Effect of fertilizer discount on fresh fruit bunches production and Nutrients in North Sumatra Indonesia

The low price of fresh fruit bunches a couple of years ago forced oil palm companies to make some efficiencies in the whole business process, including discounting the fertilizer expenses, which consequence in the Fresh Fruit Bunches (FFB) production declining in the first and second years onward. Reviewing the effect of fertilizer reduction on FFB production in commercial fields was challenging due to being confounded with many factors, including estate condition, infrastructure, climate, social, and management. Therefore it was better to assess from trials that tested the application of N, P, K, and Mg fertilizers with different rates to oil palm in a different locations on mineral soil in North Sumatra. Those trials illustrated that fertilizer discount would correct the FFB production accordingly from 3 up to 84%. In addition to this, climate, leaf nutrient uptake, and palm age play an important role in the decreasing FFB. An adequate water supply from the rainfall would stimulate a better leaf nutrient uptake and, therefore, palm more tolerant to fertilizer discounts. On the other hand, palm planted on the terraces, which are at a high risk of fertilizer leaching out and poor soil cation of subsoil, would be more sensitive to fertilizer discounts. Fertilizer discount, of course, would reduce the nutrient supply to the palm accordingly and consequently to FFB production for two to three years forward, therefore it was suggested to cut off the less response of specific nutrients only if possible.


Introduction
By the end of 2018, the price of Crude Palm Oil (CPO) had touched the bottom line, close to 508 USD per metric ton, which is the lowest price observed during the last decades, as depicted in Figure 1.This situation caused commodity got losses in regard to profitability.For this reason, many oil palm plantation companies and smallholders made some efficiencies, including reducing the fertilizer input in 2019, because the latter contributed about 35% of total production cost [10] and [17].This decision of reduction or even cessation of fertilizer would have a negative impact on FFB production in 2021 and 2022, because, according to [1], the development of an oil palm inflorescence between the stage of initiation and flower maturity last 2-3 years for both sexes, the process is initiated soon after seeding establishment.Furthermore, [14] converted a long-term NPK factorial fertilizer trial established on a highly leached and infertile soil into a fertilizer withdrawal showed that cessation would impact nutrient supply for about 12 months for K and 24 months for N, P, and Mg onward and consequently to FFB about 30 to 44 months onward under 14 years old palm.In addition to this [3], Figure 2 below shows that oil palm productivity In 2020 until 2022 declined by about 4% and 2%, respectively, relative to 2019 productivity.There were many factors involved in causing FFB stagnation, including: inadequacy of harvesters in the plantation industry, the adverse climate during lower rainfall periods, i.e.El Nino, and the effect of reduced fertilizer input or total cessation of fertilizer applications during lower CPO prices [13].However, it was difficult to review the effect of fertilizer reduction on FFB production in the commercial fields due to being confounded with many factors, including estate condition, infrastructure, climate, social, and management.Therefore, to investigate the palm response to fertilizer reduction would be better from the four series of long-term of NPKMg fertilizer trials in North Sumatra.This paper illustrated a snapshot effect of fertilizer reduction on FFB production for young and old palms.

Material and Method
Four fertilizer trials on oil palm (Deli x AVROS material; 128 stand ha -1 , ex rubber plantation) planted in 1985 were set up at different in North Sumatra sites.The site was derived from volcanic Toba acid tuff and sandstones on a range, of topography and rainfall (Table 1).Fertilizer treatments and records last from 1992 until 2009.However, this paper only showed the data set of 1994-1996 to represent young palm of 8-10 years old and data set of 2007-2009 to represent the old palm 21-23 years old.In addition to this, during 1997-2006, the four fertilizer trials were managed under similar conditions.All the trials were factorial in design with single replicates of 3 4 NPKMg (Trial 231 and 232) and 3 2 NP x 2 2 KMg (Trial 275 and 277).Plot sizes ranged from 36 to 49 plot palms of which the central 9 to 16 palms were regularly recorded for FFB production and nutrient content of leaf and rachis (N, P, K, Mg, and Ca).N fertilizer treatments were applied as Urea at annual rates of 1-4 kg palm -1 in factorial combination with 0-4 kg TSP (20% P), 0-5 kg MOP (50% K), and 0-2 kg Kieserite (16% Mg) as depicted in Table 2.The plots were surrounded by a 1meter-deep trench, which was regularly maintained.Based on the four series of fertilizer trials, it was possible to review the impact of fertilizer discounts on FFB production.The response of FFB and leaf nutrient status to additional fertilizer was supposed to be equal to the reduction of the fertilizer.The fitted FFB data for specific fertilizer combinations of young palm and old palm from four series of fertilizer trials were used to review the response of both FFB and leaf nutrients value to fertilizer discount as shown in Table 3.The fitted FFB data showed a specific fertilizer combination, and its nutrient value, which was arranged from the highest to lowest FFB yield systematically (Table 5-6).

