Changes in the Root Architecture of Oil Palm Seedlings in Response to Nitrogen Starvation

: Oil palm ( Elaeis guineensis ) is a widely cultivated crop known for its high oil yield. It is cultivated extensively across tropical regions, notably in Southeast Asia, Africa, and Latin America. It plays a vital role in global vegetable oil supply, meeting approximately 35% of the world’s demand. However, the expansion of oil palm plantations often involves the utilization of degraded soils where nutrient availability, particularly nitrogen, is limited, posing challenges to plant growth and productivity. Roots are crucial in responding to nitrogen deficiency by adjusting their growth and distribution; however, research on root system distribution patterns in oil palm still needs to be completed. This study analyzes the root system architecture using RhizoVision Explorer, a 2D root image processing software while assessing its relationship with nitrogen availability across two commercial cultivars: Deli × La M é of African oil palm ( Elaeis guineensis ) and the interspecific hybrid O × G Coari × La M é . Our findings reveal significant associations between eight root traits in oil palm seedlings and treatments with and without nitrogen availability. Notably, total root morphology (length, surface area, and volume), rotation angle, solidity, and hole characteristics decreased under nitrogen deprivation, whereas surface angle frequency increased. We highlight the variability of these traits across cultivars, suggesting genetic dependence and potential utility in breeding programs. Moreover, interactions observed in primary root morphology and hole size indicate greater differences between control and nitrogen-treated groups in C × LM than in D × LM cultivars. On the other hand, cultivar differences, regardless of nitrogen availability, influenced lateral root morphology, while nitrogen availability, irrespective of cultivar, affected inclined angle frequency. Significant differences were observed in growth and development parameters such as root and shoot biomass, root-to-shoot ratio, and leaf emission numbers between nitrogen-optimal and nitrogen-starved conditions. Nitrogen significantly affects root architecture and plant growth in oil palm, particularly in the C × LM cultivar during the nursery stage.


