Several theoretical and empirical lines of evidence indicate that root architecture (the three-dimen-
sional shape of the root system over time) may be the basis for genetic differences in P efficiency in beans
[148,172–174]. Phosphorus availability regulates many features of root architecture, including adventi-
tious rooting, aerenchyma formation, basal root elongation, basal root growth angle, lateral rooting, root
hair density, and root hair length [174].
SimRoot, an explicit geometric model of bean root growth, confirmed that root architectural traits
can influence the relationship between root C costs and P acquisition [175]. Root growth responded dy-
namically to P stress through changes in the proliferation of lateral roots and the geotropic response of
basal roots. After further research, Lynch and associates defined root gravitropism as a potentially bene-
ficial trait for P efficiency in common bean [176,177].
- Breeding for Improved Adaptation to Low Phosphorus Supply
Selection criteria to improve P efficiency in common bean can be based on physiological traits and mech-
anisms. These include the abilities to (1) mobilize P within the plant, (2) set pods and mobilize photoas-
similates to seeds, (3) minimize storage of P in seed (phytic acid), and (4) modify root architecture to ex-
ploit greater soil volume [95]. Genetic studies of P efficiency in bean showed that the P efficiency ratio
(dry matter produced per unit of tissue P content) differs among specific crosses according to parents
[137]. These studies also showed that maternal inheritance was of minor importance. Narrow-sense her-
itability estimates derived from parent offspring regression in bean families of efficient inefficient lines
were estimated to be about 40% [138], and in other studies they were high for total dry matter yield in all
families tested [178]. Studies of broad-sense heritability estimates for total dry matter yield showed that
efficiency in P use was a highly heritable trait (range 0.68 to 0.86) in bean [178,179]. Epistasis (primar-
ily, additive additive and dominance dominance gene effects) made significant contributions to the
efficiency of P use in bean [178]. Quantitative inheritance patterns and transgressive segregation for root
dry matter yields were also observed [179]. Dominance variance was more important than additive vari-
ance for P efficiency in four out of six families used in the experiments. Urrea and Singh [180] found her-
itability for seed yield under low P supply in soil to be 0.61, based on regression of F 3 populations on the
corresponding F 2 populations.
Phosphorus use efficiency has been transferred from an exotic germplasm to an adapted variety by
Schettini et al. [181] using the inbred backcross line method. They derived several tolerant lines from the
P-efficient donor (PI 206002) combined with the desirable recurrent parent ‘Sanilac’. They showed that
lines that performed well in nutrient solution culture could also perform well in a field test with soil hav-
ing a moderately deficient P supply. Saborío and Beebe [182] made efforts to breed for tolerance of low
P in soils of Costa Rica at two locations. They started with 10 segregating populations derived from
crosses with varying structures and incorporating parents from the highlands of Mexico and Peru. They
selected 14 lines and coded them as TLP (tolerant of low P) lines. Among these TLP lines, TLP 28 and
TLP 29 were superior in their adaptation to low-P soils.
Posada et al. [183] studied the heritability and mechanisms of tolerance of low-P soils in Mesoamer-
ican and Andean cultivars of common bean. They evaluated 12 parents, 6 each of Andean and Mesoamer-
ican types, and 27 of their F 2 populations for P uptake and biomass (shoot and root) production under con-
ditions of low and high P. In both parents and progeny, they observed significant differences in traits
associated with low-P tolerance, including high P uptake and efficient internal use through efficient P par-
titioning. These characteristics can therefore be used in a breeding program to improve low-P tolerance
of agronomically desirable bean cultivars.
At CIAT headquarters, more than 7000 bean germplasm accessions were evaluated for soil con-
straints, especially low P [148] (S. Beebe, personal communication). Wide differences have been found
in both P uptake and P use efficiency. Physiological studies have been combined with QTL analysis to
elucidate mechanisms of P uptake [174,184]. Recombinant inbred lines (RILs) of a mapping population
were evaluated in a greenhouse hydroponic test for a suite of traits for which the parental genotypes were
found to be contrasting. These traits were basal root number, length, and dry weight; root hair density and
coverage on the root surface; and hydrogen ion (H) exudation from roots [184] (J. Lynch, S. Beebe, and
X. Yan, unpublished results). QTLs were identified for all traits, and these were often associated with
QTLs that contributed to root length and P uptake in the field, as measured on the same RILs. Thus, P ac-
quisition reflects the interaction of several plant mechanisms. In this experiment, QTL analysis was a crit-
ically important tool for dissecting different P acquisition mechanisms.
594 RAO