Yield quantitative trait loci from wild tomato are predominately expressed by the shoot

Abstract

Plant yield is the integrated outcome of processes taking place above and below ground. To explore genetic, environmental and developmental aspects of fruit yield in tomato, we phenotyped an introgression line (IL) population derived from a cross between the cultivated tomato (Solanum lycopersicum) and a wild species (Solanum pennellii). Both homozygous and heterozygous ILs were grown in irrigated and non-irrigated fields and evaluated for six yield components. Thirteen lines displayed transgressive segregation that increased agronomic yield consistently over 2 years and defined at least 11 independent yield-improving QTL. To determine if these QTL were expressed in the shoots or the roots of the plants, we conducted field trials of reciprocally grafted ILs; out of 13 lines with an effect on yield, 10 QTL were active in the shoot and only IL8-3 showed a consistent root effect. To further examine this unusual case, we evaluated the metabolic profiles of fruits from both the homo- and heterozygous lines for IL8-3 and compared these to those obtained from the fruit of their equivalent genotypes in the root effect population. We observed that several of these metabolic QTL, like the yield QTL, were root determined; however, further studies will be required to delineate the exact mechanism mediating this effect in this specific line. The results presented here suggest that genetic variation for root traits, in comparison to that present in the shoot, represents only a minor component in the determination of tomato fruit yield.

Fig. 1

Frequency distribution of the IL and ILH means for six phenotypic traits in the dry and wet fields. Arrows on each figure indicate the mean value of M82 in the dry (D) and wet (W) fields

Fig. 2

Distributions of IL-QTL according to their mode of inheritance and direction of their effect in the dry (D) and wet (W) experiments. Each bar represents the number of QTL per trait. Above the zero line are the numbers of increasing QTL, and below are the numbers of decreasing ones. Below the bars for each trait are the numbers of IL-QTL for that trait according to their mode of expression: conserved, detected in both environments; wet specific, detected only in the wet field; dry specific, detected only in the dry field. PW plant weight; TY total yield; FW fruit weight; BX Brix; FN fruit number; BY Brix × total yield

Fig. 3

Reciprocal-grafting analysis of IL8-3 shoot and root effects on yield-related traits over four experiments. For all traits, empty bars represent a non-significant effect. Gray bars are significantly different from M82 at P < 0.05. *Significance at P < 0.01; **Significance at P < 0.001. All the values are presented as percentage difference from M82 grafted on itself. Root: root effect, shoot: shoot effect, NG: non-grafted effect. Black arrows indicate on the contrasting root effects. PW plant weight; TY total yield; FW fruit weight; BX Brix; FN fruit number; BY Brix × total yield

Fig. 4

Analysis of IL7 + 9 shoot effect on Brix × yield (BY) under the M82 and F1 rootstocks in wet and dry fields. Two-ways ANOVA for the IL7 + 9 shoot effect and the S.pennellii × M82 root effect in the wet (a) and dry (b) experiments. Each combination was tested in 20 replications under each irrigation regime

Fig. 5

Additive and non-additive interaction between shoot and root

Fig. 6

Metabolite profiles in red fruits of grafting combination of IL 8-3 scion with M82 rootstock (M82:IL 8-3) and IL 8-3. Data are normalized with respect to the mean response calculated for M82 grafted and for M82 in each case. Values presented are the mean ± SE of six replicates; values set with asterisk were determined by the t test to be significantly different (P < 0.05)