Introduction
Oil palm (Elaeis guineensis Jacq.) is a vital crop renowned for its high oil yield, making it one of the most significant sources of edible oil globally.Originating from West Africa, it is extensively cultivated across tropical regions, particularly in Southeast Asia, Africa, and Latin America [1].With annual production close to 90 million metric tons, oil palm cultivation contributes significantly to the global vegetable oil supply, meeting approximately 35% of the world's demand.The versatility of palm oil extends beyond its culinary applications; it serves as a key ingredient in various consumer products, including food, cosmetics, and biofuels [2].Moreover, oil palm cultivation is crucial to socioeconomic development, providing livelihoods for millions of smallholder farmers and supporting the economies of palm-growing regions [3].
Oil palm demonstrates remarkable adaptability to various soil conditions, including degraded soils, making it a promising candidate for land rehabilitation efforts.Studies have shown that oil palm can thrive on soils with low fertility and high acidity, commonly associated with degraded lands resulting from deforestation, mining, or other anthropogenic activities [4].The extensive root system of oil palm enables efficient nutrient uptake, facilitating growth even in nutrient-poor soils.Additionally, the ability of oil palms to form symbiotic relationships with beneficial microorganisms further enhances their resilience to adverse soil conditions [5,6].
While oil palm exhibits a notable tolerance to poor acidic soils, several limiting factors can hinder its growth and productivity.One primary challenge is the reduced availability of essential nutrients, particularly nitrogen, phosphorus, and potassium, which are crucial for plant growth and development [7].Acidic soils often exhibit decreased nutrient solubility and increased nutrient leaching, leading to deficiencies that can negatively impact oil palm performance [8].Furthermore, the elevated levels of aluminum and manganese in acidic soils can be toxic to oil palm roots, impairing their ability to absorb nutrients and water efficiently [9,10].These adverse soil conditions may also affect soil microbial populations, disrupting important symbiotic relationships that aid nutrient uptake and plant health.
Oil palm cultivation represents an alternative option for rehabilitating degraded land and preserving biodiversity [11].Despite the diverse soil conditions of Colombia, including acidic soils with lower nutrient availability (pH 5.5-4.5), the expansion of oil palm cultivation has not been impeded, with nearly 580,000 hectares cultivated, many on degraded or acidic soils [12] and under varied climatic, soil, and biotic stress conditions.
The high yield of fresh fruit bunches (FFB) in oil palm requires significant chemical fertilizers, constituting approximately 40% of variable costs [13], with nitrogen fertilizers being predominantly utilized due to their role as a limiting element for agricultural production systems [14].
The root is the primary organ involved in root-soil interactions.It plays an important role in nitrogen acquisition, particularly under limited conditions.However, due to the complexity of phenotyping, it still needs to be studied.In oil palm seedlings, the root system comprises mainly vertical primary roots accompanied by lateral roots.The proportions and distribution vary over time [15].
Recently developed computational software, such as GiARoots, REST, DIRT, RHIZO, and RhizoVision Explorer, facilitate easy and detailed root architecture phenotyping [16].Among these tools, RhizoVision Explorer stands out for its user-friendly interface, speed, cost-free accessibility, batch analysis capabilities, and whole mode analysis, which enable the extraction of numerous features, especially for intact roots.Moreover, its accuracy has been demonstrated in recent studies [17].
Previous research [18] has examined the effects of varying nitrogen concentrations on oil palm seedling performance and physiological responses.However, there is a notable research gap related to root morphology or architecture in oil palm seedlings under nitrogen starvation.This gap may be attributed to unfamiliar methodologies assessing biometric parameters associated with root changes.The implication of knowing root morphology and architecture is critical in its application in crop management.For example, Pinto Gloria et al. [19] noted that oil palm cultivars with smaller root systems may be more severely affected by the fatal yellowing (FY) disease.
In this context, we conducted root architecture analyses on the African oil palm (Elaeis guineensis) cultivar Deli × La Mé (D × LM) and the interspecific O×G hybrid (Elaeis oleifera × Elaeis guineensis) Coari × La Mé (C × LM) roots grown under nitrate nutrition at a concentration of 10 mM, which was previously identified as the optimal nitrogen condition [18].Nitrate was chosen as the nitrogen source due to its frequent presence in agricultural soil and its dual role, as it must convert to ammonium for plant assimilation.The objective was to elucidate how root architecture responds to nitrogen deprivation.The findings from this investigation will enable the identification of root traits dependent on the genotype during the nursery phase for future breeding programs and provide insights into the association between nitrogen availability and changes in root architecture.

Growth Conditions and Experimental Design
The study was performed at the Palmar de la Sierra Experimental Center in the Banana Zone, Colombia.The experiment was carried out in a mesh house with a 40% polyshade.Seeds at phenological stage 004 [20,21] of the interspecific O×G hybrid Coari × La Mé and the E. guineensis cultivar Deli × La Mé were grown in semi-hydroponic conditions as described by De La Peña et al. [18].Briefly, seeds were sown in 3.2 L pots with a mixture of the perlite-vermiculite substrate (1:1, v:v).After germination, the seedlings were irrigated for six weeks with deionized water three times per week.Then, in the seventh week, the seeds were removed from the seedlings, and nutrient solution treatments with (N+) and without (N−) nitrogen were applied., 0.12 µM CuSO 4 , and 0.06 µM KI (all the chemicals used were Sigma-Aldrich, St. Louis, MO, USA).The pH of the solution was 6.3.For the treatment with 10 mM of N, 5 mM Ca (NO 3 ) 2 was applied.Weekly, the seedlings were watered with 300 mL of the nutrient solution, divided into 100 mL doses every two days, keeping the substrate within the limits of a usable moisture fraction of -0.196 MPa on average.After cultivating for 14 weeks, the seedlings were carefully removed from the substrate, maintaining the root architecture as much as possible.They were then positioned against a blue background and photographed from a consistent distance each time.Finally, the roots were washed with deionized water to remove residues from the substrate and dried in an oven at 80 • C for 48 h to obtain the dry weight.The chemical and physical properties of the substrate are detailed by Ruiz-Romero et al. [22].Because the nutrient content in the substrate, specifically nitrogen, is poor, it was ensured that the results obtained are a consequence of the imposed treatments.A randomized complete block design was employed, utilizing a 2 × 2 factorial arrangement (two cultivars × two nitrogen conditions), comprising five replications and three plants per experimental unit.

