Genetic and geographic variation in growth of Balanites aegyptiaca in Niger: comparing results from provenance/progeny tests in the nursery and field

Some tree improvement programs in Africa use nursery tests to investigate genetic and geographic variation in growth, but do they lead to the same conclusions as field tests? We investigated this question using provenance/progeny tests (12 provenances, 108 families) of Balanites aegyptiaca from semi-arid Niger. The nursery test included treatments for time (12–16 months) and watering regime (reduced and normal). Family variation was significant for shoot diameter (Sdia), height (Sht), dry weight (Swt) and root dry weight (Rwt) in the nursery, and for tree height at 1 and 2 years (Fht-1, Fht-2) but not at 13.5 years (Fht-13.5) in the field. Provenance variation was significant only for root/shoot weight ratio (RSwt) in the nursery. Family mean Fht-1 and Fht-2 were positively correlated with all nursery growth variables except RSwt. Provenance mean Fht-2 was positively correlated with Sht, while provenance mean Fht-13.5 was negatively correlated with Swt and positively correlated with RSwt. Family mean survival at 13.5 years was positively correlated with Swt, Sdia, Fht-1 and Fht-2. Family mean Rwt, RSwt and Fht-2 increased from the more humid western to the drier eastern locations. Most correlations with nursery growth variables were stronger at 16 than at 12 months and in the normal compared with the reduced watering regime. Results suggest that Swt and Sdia may be useful for predicting family survival in the field, and Rwt and RSwt may be useful for investigating geographic variation in growth in the field. We recommend conducting both nursery and long-term field tests.