Towards Sustainable Oil Palm Plantation Management: Effects of Plantation Age and Soil Parent Material

This study was conducted to generate information required to guide plantation management in relation to replanting on soils derived from different parent materials. Six oil palm estates in coastal lowlands of southwest Cameroon were considered. Oil palm yield data (in t·ha −1 of fresh fruit bunch, FFB) and corresponding age of palms (in years after planting, YAP) were obtained for the various estates. In all the estates, average yields were <10 t·FFB·ha −1 and highly variable. Plantation age, solely, explained between 20% - 58% of the variation in yield. The highest average yields (11.5 t·FFB·ha −1 ) were obtained in plantations aged between 9 and 18 YAP and the lowest (4.66 t·FFB·ha −1 ) were obtained in old plantations (>23 YAP). Plantations located on volcanic parent materials generally had higher yields compared to those established on sedimentary parent materials. In order to intensify production and increase yields while conserving the environment, one important measure to consider is the replacement of aged palms, and the recommended optimal replanting age in coastal plains of southwest Cameroon should be at most 23 YAP. Estimated mean yields, if aged palms are replanted on time, can increase by 43% - 65%. Additionally, site-specific nutrient management options should be considered in plantation intensification programs.

digenous to West Africa and happens to be the most productive vegetable oil crop. Annual global production is more than 35 million metric tons of palm oil with Malaysia and Indonesia contributing >80% of global production [1]. During the past decade, the total area under production was about 14 million hectares [2]. Between 2000 and 2012, the global oil palm planting area increased from 10 to 17 million hectares [3], following the conversion of most tropical forests, pastures and peatlands. Consequently, global oil palm expansion has been a major driver of deforestation, severe loss of biodiversity [4] [5] and greenhouse gas emissions [6]. Additionally, ecosystem functions (such as soil erosion prevention, water regulation and water supply, climate regulation and mitigation of extreme events, etc.) are greatly reduced following the conversion of natural forests to oil palm plantations [7]. The rapid expansion of oil palm planting areas and the high quantities of oil palm production have been due to increased demand of palm oil for food, industrial transformations, and biofuel production [8], and also because it is the least expensive of all vegetable oils. The oil palm is the most productive oil crop in terms of oil yield per hectare and resource use efficiency due to its high efficiency at transforming solar energy into vegetable oil [9]. The average yield of palm oil is approximately 4.2 t·ha −1 ·yr −1 , with yields exceeding 6.0 t·ha −1 ·yr −1 in the best-managed plantations. Oil palm yields are far greater than those of vegetable oils such as rapeseed and soybean that produce only 1.2 and 0.4 t·ha −1 ·yr −1 , respectively [10]. However, optimal oil palm yields are hardly achieved in both smallholder farms and agro-industrial estates due to several factors that cause yield gaps. Although Cameroon has very favourable environmental conditions (climate and soil) for achieving optimum oil palm production, actual yields are far below the potential yields. For example, as of 2012, Cameroon recorded a total production of 230,000 t of crude palm oil (CPO) from approximately 190,000 ha exploited by both agro-industries and small/medium holdings [11], giving approximately 2.6 t·CPO·ha −1 . Considering smallholders only, oil palm yields hardly exceed 1.0 t CPO·ha −1 ·yr −1 . It has also been reported that agro-industries in Cameroon such as the Cameroon Development Corporation (CDC) and PAMOL plantation Plc hardly produce 2 t·CPO·ha −1 ·yr −1 meanwhile smallholders in Indonesia reach much better yields of 3.0 to 3.5 t·CPO·ha −1 ·yr −1 [12]. The low yields in oil palm have rendered Cameroon a net importer of palm oil due to the high national demand, mostly for industrial transformation (especially in soap production). From 1999 to 2004, the import of vegetable oils increased by 366% with palm oil accounting for 55% of the imports [13]. Again, imported CPO in Cameroon increased from 16,000 tons in 2016 to 95,000 tons in 2017 [14].
Faced with the challenge of achieving optimum oil palm yields, it is imperative to identify factors causing the yield gaps in order to generate information required to guide on appropriate management strategies that will increase oil palm yields and sustain production. According to Ruiz et al. [15], a gap in crop yield is defined as the difference between yields obtained by producers in their  [16]. Different factors create yield gaps in oil palm, including climate constraints, soil constraints, planting materials and their characteristics, and adoption of different technologies [15]. An extensive review of factors contributing to oil palm yield gaps has been done by Woittiez et al. [17], wherein four production levels have been distinguished; the potential, water-limited, nutrient-limited, and the actual yield. According to Woittiez et al. [17], closing these yield gaps could increase global production by 15 -20 Mt·oil·yr −1 , which would limit the drive for further area expansion at a global scale, and hence limit the negative consequences of deforestation.
In Cameroon, many factors contribute to declining oil palm yields, including climatic variations [18] [19] [20], inappropriate plantation management, limited inputs (fertilizers), labour shortage, pests and diseases [11], and the continued exploitation of old plantations (>25 years) [21].  [26] reported that the optimum age of replanting corresponds to the time when the marginal net revenue in year of the present stand is equal to or exceeds the amortized value of net revenue in year of the second stand. Thus, the optimum replanting age depends on the price of FFB, cost to establish and maintain new palms, technology that changes the yield profile, and the discount rates adopted.
Although the useful life of a palm tree with respect to fresh fruit bunch (FFB) production has been established by different authors, a common practice in most oil palm plantations and small farm holdings in Cameroon is the non respect of age limit required for renewing aged plantations. This practice is more pronounced on soils with high fertility status. As a result, FFB yields are low and continue to decline even on volcanic soils which are generally fertile. Increase in oil palm production is now synonymous with the opening of new farmlands despite numerous environmental consequences associated with it. Additionally, there is an existing myth that very old palms can be productive and profitable on fertile soils. The objective of this study was to generate information required to guide plantation management in relation to replanting on soils derived from different parent materials. More specifically, the study aimed to identify the age range of plantations, to establish age-FFB yield relationships and to ascertain the age limit of a palm tree for optimum productivity.