Root Phenotyping
Root images were captured using a camera at a resolution of 72 pixels per inch (ppi) when the seedlings were 14 weeks old.The roots were positioned vertically against a blue background, with a ruler included for scaling purposes.Image analysis was conducted using RhizoVision Explorer Version 2.0.2 [23] in the whole root analysis mode.The parameters for root diameter ranged in lateral roots from 0 to 1.5 mm, and for the primary root from 1.5 mm and above.The features analyzed included total root length, total surface area, total volume, orientation angle, length of primary roots (PR), surface area of PR, volume of PR, surface angle frequency, length of lateral roots (LR), surface area of LR, volume of LR, inclined angle frequency, solidity, interaction area, hole size, and holes.

Statistical Analyses
Statistical analyses were conducted using R statistical software Version 4.3.0.Significant differences, resulting from nitrogen condition, cultivar, and the interaction of nitrogen condition × cultivar (N × C), were analyzed using a linear model through the 'lm' function, along with an analysis of variance (ANOVA) using the 'anova' function within the 'stats' package in R. For visualization purposes, in cases where the interaction was not significant, the main effects plot shows the boxplot outcomes for each factor (nitrogen condition or cultivar), displaying the values of the experimental units at each level of the main factor.Conversely, when the interaction was significant, the boxplot shows values for each factor and levels of the factor.The plots were generated using the 'ggplot2' package in R.

Effect of Nitrogen Starvation on Biomass Allocation
Seedlings at 14 weeks of age, after 7 weeks of nitrogen application, displayed 3-4 fully expanded lanceolate leaves, contrasting with the limited development observed in seedlings under nitrogen starvation, which had approximately 2 fully expanded lanceolate leaves (Figure 1).This result underlines the crucial role of nitrogen in leaf development, as it is an integral component of primary and secondary metabolites and nucleic acids.Notably, the results observed under nitrogen application were consistent with those reported for 3-month-old seedlings [24].On the other hand, the cultivar type did not affect the leaf number, so genetic variation among cultivars does not significantly influence this trait.

