymenoptera: Apidae), along with four species of non-bees, including Dolichoderus thoracicus (Hymenoptera: Formicidae), Papilio demoleus (Lepidoptera: Papilionidae), Delias belisama (Lepidoptera: Pieridae), and Hypolimnas misippus (Lepidoptera: Nymphalidae). However, the diversity of wild insect visitors with meliponiculture was maintained, and Tetragonula laeviceps (Hymenoptera: Apidae) emerged as a new pollinator for lemon flowers with the highest relative abundance (t₍₃₎ = 60.00, p = 1.02E⁻⁵). The relative abundance of wild bees visiting lemon flowers decreased significantly in the first period, including A. cerana (t₍₆₎ = 31.84; p = 6.38E⁻⁸), L. albescens (t₍₆₎ = 28.54; p = 1.23E⁻⁷), M. laticeps (t₍₆₎ = 8.69; p = 1.28E⁻⁴), X. confusa (t₍₆₎ = 31.72; p = 6.52E⁻⁸), X. confusa (t₍₆₎ = 31.72; p = 6.52E⁻⁸), X. caerulea (t₍₆₎ = 18.04; p = 1.87E⁻⁶), while the non-bees remained constant (Table 1). There was no change in the relative abundance of wild insect visitors between the first period with meliponiculture and second period with meliponiculture (p > 0.05).

Foraging behavior of bee pollinators

The daily foraging activity of bee pollinators visiting lemon flowers starts from 7:00 to 16:00 (Fig. 1). The highest number of bee pollinators visiting lemon flowers occurred at 11:00 in both the first and second periods. However, the highest number of bee pollinators visiting lemon flowers in the first period significantly decreased between without and with meliponiculture, including A. cerana (t₍₆₎ = 85.90; p = 1.56E⁻¹⁰), L. albescens (t₍₆₎ = 48.18; p = 5.36E⁻⁹), M. laticeps (t₍₆₎ = 51.30; p = 3.68E⁻⁹), X. confusa (t₍₆₎ = 47.82; p = 6.60E⁻⁹), X. latipes (t₍₆₎ = 18.59; p = 1.56E⁻⁶), and X. caerulea (t₍₆₎ = 18.64, p = 1.54E⁻⁶). There were no significant differences in daily foraging activity between with meliponiculture in the first and with meliponiculture in the second period (p > 0.05).

Each visit to lemon flowers by T. laeviceps follows a distinct pollination sequence (Fig. 2). It starts with approaching (Fig. 2A) the flower at a position parallel to 45 degrees from its original position, followed by perching on a petal (Fig. 2B) or an anther (Fig. 2C). Subsequently, it enters the nectary flower to collect nectar (Fig. 2D), then climbs to the anther to collect pollen (Fig. 2E). When solely interested in pollen, T. laeviceps directly perching goes to the anther. Upon completing its activities, it leaves the lemon flower from the anther (Fig. 2F) without buzzing and moves to search for the nearest flower to resume collecting nectar and pollen (Fig. 2G). T. laeviceps exhibited the longest time visiting lemon flower (F_{(6, 105)} = 12.22; p = 0.000), while M. laticeps exhibited the fastest time (F_{(6, 105)} = 26.01; p = 0.000) (Table 2).

Lemon flowers produce 12,539 ± 376 pollen grains per flower. Each bee pollinator carries pollen on its body after visiting a lemon flower and the pollen load of each bee pollinator differs (p < 0.05). Based on pollen load on their bodies, bee pollinators seem to visit not only lemon flowers but also other flowers in the lemon orchard, such as tomatoes and eggplants. The honey bee A. cerana has the highest pollen load in lemon flowers, with 84,875 pollen grains (F_{(6, 105)} = 71,728.25; p = 0.000), while the stingless bee T. laeviceps carries only 6,124 pollen grains (F_{(6, 105)} = 5,239.94; p = 0.000). Furthermore, M. laticeps (F_{(6, 105)} = 12,372.51; p = 0.000) carries the highest number of pollen grains from various plant flowers. The volume of nectar on lemon flowers continued to increase from 7:00 to 11:00 and decreased until 16:00 when the last bee pollinators visited lemon flowers (Fig. 3A). However, the nectar concentration continued to increase from 7:00 to 16:00. The highest pollinator visitation rates occurred at noon (10:00 to 13:00), with T. laeviceps being the most significant pollinator visiting lemon flowers (F_{(12, 483)} = 7.42; p = 0.000). However, it was observed that the lowest pollinator visitation rate was for X. caerulea (F_{(12, 483)} = 0.01; p = 0.000) during the afternoon (13:00 to 16:00).

The number of bee pollinators visiting lemon flowers was then analyzed in correlation with environmental factors such as microclimate conditions and nectar contents. Microclimate conditions including temperature, light intensity, and relative humidity during observations from March to October 2023, ranging from 20.27–24.29 °C, 589.3–5,442.5 lux, and 71.95–87.43%, respectively. The temperature followed the same pattern as the number of pollinators and also showed a very high positive correlation (r = 0.83; p = 2.22E⁻¹⁶), as with performed light intensity (r = 0.93; p = 2.22E⁻¹⁶) (Fig. 4). Whereas relative humidity follows the opposite pattern and showed a high negative correlation (r = −0.68; p = 3.26E⁻¹¹). In addition, the number of bee pollinators visiting lemon flowers also correlates with nectar content, with a high positive correlation with volume (r = 0.65; p = 6.20E⁻¹⁰), and a low positive correlation with concentration (r = 0.24; p = 0.04). Furthermore, a principal component analysis (PCA) was performed to explore the relationship between microclimate conditions, nectar content, and the number of pollinators visiting lemon flowers (Fig. 5). The results indicated that light intensity (Dim1 = 2.29; Dim2 = −0.36) is the most significant component that influenced on the number of pollinators (Dim1 = 2.33; Dim2 = −0.02), followed by temperature (Dim1 = 2.08; Dim2 = 0.34), relative humidity (Dim1 = −1.89; Dim2 = −0.87), volume (Dim1 = 1.71; Dim2 = −1.62) and concentration of nectar (Dim1 = 0.74; Dim2 = 2.26).

Population of bee pollinators

The monthly population of bee pollinators visiting lemon flowers shows fluctuations (Fig. 6). The population of bee pollinators and lemon flowers increased from March to April in the first period, but then decreased until June. In the second period, a similar pattern was observed. This indicates that the lemon production cycle lasts for 4 months, with phenological stages lasting 115–120 days in the first period and 121–125 days in the second period from full bloom to harvest from 20 flowers observed. The total number of bee pollinators exhibited a very high positive correlation (r = 0.98; p = 2.25E⁻⁵) with the total number of blooming lemon flowers.

Pollination efficacy and productivity estimation of lemon

The pollination efficacy on 64 lemon plants was evaluated using 15 flowers per plant (Table 3). The results of the first period showed a significant difference in the fruit set (t₍₁₂₆₎ = 26.47; p = 1.29E⁻⁵²), fruit weight (t₍₁₂₆₎ = 118.49; p = 4.44E⁻¹³¹), and yield per plant (t₍₁₂₆₎ = 108.63; p = 2.27E⁻¹²⁶) between without and with T. laeviceps. There were no significant differences in the parameters of fruit sets, fruit weight, and yield per plant in both sampling periods with T. laeviceps (p > 0.05). Estimated lemon productivity without meliponiculture is 11.62 ± 0.24 tons per hectare per year, while with meliponiculture is 15.19 ± 0.12 tons per hectare per year in the first period and 14.92 ± 0.16 tons per hectare per year in the second period.