Location and Description of Study Area
The study was carried out in the lowland plains of Southwest region of Cameroon, a region constituting the major oil palm growing areas of the country.
Considering climatic and soil suitability, the area is in general suitable for oil palm cultivation and growth [27] [28] and can thus be considered a representa- allowed to be carried out, will lead to severe loss of forest and biodiversity. According to the Koppen climate classification, the area has the equatorial climate type (Am), precisely the Cameroon type, which is quite hot and humid, characterized by the existence of two distinct seasons-one wet (rainy) season (March

Source of Data and Statistical Analysis
Due to lack of reliability of smallholder oil palm yield data, yield data (measured in t·ha −1 of FFB) and corresponding age of plantations (in years after planting, YAP) was obtained for different oil palm estates of the Cameroon Development Corporation (CDC) in the Southwest region of Cameroon. The estates were Mondoni, Debundscha, Bota, Mungo, Idenau and Illoani-Boa. Yield data was mostly recorded on a monthly basis. The data included yields recorded from 2005 to 2017. The data was subjected to descriptive analysis in order to appreciate mean values, range in yields, variation in yields, etc. For a particular study site, simple and polynomial regression analysis was performed to appreciate the relationship existing between yield and plantation age. Correlation analysis was performed to measure the strength of the relationship existing between yield and age. In order to further appreciate the relationship existing between yield and age, yield data was grouped according to different age classes as defined by [25]. These classes included: young trees (3 -8 YAP), prime (9 -18 YAP), ageing (19 -23 YAP) and old (>23 YAP). Variability in oil palm yield within each of the classes was evaluated using the coefficient of variation (CV). Various oil palm plantations were grouped according to the dominant soil parent materials (mainly sedimentary and volcanic) in order to investigate the influence of parent material on oil palm yield.