Effect of Nitrogen Starvation on Biomass Allocation
Seedlings at 14 weeks of age, after 7 weeks of nitrogen application, displayed 3-4 fully expanded lanceolate leaves, contrasting with the limited development observed in seedlings under nitrogen starvation, which had approximately 2 fully expanded lanceolate leaves (Figure 1).This result underlines the crucial role of nitrogen in leaf development, as it is an integral component of primary and secondary metabolites and nucleic acids.Notably, the results observed under nitrogen application were consistent with those reported for 3-month-old seedlings [24].On the other hand, the cultivar type did not affect the leaf number, so genetic variation among cultivars does not significantly influence this trait.Biomass gain in dry weight was assessed by comparing the weight of seedlings growing under nitrogen application (N+) and the control (N−) for each cultivar.In the aerial Biomass gain in dry weight was assessed by comparing the weight of seedlings growing under nitrogen application (N+) and the control (N−) for each cultivar.In the aerial part, the gain was substantial, with increases of 170% and 67% for C × LM and D × LM, respectively (Figure 1).The two-way ANOVA revealed a significant interaction between nitrogen condition and cultivar (N × C), indicating that the difference between the N+ and N-treatments was significantly more pronounced in C × LM than in D × LM.On the other hand, root biomass gain was moderate, with increases of 33% and 30% for C × LM and D × LM, respectively.Both nitrogen condition and cultivar had a significant effect on root biomass, while the N × C interaction was not significant.This suggests that both nitrogen application and cultivar type influence root biomass.Specifically, the aerial part biomass of C × LM responded more prominently to nitrogen fertilization than D × LM.In contrast, root biomass was more closely associated with the main effect of nitrogen condition and cultivar.
Regarding the distribution of dry matter, it was determined that the highest translocation of carbohydrates was toward the aerial part, presenting 51-68% and 58-64% of the total dry weight of the plant for the N− and N+ treatments in C × LM and D × LM, respectively.The above reveals the importance of nitrogen fertilization in the growth of nursery seedlings, especially in the aerial part, where nitrogen is essential for photosynthesis through enzymes, pigments, and other components that increase assimilation.However, in the control (N−), the dry weight of the aerial part was reduced more than in the root.In addition, an increase of 96% and 26% of the root-to-shoot ratio was observed in the N− condition compared to N+ for C × LM and D × LM, respectively (Figure 1.This agrees with the findings of Jackson et al. [25] that plants prioritize biomass distribution toward the roots to optimize efficient nitrogen uptake at the cost of carbon from the aerial part.This photoassimilate translocation response to the roots is also typical in plants under water stress, so this trait in nitrogen deficiency could be used in breeding methodologies to confer resistance to drought [24,25].Overall, the whole plant growth of C × LM was higher than D × LM (Figure 1).In relation to this, Ibarra-Ruales et al. [26] also observed higher O × G hybrid growth in nurseries compared to the E. guineensis cultivars under different soil conditions in Colombia.Similarly, though not statistically significant, Pinto Gloria et al. [19] found a trend of higher biomass in C × LM compared to E. guineensis seedlings at 8 months.This suggests that during the nursery stage, C × LM might require distinct management practices compared to D × LM, including variations in fertilization, irrigation, planting density, container volume, disease management, and other factors.

Root System Overview
Figure 2 shows the root architecture of C × LM and D × LM seedlings with nitrogen additions when they were 14 weeks old.Root identification followed the criteria outlined by Jourdan et al. [15].The primary root (PR) typically descended downward, exhibiting dense branching along its surface, except for the unbranched apical region.Lateral roots (LR) emerged at an approximate 90 • angle.Furthermore, younger tissues, predominantly lateral roots or the primary root tip, displayed a creamy white color, while older sections, mainly within the primary root, appeared dark brown.To adhere to the criteria for distinguishing true primary and lateral roots in the RhizoVision Explorer software, a diameter threshold of <1.5 mm successfully selected lateral roots, while >1.5 mm identified primary roots (Figure 3).This criterion aligns closely with that of Pinto Gloria et al. [19], who chose a root diameter >1 mm to separate thick roots in C × LM and E. guineensis seedlings, although the authors did not specify if this criterion solely identified primary roots.In the present study, differentiating between radicle and adventitious roots was not feasible due to similar branching structures at the developmental stage during sampling.Another limitation was related to the RhizoVision Explorer software, as documented by Martin et al. [17], which failed to accurately estimate the length of finer roots, such as the tertiary roots originating from the secondary ones.Consequently, we combined and referred to these as lateral roots.Nonetheless, the software was invaluable for isolating the root from the background in the segmented image (Figure 3) and extracting various features.

Root Plasticity in Response to Nitrogen Fertilization
RhizoVision Explorer is a software that analyzes both broken and whole roots to vide a comprehensive view of root structures.It is crucial to clarify that the concept cussed here operate within a 2D space, not 3D, to ensure understanding.For instanc parameter "volume" might imply working in a three-dimensional (3D) space.How in the context of our study, "volume" quantifies the spatial distribution or presence root in the 2D image rather than calculating an actual 3D volume.Further elaborati this concept can be found in Seethepalli et al. [23].

