Patterns of recombination can be inferred from genome-wide polymorphisms data using different approaches: meiotic recombination rates can be estimated by identifying crossovers within individual meioses from pedigree data; more precise estimates can be obtained using population genetic approaches in densely genotyped samples of unrelated individuals. However these population-based estimates can be biased as they are affected by evolutionary pressures, most notably, selection. In this work, we present a strategy to establish fine scale recombination maps by combining these two approaches. We exploiting two different datasets from one Sheep breed, the Lacaune: a large pedigree genotyped with a 50K SNP array and a sample of unrelated individuals genotyped with a 600K SNP array. Our analyses of these two datasets show that recombination patterns in the Sheep resemble of those observed in other mammals, in particular we demonstrate the presence of small intervals of a few kilobases exhibiting very large recombination rates (i. e. harbouring crossover hotspots). Then, we developped a statistical model aimed at formally combining the two datasets. This model allows first to scale the population-based estimates from the meiotic ones while accounting for their respective uncertainties and to produce high-resolution recombination maps. Second, we show that the combination of meiotic and population-based inference highlights the effect of selection on the genome, revealing already known selection signatures, but also new ones. Finally, we compared our recombination maps to those obtained from an analysis of pedigree data in Soay Sheep. This comparison allows to demonstrate that Soay Sheep exhibit about 20% more crossovers per meiosis compared to Lacaune Sheep and that this inflation is most likely due to selection in the Soay. Despite this strong past selection, we were able to show that the genetic determinism of recombination rate in the two populations remain very similar, two major loci (RNF212 and HEI10) contributiong greatly to its inter-individual variation in both populations. This suggest that recombination rate is genetically determined by multiple independant biological pathways