Recombination hotspot variation across populations

They find some interesting and strong population differences in recombination rates, but don't really give any clues as to any possible explanations other than:

Differences in recombination rates and presence or absence of hotspots may follow a more complex pattern related to the genetic diversity among humans due to their evolutionary history and its impact on the DNA sequences, that may be at the base of differences in recombination rates.

Variation in estimated recombination rates across human populations

Jan Graffelman, David J. Balding, Anna Gonzalez-Neira and Jaume Bertranpetit

Human Genetics Volume 122, Numbers 3-4 / November, 2007

Abstract: Recently it has been reported that recombination hotspots appear to be highly variable between humans and chimpanzees, and there is evidence for between-person variability in hotspots, and evolutionary transience. To understand the nature of variation in human recombination rates, it is important to describe patterns of variability across populations. Direct measurement of recombination rates remains infeasible on a large scale, and population-genetic approaches can be imprecise, and are affected by demographic history. Reports to date have suggested broad similarity in recombination rates at large genomic scales and across human populations. Here, we examine recombination rate estimates at a finer population and genomic scale: 28 worldwide populations and 107 SNPs in a 1 Mb stretch of chromosome 22q. We employ analysis of variance of recombination rate estimates, corrected for differences in effective population size using genome-wide microsatellite mutation rate estimates. We find substantial variation in fine-scale rates between populations, but reduced variation within continental groups. All effects examined (SNP-pair, region, population and interactions) were highly significant. Adjustment for effective population size made little difference to the conclusions. Observed hotspots tended to be conserved across populations, albeit at varying intensities. This holds particularly for populations from the same region, and also to a considerable degree across geographical regions. However, some hotspots appear to be population-specific. Several results from studies on the population history of humans are in accordance with our analysis. Our results suggest that between-population variation in DNA sequences may underly recombination rate variation.