Effect of fertilizer discount on the fresh fruit bunches production.
The FFB data on young and old palm response to fertilizer reduction from 0-100% of four fertilizer trials were depicted in Figure 3-6.It was clear that the impact of the fertilizer discount on FFB was much more substantial on the old palm rather than the young palm in all trials, because the rate of leaf production in which inflorescence was initiated in the axil of each leaf of the palm varies with age, with typically two leaves produced per month in the case of older individual palm and three leaves produce per months for the young palm [1] , consequently, the old palm produce less number of bunches even though it had a greater bunch weight, and on the contrary, the young palm is more energetic in producing more bunches.So the old palm is more sensitive to fertilizer discounts.Figure 3-4 depicted that the FFB yield response to the fertilizer discount on trials 275 and 277 was quite similar, probably due to both sites having similar environmental properties as shown in Table 1, including flat topography, good soil reserve of high exchangeable-K from Rhyiolite soil parent material, moderate level of N total in the soil and rainfall intensity about 1700 mm per year with 200 mm water deficit per year.
Fertilizer discount up to 50% depressed FFB by about 5-10% for young palm and 15-20% for old palm, furthermore stoppage fertilizer depressed FFB to a more severe level of about 25% for young palm and 35% for old palm.
Similar to the trials 275 and 277, the FFB response to fertilizer discount of 50% on trials 231 (Figure 5), situated on high rainfall and flat topography, depressed FFB by about 8% and 12% for young and old palm, respectively.Interestingly stoppage of fertilizer only depressed FFB by about 23% both for young and old palm.This finding was significantly difference from trials 275 and 277, where FFB production declined up to 35% when fertilizers were stopped.Probably due to adequate water supply, as indicated by no water deficit accident and a high organic matter in trial 231, as shown in Table 1, those conditions improved leaf N and P status higher compared with the other site trials, as depicted in Table 4. Table 4 also showed that the leaf levels of N and K, which play an essential roles in driving FFB production in trial 231, was not declined below the minimum critical level at cessation fertilizer, probably due to a good supply of Nitrogen From organic matter, high K-exchangeable from rhyolite soil parent material and adequate water supply throughout the year.That situation explained why the yield decline was not as severe as the other trial, and the yield decline of the old palm was quite similar to the young palm.
On the contrary, the FFB response to fertilizer discount on trial 232, as depicted in Figure 6, was Severe, probably due to the steep slope and sandy texture soil, which contribute to the low efficiency of fertilizer application [7].In addition to this, sand stone soil parent material also had poor N total and K exchangeable in the soil.Fertilizer discount of up 50% decreased FFB by about 20% for both young and old palms.Further more cessation fertilizer depressed FFB by about 30% and 80% for young and old palms, respectively.Table 4 showed that the nutrient leaf level at frond #17 in the fertilizer cessation plot on trial 232 was the most severe nutrient deficiency, where leaf levels of N, P, and K were droped below the minimum critical level, especially on the old palm; therefore this trial had the most severe FFB yield declined compared with the other trials.In addition, [13] also revealed that in past experiments carried out during the 1970s, 1980s and 1990s over expansive range of soils and climate regions, cessation of fertilizer could result in the FFB yield dropping over 40%, which translates to about 13-14,5 ton FFB per hectare per year.Bold with Underline = defeciency of leaf nutrient.
Table 4 showed that 100% fertilizer discount depressed N and P leaf levels below the minimum recommended value except for N at trial 231 on young and old palm; this, finding explained that trial 231 had a better buffer fertility zone that was able to minimized the FFB declining when fertilizer was stopped, that buffer fertility was supported by a high content of organic matter and adequate water supplied (Table 1).Surprisingly K leaf level was not declined below the minimum recommended value in all trial sites except on sandstone parent material (232), probably due to soil reserve K from Rhyolite parent material fulfilled the K demand for the young and old palm.The recommended minimum nutrient value was shown in Appendix A.
Tables 5 and 6 depict the response of fertilizer discount to FFB yield and nutrient status of the leaf and rachis of the young and old palm.Three of four trials (277, 231, and 232) demonstrated that the fertilizer discount impacted substantially depressed FFB production of the young palm by 16%, 12%, and 21%.respectively, and for the old palm by 21%, 13% and, 35% respectively when K fertilizer was a stoppage.The uptake of K in the rachis of those three trials (277, 231, and, 232) for the young and old palm declined markedly.However, the leaf K level was not down to below the minimum recommended value in the absence of K fertilizer except in the trial on sandstone soil (232), probably due to the positive effect of N in the leaf, which stimulated the translocation of K from the rachis into the leaf especially in trial 231 