Estimation of Oil Palm Yield Gaps
Yield gaps in the various sites were quantified as the ratio of real or actual yield obtained with respect to potential yield, expressed in percentage (Equation (1)) [15].
Actual real yield in of FFB ha Yield gap % = 100 Potential yield in of FFB ha Potential yield in southwest Cameroon was obtained from Feintrenie et al. [31]. According to the latter, taking into consideration climatic, topographic, pedological and management (conventional cropping practices and required fertilization, etc.) conditions of the entire study area (Southwest region of Cameroon), potential yields are estimated between 12 (minimum) and 25 (maximum) t·ha −1 ·yr −1 . Rating of oil palm yields in southwest Cameroon was established as follows: Low (<12 t·ha −1 ), Intermediate (12 -25 t·ha −1 ) and High (>25 t·ha −1 ).

Relationship between Oil Palm Yield and Plantation Age
Descriptive statistics of oil palm yield data and age are shown in Results of correlation analysis showed that in all sites, there was a significant correlation between FFB yield and age (Table 3). There was a significant negative correlation between tree age and yield in Mondoni, Bota, Mungo and Idenau    The relationships between age and yield ( Figure 2) revealed that plantation age could explain between 20.6% and 58.0% of the variation in oil palm yield.
Woittiez et al. [32] investigated the influence of plantation age (measured in YAP), location (contrasting soil types) and tissue nutrient concentrations in leaf (N, P, K, Mg, and Ca) on vegetative growth and oil palm yield. In their results, a full model explained 64% of the variation in vegetative growth meanwhile 56% of the variation was explained by location and palm age only, with palm age having a significant negative effect on yield.
These results are in concordance with the observations reported by Euler et al. [33]. However, the yields in various classes are very low compared to the potential yields. In all sites, actual yields are far smaller than the minimum potential yield of 12 t·ha −1 .

Estimated Yield Gaps in Oil Palm
The average yield gap for different estates is shown in Table 5, while the yield gaps in the different age classes are shown in Table 6. In general, yields in all es- of prime plantations (6 -10 years) had high yields (>33 t·ha −1 ), 43% had intermediate yields (21 -33 t·ha −1 ) and only 7% had low yields (<21 t·ha −1 ). As concerns ageing plantations, Ruiz et al. [15] reported that more than 50% of ageing plantations (>16 years) had high yields (>33 t·ha −1 ).

Influence of Soil Parent Material on Oil Palm Yield
The relationship between oil palm yield and plantation age grouped under volcanic and sedimentary parent materials is shown in Figure 4. Mean yields in the two areas were significantly different at the 95% confidence interval (t = 3.704, p < 0.05). The average yield on volcanic parent materials is 7.94 ± 0.32 t·ha −1 (standard deviation = 4.79 t·ha −1 ), while the average yield in plantations located on sedimentary parent materials is 7.43 ± 0.32 t·ha −1 (standard deviation = 4.30 t·ha −1 ). In both sites, there was a significant negative correlation between plantation age and yield at the 99% confidence interval (r = −0.379** for plantations on volcanic parent material and r = −0.287** for plantations on sedimentary parent material), indicating that yields decrease with increase in plantation age. In plantations located on sedimentary parent materials, plantation age explained more than 40% in the yield variation, compared to about 24% in sites with volcanic parent materials, following a quadratic model. Coefficient of variation was very high for yields obtained in both sites (>50%).    As concerns mean yields obtained in different age classes from the two locations, it was observed that except for old plantations, all age classes (from young to ageing) had higher yields in plantations located on volcanic parent materials compared to those on sedimentary parent materials. The difference in mean yields was however significant at the 95% confidence interval only in the young phase (t = 3.724, p = 0.000). The prime phases in both areas had medium coefficients of variation (between 15% and 35%), indicating that yields are more stable during this phase of oil palm growth compared to other phases. In general, the coefficient of variation was high (>35%) in all phases of plantations established on volcanic and sedimentary parent materials. Normality plots of FFB yield in the various age classes showed that the data was not normally distributed in the young and old phases ( Figure 5 and Figure 6). The median yields in all age  (Table 7).

