3.1. Effect of KCl and MgSO4 fertilization on leaflet K and Mg concentrations
Progenies displayed different leaflet K concentrations in response to the applied KCl levels. Treatment K1 significantly (p < 0.0001) increased leaflet K concentrations in the palms of progenies C1, C2 and C3 compared to treatment K0, whereas no significant differences were observed in leaflet K concentrations between oil palms with treatment K2 compared to the palms with treatment K1 in the same progenies. Leaflet K concentrations of progeny C4 significantly increased between treatments K0 and K1 and between treatments K1 and K2. With data pooled for all progenies, average leaflet K concentrations significantly increased from 0.715–0.829% and 0.864% of DM for K0, K1 and K2, respectively (Fig. 1A).
Without KCl application (K0) over the same period (5 to 7 YAP), progeny C4 had a leaflet K concentration (0.82%) that was 11%, 18% and 25% higher than in progenies C1, C2 and C3, respectively. Respective increases were 12%, 18% and 24%, for treatment K1 and 11%, 20% and 26% for treatment K2 (Fig. 1A).
MgSO4 applications (data pooled for all progenies) had a significant effect (p = 0.0002) on leaflet Mg concentrations. Across all progenies, average leaflet Mg concentrations in oil palms with treatment Mg2 (0.288% of DM) differed significantly (p = 0.0002) from those with treatment Mg1 (0.270% of DM) and those with treatment Mg0 (0.263% of DM), with the last two concentrations not significantly differing from one another (Fig. 1B).
With treatment Mg0, the palms of progeny C3 had a leaflet Mg concentration (0.292%) that was 13%, 18% and 10% higher than that of the palms of progenies C1, C2 and C4, respectively. Respective increases were 12%, 18%, and 13%, for treatment Mg1 and 13%, 16%, and 10%, for treatment Mg2, respectively (Fig. 1B).
Average leaflet Mg concentrations significantly (p = 0.002) decreased from 0.302–0.266% and 0.255 of DM, for treatments K0, K1 and K2, respectively. However, in all progenies there were no significant differences in leaflet Mg concentrations between treatments K1 and K2. Leaflet Mg concentrations decreased in all progenies from K0 to K1, whereas there was no significant difference between treatments K1 and K2 for none of the progenies (Fig. 1C).
Increasing MgSO4 levels had no significant effect on leaflet K concentrations in the progenies. However, leaflet K concentrations differed between progenies for each Mg level (Fig. 1D). Statistical analysis did not reveal any significant K*Mg interactions for leaflet K or Mg concentrations.
3.2. Influence of KCl fertilization on oil palm progeny yield components
Progeny C1 produced the largest number of bunches per year (14.4 ± 6.0), which was not statistically different from those of progeny C2 (14.3 ± 5.1). The palms of progeny C4 produced the lowest annual bunch number (11.4 ± 2.5), which was significantly lower than that of progenies C1 and C2, whereas progeny C3 palms had an intermediate average annual bunch number (13.6 ± 5.3), which significantly differed from that of progenies C1, C2 and C4 (Table 1).
The number of male inflorescences significantly (p < 0.0001) differed between progenies. Progeny C1 (Deli x La Mé origin) exhibited the lowest annual number of male inflorescences (6.8 ± 4.3). Progenies C2 and C3 had intermediate male inflorescence numbers (9.1 ± 4.6 and 9.5 ± 4.4), whereas progeny C4 produced the largest number of male inflorescences (15.3 ± 3).
The “bunch number to emitted frond number” ratio was 0.40, 0.37, 0.36 and 0.31 in the palms of progenies C1, C2, C3 and C4, respectively.
Progeny C3 palms had the highest average bunch weight (8.9 ± 2.8 kg bunch-1) over the 5 to 8 YAP period and an average weight of 10.7 and 11.9 kg bunch-1 at the age of 7 and 8 YAP (last two years of bunch harvesting), respectively (Table 1). On average, the palms of progenies C1 and C2 produced the lightest bunches, with bunch weights for progeny C1 that were significantly lower than those of progeny C2 (7.3 ± 1.8 and 7.7 ± 2.4 kg bunch-1, respectively). Average bunch weight for progeny C4 palms at 8 YAP (11.8 kg bunch-1) was similar to that of progeny C3 at the same age. However, at 7 YAP, the palms of progeny C4 had an average bunch weight of 8.07 kg, which was significantly lower than that of progeny C3 at the same age.