Root Plasticity in Response to Nitrogen Fertilization
RhizoVision Explorer is a software that analyzes both broken and whole roots to provide a comprehensive view of root structures.It is crucial to clarify that the concepts discussed here operate within a 2D space, not 3D, to ensure understanding.For instance, the parameter "volume" might imply working in a three-dimensional (3D) space.However, in the context of our study, "volume" quantifies the spatial distribution or presence of the root in the 2D image rather than calculating an actual 3D volume.Further elaboration on this concept can be found in Seethepalli et al. [23].

Root Plasticity in Response to Nitrogen Fertilization
RhizoVision Explorer is a software that analyzes both broken and whole roots to provide a comprehensive view of root structures.It is crucial to clarify that the concepts discussed here operate within a 2D space, not 3D, to ensure understanding.For instance, the parameter "volume" might imply working in a three-dimensional (3D) space.However, in the context of our study, "volume" quantifies the spatial distribution or presence of the root in the 2D image rather than calculating an actual 3D volume.Further elaboration on this concept can be found in Seethepalli et al. [23].
In this context, after analyzing the images in whole mode within the RhizoVision Explorer software, we extracted 16 features.There were significant differences in the total length, surface area, and volume of roots due to nitrogen treatment and cultivation, with the highest values found in the N+ condition and the C × LM cultivar (Figure 4).The main effect of the nitrogen condition, cultivar, and interaction effect were observed in the primary roots' length, surface area, and volume.Consistent with the findings of the effect of nitrogen application, Naulin et al. [27] demonstrated that nitrogen nutrition stimulates primary root growth through the cytokine signaling pathway.
The interaction effect indicates that the difference in the primary root length between the N+ and N− treatments was more in C × LM than in D × LM.On the other hand, for the length, area, and root volume of lateral roots, the interaction effect did not present significant differences.However, these parameters are of great importance to identifying the genetic material since a considerable effect of the cultivar was observed (Figure 4).Similarly, Pinto Gloria et al. [19] noted that the hybrid C × LM has more lateral roots than the E. guineensis plants.
In Arabidopsis, Gruber et al. [28] found that the lengths of primary and lateral roots increased under moderate nitrate deficiency conditions, while they decreased under severe conditions.In rice, Müller [29] reported an increase in the lengths of primary and lateral roots under conditions of nitrogen deficiency.Therefore, it has been observed that the length of the roots in response to nitrogen deficiency varies according to the genetics of the plants and the severity of the deficiency.In this sense, in our specific study, the observation that lateral roots are not affected by nitrogen starvation could indicate a prioritization of the plant, regardless of the cultivar, in maintaining the growth of lateral roots to search for nitrogen.Nevertheless, since C × LM typically invests more in lateral roots than D × LM regardless of nitrogen conditions, this could also lead to a decrease in the growth of primary roots under nitrogen starvation, as resources are directed towards maintaining lateral root morphology.However, we did not separately weigh the lateral roots to support this hypothesis.A study by Pinto Gloria et al. [19] found that C × LM translocated more biomass to the fine and very fine roots, indicating a potential genetic factor.Interestingly, this trait is not exclusively expressed under favorable growth conditions; Silva et al. [30] demonstrated in two oil palm hybrids that root length tends to increase or remain stable under drought stress compared to control conditions.
On the other hand, the nitrogen condition and cultivar slightly altered the root angle, being on average 1.5 • steeper under the N+ condition than the N− and 0.8 • steeper in C × LM than D × LM (Figure 4).Likewise, the frequency of surface roots was significantly higher in the N− condition than N+ and in the cultivar D × LM than in C × LM, while the frequency of steep-angled roots was significantly higher in the N+ than the N− condition (Figure 4).Contrary to these results, root insertion angles are typically steeper under nitrogen deficiency conditions than the average [31], a desirable strategy for roots to acquire nitrogen from deeper areas where it tends to mobilize.Trachsel et al. [32] found that the roots were 18 • more inclined under nitrogen deficiency conditions than normal conditions in maize.The results of this experiment may differ from those reported in other studies due to the different ways of creating a nitrogen deficiency.These studies induced nitrogen deficiency at very low concentrations, in which a small amount of nitrate was mobilized to the bottom of the container.In contrast, in the present experiment, nitrogen was only applied to the substrate with the N+ condition, while the substrate with the N− condition never received N. Therefore, the concentration of the element was almost zero in all its directions, a condition that was probably sensed by the plant, thus avoiding investing energy in deepening the roots.Features extracted from root image processing of Coari × La Mé and Deli × La Mé under both nitrogen-deficient and nitrogen-applied conditions after 14 weeks.In cases where there is no interaction in the two-way ANOVA analysis, the values are represented by the main effect of the nitrogen condition (n = 10) or the main effect of the cultivar (n = 10).When an interaction is present,