as mentioned by [7] that N fertilizer stimulated translocation of nutrients from the reserve tissue to the leaf tissue, it was essential to making fertilizer recommendation base on both leaf and rachis nutrients levels.Furthermore, [19] revealed that in the absence of K fertilizer (but at the optimal level of all other fertilizer), leaf and rachis K levels declined markedly.In addition to this, cessation of K fertilizer depressed N and P leaf nutrients levels below the minimum critical level in all trials.One of the four trials (275), which was derived from volcanic ash soil rich in K from rhyolite parent material, showed no FFB yield response to K fertilizer for the young palm and moderate FFB yield response for the old palm, in agreement with the result of [18] that no response to K fertilizer on rhyolitic soil when reserving K > 4 meq/100g.Furthermore stoppage of P fertilizer in trial 275 significantly depressed FFB yield by about 19% relative to the control plot (0000) for the young palm, and the stoppage of N and P fertilizer in trial 275 depressed FFB yield about 17% for the old palm, it was indicated that N and P play the most critical role in this specific-site (trial 275).
[8] also found the most frequent FFB yield response to potash in all soils tested in West Malaysia, furthermore [2] showed that MOP application has the most substantial impact in terms of yield improvement (42,3 kg FFB/palm/year), [12] also reported application of K fertilizer as MOP at 2 kg/palm /year significantly increased number of bunches and bunch weight and consequently FFB compared with control, which received no K-fertilizer.Thus [7] revealed that K fertilizer had the most excellent fertilizer nutrient for the first 1 kg per palm fertilizer nutrient applied, recovery efficiency was generally greater than 80%, 70%, and 40% for K2O, N and P2O5 respectively, in situation where no major site limitation.However at site with steep slopes or high rainfall the recovery of the first 1kg per palm fertilizer was about 50-60% , 40%, and 30% for K2O, N and P2O5, respectively.In a situation where no major site limitation.However, at sites with steep slopes or high rainfall, the recovery of the first 1kg per palm fertilizer was about 50-60%, 40%, and 30% for K2O, N and P2O5, respectively.
While N fertilizer plays an essential role in trial 275 for both the young and old palm and trial 232 on the old palm, however, it did not mean that the other trials required no N fertilizer; all trials still need N to stimulate the other nutrient uptake and yield.
[15] revealed that N uptake has a positive synergic relationship with P and K and then occasionally with Mg, where one increased the other uptake in all trials.It can be understood that in some cases, palm deficiency of specific, nutrients should not necessarily cure with an the application of corresponding nutrient sources.Furthermore, of the three major nutrients, P seems not as crucial as K and N due to lower fertilizer nutrient recovery, presumably due to the high P buffering capacity of the soils [15] .The last was Mg which is a complementary fertilizer to obtain the optimum FFB, as showed in Tables 5 and 6.
Furthermore, [14] also found that the withdrawal of fertilizer caused a negative impact on FFB production at the different time frames.N fertilizer was going to depress FFB in the third year, commencing at 30 months after ceasing fertilizer, while K and P fertilizer were going to depress FFB in the fourth year commencing 39 and 44 months after ceasing fertilizer, respectively.Complete stoppage of all fertilizer (N, P, K) simultaneously had a similar time frame with N fertilizer.However, [11] showed that the reducing K fertilizer resulted in a reduction in the bunch weight of oil palm by up to 30%, and the number of bunches by about 10% in the second year after application and [9] reported the yield response to Urea of 17 years old palm was only 1% and 8% in the first and second years after application respectively, only data from the third year the yield response achieved 18%.
The above Tables also depicted that the optimum fertilizer rate was about 6-8 kg per palm annually, both for young and old palms, according to the nutrient status both in the leaf and rachis, attributed to the sufficient water supply, favorable climate conditions, and indigenous nutrient supply.The increase in fertilizer rate will eventually lead to a decrease in nutrient recovery efficiency [5].
Therefore, in complex economic situation, discount fertilizer must be very selective of specific nutrients, which had a minor impact on FFB production in the specific site and considering the leaf nutrients status of the palm.Tables 5 and 6 showed that Mg fertilizer was the priority to be cut off due to having the slightest impact on FFB yield, then P fertilizer in the second priority; the latter had a long residual effect due to high P buffering capacity in the soil except for specific site which has strong response to P fertilizer like trial 275.[6] found that reduction of Rock Phosphate's frequency application with the same rate from Annually to once every four years did not depress leaf P level significantly but depressed FFB yield on average of 2,6 tons / ha/ year.[2] suggested rationing fertilizer must be very selective in order to reduce as far as possible any negative impact on the future performance of the palm.It has to be based on the result of field trials and taking account palm status and the past of fertilizer regime.