Between 5 and 8 YAP, progeny C3 (all fertilizer treatments combined) produced 15.80 t of FFB ha-1 year-1 and 5.16 t of crude palm oil (CPOLab) ha-1 year-1 on average (Table 2), which was significantly higher than the FFB and CPOLab of the palms of the other progenies, particularly those of progeny C4, which produced the lowest average yield (12.20 t of FFB ha-1 year-1 and 3.78 t of CPOLab) ha-1 year-1. The highest bunch producing progenies (C3 > C2) were also the highest oil producers, with similar oil extraction rates (Table 2). The lowest oil extraction rate (OERLab: 31.65%) was found for progeny C4, whereas progenies C2 and C3 had significantly higher extraction rates (32.46% and 32.74%, respectively). No KCl effect on the oil extraction rate was observed for any progeny (p = 0.19).
KCl treatments were associated with bunch weight increases for the palms of progenies C2 (+0.6 kg from K0 to K1; p = 0.017), C3 (+0.8 kg from K0 to K2; p = 0.0001) and C4 (+1.0 kg from K0 to K1; p < 0.0001) and bunch number increases for the palms of progeny C1 (+0.7 bunches year−1 from K0 to K2) with a significant linear contrast test result (F = 4.37; p = 0.038) over the 5 to 8 YAP period.
In none of the progenies did KCl applications have a significant effect on the average number of male inflorescences produced per palm, except at 5, 6 and 7 YAP for progenies C4, C2 and C3, respectively (Table 1).
The linear contrast test on the average bunch weight for the KCl treatment of progeny C3 was significant (F = 19.7; p < 0.0001). The quadratic contrast test was significant (F = 10.0; p = 0.002) for progeny C4, but not for progeny C2 (F = 3.2; p = 0.074).
3.3. Yield responses to KCl applications
Over the 5 to 8 YAP period, we found significant (p < 0.0000) differences in bunch production between progenies. For all KCl levels applied, annual bunch production per palm could be ranked as C3 > C2 > C1 > C4 (Fig. 2), with a difference of 25 to 30 kg palm−1 year−1 between the highest (C3) and the lowest yielding progeny (C4).
In all progenies, bunch production was significantly lower for treatment K0 than for treatments K1 and K2. Treatment K1 significantly improved bunch yields for the palms of all progenies (p = 0.0007) compared to treatment K0. For progenies C2 (p = 0.037) and C4 (p = 0.015), there was a significant increase in bunch production only between treatments K0 and K1 (Fig. 2). For progenies C1 (p = 0.045) and C3 (p = 0.003), significant increases in bunch production gains were also observed between oil palms receiving treatments K1 and K2.
3.4. Progenies’ optimum K levels
The quadratic contrast test results (F = 2.56; p = 0.111 and F = 3.11; p = 0.079, for C2 and C4, respectively) were not significant, indicating that K1 was the optimum level for these two progenies (Fig. 3). Leaflet K concentrations of palms with the K1 treatment were 0.78% and 0.95% for progenies C2 and C4, respectively.
For progenies C1 and C3, maximum yield was only reached with treatment K2 (Fig. 3). The contrast test results showed KCl response linearity (F = 6.16; p = 0.014 and F = 11.8; p = 0.000 for C1 and C3 respectively), indicating that in our experiment, K2 was the optimum KCl treatment for progenies C1 and C3 (Fig. 3). Leaflet K concentrations at K2 were 0.91% and 0.77% for progenies C1 and C3, respectively.
MgSO4 had no effect on the yield of any progeny, except for the average bunch weight of progeny C3 over the 5 to 8 YAP period (p = 0.049), with an increase of 0.35 kg bunch−1 between treatments Mg0 and Mg1.