Response of in vitro pollen germination and pollen tube growth of groundnut (Arachis hypogaea L.) genotypes to temperature

Air temperatures of greater than 35 °C are frequently encountered in groundnut‐growing regions, especially in the semi‐arid tropics. Such extreme temperatures are likely to increase in frequency under future predicted climates. High air temperatures result in failure of peg and pod set due to lower pollen viability. The response of pollen germination and pollen tube growth to temperature was quantified in order to identify differences in pollen tolerance to temperature among 21 groundnut genotypes. Plants were grown from sowing to harvest in a poly‐tunnel under an optimum temperature of 28/22 °C (day/night). Pollen was collected at anther dehiscence and was exposed to temperatures from 10° to 47·5 °C at 2·5 °C intervals. The results showed that a modified bilinear model most accurately described the response to temperature of percentage pollen germination and maximum pollen tube length. Genotypes were found to range from most tolerant to most susceptible based on both pollen characters and membrane thermostability. Mean cardinal temperatures (T_min, T_opt and T_max) averaged over 21 genotypes were 14·1, 30·1 and 43·0 °C for percentage pollen germination and 14·6, 34·4 and 43·4 °C for maximum pollen tube length. The genotypes 55‐437, ICG 1236, TMV 2 and ICGS 11 can be grouped as tolerant to high temperature and genotypes Kadiri 3, ICGV 92116 and ICGV 92118 as susceptible genotypes, based on the cardinal temperatures. The principal component analysis identified maximum percentage pollen germination and pollen tube length of the genotypes, and T_max for the two processes as the most important pollen parameters in describing a genotypic tolerance to high temperature. The T_min and T_opt for pollen germination and tube growth, rate of pollen tube growth were less predictive in discriminating genotypes for high temperature tolerance. Genotypic differences in heat tolerance‐based on pollen response were poorly related (R² = 0·334, P = 0·006) to relative injury as determined by membrane thermostability.