A SNP-based honey bee paternity assignment test for evaluating the effectiveness of mating stations and its application to the Ataun valley, Basque Country, Spain

A SNP-based honey bee paternity assignment test for evaluating the effectiveness of mating stations and its application to the Ataun valley, Basque Country, Spain

Abstract

Geographically isolated mating stations are deployed across Europe to facilitate controlled mating with selected drone-producing colonies. To evaluate the effectiveness of these stations, we developed a paternity assignment test using a custom Illumina genotyping chip with 6,457 SNPs. The test relies on two key metrics: the number of mismatch alleles and kinship values. The method demonstrated remarkable accuracy during validation with an independent dataset of known parent-offspring pairs, with an accuracy rate of 97.7%. We then applied the developed paternity assignment test to the Apis mellifera iberiensis mating station in the Ataun Valley, Basque Country, Spain, in 2021. Drone-producing colonies in the valley were sampled and genotyped, as well as 156 worker offspring of queens mated at the station, and 56 drones collected in the drone congregation area. Out of the 156 worker samples, we could assign paternity of 120 (76.9%) to one of the drone-producing colonies in the valley, while 23.1% were of unknown patriline. Out of the 56 drones collected in the air, 52 (92.9%) were assigned to drone-producing colonies. We were also able to determine the colonies and apiaries that made the most significant contributions to the matings. This information aids in effective apiary management, including the selection of suitable mating station locations and the positioning of drone-producing colonies therein. Overall, our SNP-based paternity assignment test offers a valuable tool for evaluating mating station effectiveness across Europe, crucial for advancing breeding objectives in honey bee populations.







Abstract

Geographically isolated mating stations are deployed across Europe to facilitate controlled mating with selected drone-producing colonies. To evaluate the effectiveness of these stations, we developed a paternity assignment test using a custom Illumina genotyping chip with 6,457 SNPs. The test relies on two key metrics: the number of mismatch alleles and kinship values. The method demonstrated remarkable accuracy during validation with an independent dataset of known parent-offspring pairs, with an accuracy rate of 97.7%. We then applied the developed paternity assignment test to the Apis mellifera iberiensis mating station in the Ataun Valley, Basque Country, Spain, in 2021. Drone-producing colonies in the valley were sampled and genotyped, as well as 156 worker offspring of queens mated at the station, and 56 drones collected in the drone congregation area. Out of the 156 worker samples, we could assign paternity of 120 (76.9%) to one of the drone-producing colonies in the valley, while 23.1% were of unknown patriline. Out of the 56 drones collected in the air, 52 (92.9%) were assigned to drone-producing colonies. We were also able to determine the colonies and apiaries that made the most significant contributions to the matings. This information aids in effective apiary management, including the selection of suitable mating station locations and the positioning of drone-producing colonies therein. Overall, our SNP-based paternity assignment test offers a valuable tool for evaluating mating station effectiveness across Europe, crucial for advancing breeding objectives in honey bee populations.


Introduction

Honey bees have a complex mating biology, with queens mating multiple times in flight away from their colonies, at so-called drone congregation areas (DCAs) (Koeniger et al., Citation2014). This makes their controlled breeding for conservation or genetic improvement programs quite challenging (Neumann, Moritz et al., Citation1999). For conservation purposes of local native stock, it is crucial to control mating in order to avoid hybridization with foreign honey bees (Henriques, Browne et al., Citation2018; Parejo et al., Citation2016). For directed breeding programs, increased mating control and thus more accurate pedigree information ensures higher accuracy in estimating breeding values and ultimately leads to faster genetic progress (Plate et al., Citation2019).

To achieve controlled mating, various methods are employed. One approach involves setting up mating stations in geographically isolated locations, such as valleys or islands (Neumann, van Praagh, et al., Citation1999; Jensen et al., Citation2006). Geographic barriers help ensure that mating occurs with a certain degree of isolation, however, finding completely isolated locations is not straightforward given the high colony density in many regions (Chauzat et al., Citation2013) and the large mating flight distances of drones and queens (Ruttner & Ruttner, Citation1972; Jensen et al., Citation2006). Temporal isolation, another technique, involves controlling the timing of mating (Oxley et al., Citation2010; Musin et al., Citation2023). Artificial insemination is the most direct and controlled method, enabling total control over the genetic origin of the semen (Gillard & Oldroyd, Citation2020).