Influence of Plantation Age and Soil Parent Material on Oil Palm Yield
The high yield gaps observed in oil palm plantations of Cameroon are due to a combination of different factors such as inappropriate soil fertility management, the impact of pests and diseases, and other factors among which plantation age contributes from 20% to 50% in yield variation. In Andhra Pradesh (India), Latha et al. [34] observed that 50% of 5 years old plantations produced 10 -15 t·ha −1 of FFB, while 78% of 9 years old plantation produced more than 20 t·ha −1 and in 24 years old plantation, 66% of palms recorded more than 30 t·ha −1 of FFB. The low yields and large yield gaps observed in southwest Cameroon compared to other parts of the world such as Malaysia and Indonesia is a clear indication that oil palm production needs to be intensified in order to produce acceptable yields. The first measure to intensify oil palm production and limit deforestation should be the replacement of old palms. From Figure 2 and Figure   3, it is suggested that, based on yields observed at various ages, the oil palm replanting age should be 23 YAP. With respect to the influence of plantation age on oil palm yield (over a 25-year period), a comparison of our results with those from other studies (as reviewed by [17]) indicates that mean actual yields in southwest Cameroon are about 2 -3 times lower than actual yields elsewhere and about 4 times lower than nutrient/water limited yields and potential yields. Soil physical properties (such as texture, structure, drainage, porosity, effective soil depth) and chemical properties (such as soil pH, C content, N content, P content, CEC, base saturation and exchangeable bases) as influenced by parent material, have a direct influence on plantation management and yield [35].
Woittiez et al. [17] also reported that oil palm growth and yield greatly depends on the potential of the soil parent material. Apart from the influence of soil type on yield, the areas with volcanic parent materials also witness a well distributed rainfall pattern throughout the year with mean annual rainfall exceeding 8000 mm. It is therefore evident that in addition to plantation age, soils developed from volcanic parent materials, alongside high amounts of rainfall favour oil palm yields. According to [17], soil type (parent material) and rainfall are major contributing factors to the water-limited yield in oil palm. Although we did not measure the soil water holding capacity of the soils, it appears that the plantations located on sedimentary parent materials (mostly alluvial sediments) suffer from impeded drainage due to high water table at some periods of the year (field observation), especially in the Boa plain (Illoani-Boa estates). Henson et al. [36] observed that impeded drainage as caused by high water tables can significantly reduce oil palm yields (20% -30% yield loss) by affecting net photosynthesis, stomata conductance and transpiration. Lee and Ong [37] observed that water logging in low-lying areas can cause the death of up to 75% of immature palms.
The lateral movement of soil water as influenced by a high water table can also influence plant nutrient availability. On the other hand, the plantations located on volcanic parent materials have soils with good drainage, which favours the growth of palms. Based on the type of parent material, specific measures such as water drainage can be used to improve on the soil condition so as to favour oil palm growth and productivity. According to Chuah and Lim [38], drainage of frequently flooded plantations can increase fresh fruit yield by >5 t·ha −1 .

Closing Yield Gaps through Oil Palm Replanting
Estimated yield recovery through oil palm replanting (assuming similar management type) is shown in  [39] reported that one of the immediate strategies to increase oil palm yields in Malaysia is through the identification of the optimal age profile and then the establishment and practice of a replanting program. With yield gaps ranging between 1.0 and 2.4 t·ha −1 ·yr −1 of palm oil observed in Malaysia, one of the best measures adopted to close the yield gaps has been the replanting of more than 200,000 ha of palms exceeding  lanted such that 15% of the land is covered with the immature phase (<3 YAP) at any one time [40]. Recently in Indonesia, the government has launched huge replanting programs as a measure for intensifying oil palm yield while conserving the environment. Afriyanti et al. [41] reported that although oil palm replanting with good planting materials is a sure means of reaching potential yields (27 -38 t·FFB·ha −1 ) in Indonesia, this may not be the case with smallholders due to the very high costs required for replanting, which they cannot afford. Notwithstanding, replanting proves to be an effective means of increasing oil palm yields without necessarily increasing the cultivation area.

Conclusion
This study investigated the quantitative relationship between oil palm age and yield, and the influence of location (parent material) on oil palm yield. The results indicated that oil palm age has a significant effect on yield and accounts for at least 20% in the variation of yield. Additionally, there was also a significant difference in oil palm yields within the different locations, with the highest yields obtained on volcanic parent materials. The very low yields (<12 t·FFB·ha −1 ) and the large yield gaps observed necessitate urgent measures for plantation intensification in order to increase yields and sustain production.