Conclusion
1. Fertilizer discount certainly depressed the level of N and P in the leaf drop below the minimum recommended values in all; furthermore it was also depressing the K in the rachis markedly, consequently depressed fresh fruit bunches production by about 20-38% for young palm 35-80% for old palm.2. The thoughtful limiting factor, including palm age, water stress, steep slope, and coarse texture, trigger a worse FFB decline.3. The four NPKMg fertilizer trials indicated that the most critical nutrient was K, N, P, and Mg respectively; therefore; in difficult economic situations, it was suggested to consider the nutrient status and the less responsive fertilizer according to the site-specific.4. The more significant discount would impact the worse FFB yield in the next two and three years.

Figure 3 .
Figure 3. FFB's response to fertilizer discounts both for young and old palm in trial 275.

Figure 4 .
Figure 4. FFB response to fertilizer discount both for young and old palms in trial 277.

Figure 5 .
Figure 5. FFB's response to fertilizer discounts both for young and old palm in trial 231.

Figure 6 .
Figure 6.FFB response to fertilizer discount both for young and old palms in trial 232.

Figure 7 .
Figure 7. Number of the bunches (NOB) and average bunch weight (ABW) response to fertilizer discount for young and old palm in trial 275.

Figure 8 .
Figure 8. Number of the bunches (NOB) and average bunch weight (ABW) response to fertilizer discount for young and old palm in trial 277.

Figure 9 .
Figure 9. Number of the bunches (NOB) and average bunch weight (ABW) response to fertilizer discount for young and old palm in trial 231.

Figure 10 .
Figure 10.Number of the bunches (NOB) and average bunch weight (ABW) response to fertilizer discount for young and old palm in trial 232.

Appendix A .
The recommended minimum and maximum values in trials are275,277,231,232.

Table 1 .
General characteristics of the trial sites in North Sumatra.

Table 2 .
Treatment of four NPKMg fertilizer trials.

Table 3 .
Treatment of fertilizer discount of young and old palm in the four trials.

Table 4 .
Nutrient status at the cessation of fertilizers (control plot).

Table 5 .
The effect of specific fertilizer reduction on FFB yield and nutrient level in the leaf and rachis on the young palm.