In Europe, numerous mating stations have been established to facilitate controlled honey bee mating (Bouga et al., Citation2011; Bienefeld et al., Citation2007). One such station is the Ataun mating station located in the Ataun Valley, Basque Country, Spain, which was set up for the genetic improvement program of the native bee of the Iberian Peninsula, Apis mellifera iberiensis. The location was chosen because of its orographic characteristics and its scarcity of pre-existing apiaries. The mating station is managed by the bee breeding association ERBEL (www.erbel.eus) and has been running since 2019.

The effectiveness of the Ataun mating station and other stations alike must be evaluated to ensure that mating with the preferred drone-producing colonies is indeed controlled. Genetic monitoring tools that have been developed for estimating hybridization in A. m. mellifera (Parejo et al., Citation2016; Henriques, Browne et al., Citation2018) and A. m. iberiensis (Henriques, Parejo et al., Citation2018) can give information on the percentage of pure matings. Observations of nuptial flights can also give information on the queen’s mating (Heidinger et al., Citation2014; Uzunov et al., Citation2024). However, a more direct way to evaluate the effectiveness of mating stations is the analysis of paternity between offspring of the queens mated at the station and the drone-producing colonies. Paternity analyses aim to answer critical questions, such as drones from which drone-producing colonies have mated with the virgin queens and what percentage of the mating drones are of foreign (uncontrolled) origin.

Traditionally, paternity in various species, including humans and animals, has been assigned using microsatellite markers (Pena & Chakraborty, Citation1994; Tian et al., Citation2008). While these markers are informative due to their multiallelic nature, they also have limitations: Their use is labor-intensive, tedious, and requires calibration among different laboratories (Ashley & Dow, Citation1994). In recent years, single nucleotide polymorphism (SNP) genotyping has become the standard method for assigning paternity (Clarke et al., Citation2014; Kaiser et al., Citation2017; Zhao et al., Citation2018). It allows for the simultaneous genotyping of thousands of loci, making it a more efficient and cost-effective approach. This shift to SNP genotyping is driven by the desire for greater accuracy, ease of use, and compatibility across different research settings. In honey bees, it has already been shown that SNPs outperform microsatellites in estimating hybridization (Muñoz et al., Citation2017; Parejo et al., Citation2018), and the use of SNPs to estimate kinship and pedigree information has already been performed in the context of genomic selection (Bernstein et al., Citation2022; Bernstein et al., Citation2023).

In this study, we present a paternity assignment test for honey bees based on SNP markers and built on the calculation of two metrics: (1) the KING kinship coefficient and (2) the probability of identity-by-state equal 0 (P(IBS = 0)). To this end, we employ the KING-robust algorithm developed by Manichaikul et al. (Citation2010) based on large-scale unlinked SNP sets that can be easily calculated between any pair of individuals. The use of the KING-robust kinship coefficient for parentage testing has the advantage that it does not depend on population-estimated minor allele frequencies, i.e., the estimation of kinship coefficients is robust and independent of sample composition or population structure and thus reliable across different groups of individuals and populations (Manichaikul et al., Citation2010). This property, known as sample invariance, is important for ensuring that the method can be applied widely and accurately. The second metric used in our approach is the probability of identity by state equalling zero P(IBS = 0). This metric measures the proportion of incompatible (mismatch) alleles shared between a pair of individuals, i.e., two individuals having homozygous opposite alleles. Unlike the kinship coefficient, this metric can differentiate parent-child relationships from sibling relationships.

Here, we developed a SNP-based paternity assignment test for honey bees based on the above-mentioned metrics and a honey bee genotyping chip with a total of 6457 SNPs. We first calculated kinship and P(IBS = 0) on known parent-offspring relationships to set sensible thresholds for paternity assignment, then evaluated the performance of our method in an independent test set of known parent-offspring and unrelated pairs, and finally, applied the method to samples from the mating station in the Ataun Valley, Basque Country, Spain. The newly developed method proves valuable for optimizing apiary management, specifically aiding in the selection of appropriate mating station locations and the strategic placement of drone-producing colonies therein. The SNP-based paternity assignment test holds great potential in assessing the effectiveness of mating stations throughout Europe, playing a pivotal role in progressing towards established breeding objectives.