Figure 1 .
Figure 1.Number of fully expanded leaves and biomass accumulation in shoots and roots of the African oil palm cultivar Deli × La Mé and the interspecific O×G hybrid Coari × La Mé under nitrogen-deficient (N−) and nitrogen-applied (N+) conditions after 14 weeks.In cases where there is no interaction in the two-way ANOVA analysis, the values are represented by the main effect of the nitrogen condition (n = 10) or the main effect of the cultivar (n = 10).When an interaction is present, the values are depicted with the effect of each factor and the levels of the factor (n = 5).The variables included are leaf number, shoot biomass, and root biomass.Asterisks indicate the level of significance of the ANOVA due to the factors nitrogen condition (N), cultivar (C), and the interaction between nitrogen condition and cultivar (N × C). ***: p ≤ 0.001.

Figure 1 .
Figure 1.Number of fully expanded leaves and biomass accumulation in shoots and roots of the African oil palm cultivar Deli × La Mé and the interspecific O×G hybrid Coari × La Mé under nitrogen-deficient (N−) and nitrogen-applied (N+) conditions after 14 weeks.In cases where there is no interaction in the two-way ANOVA analysis, the values are represented by the main effect of the nitrogen condition (n = 10) or the main effect of the cultivar (n = 10).When an interaction is present, the values are depicted with the effect of each factor and the levels of the factor (n = 5).The variables included are leaf number, shoot biomass, and root biomass.Asterisks indicate the level of significance of the ANOVA due to the factors nitrogen condition (N), cultivar (C), and the interaction between nitrogen condition and cultivar (N × C). ***: p ≤ 0.001.

Figure 2 .
Figure 2. Image of the root system of oil palm seedlings grown in perlite-vermiculite (1:1) m with supplemented nitrogen.

Figure 3 .
Figure 3. Two-dimensional root images of Coari × La Mé and Deli × La Mé under conditi deficiency (−) and application (+) of nitrogen after 14 weeks using the RhizoVision Exp (A,D,G,J) Original image of the roots.(B,E,H,K) Processed images.The image shows a blue c polygon wrapping around the entire root system for extracting the convex area.Lines in dar indicate the primary roots (diameter >1.5 mm), and lines in red indicate the lateral roots (dia <1.5 mm).The "holes" or patches in the image are randomly colored to distinguish them.(C Segmented image of the root.

Figure 2 .
Figure 2. Image of the root system of oil palm seedlings grown in perlite-vermiculite (1:1) mixture with supplemented nitrogen.

Figure 2 .
Figure 2. Image of the root system of oil palm seedlings grown in perlite-vermiculite (1:1) mixture with supplemented nitrogen.

Figure 3 .
Figure 3. Two-dimensional root images of Coari × La Mé and Deli × La Mé under conditions of deficiency (−) and application (+) of nitrogen after 14 weeks using the RhizoVision Explorer.(A,D,G,J) Original image of the roots.(B,E,H,K) Processed images.The image shows a blue convex polygon wrapping around the entire root system for extracting the convex area.Lines in dark blue indicate the primary roots (diameter >1.5 mm), and lines in red indicate the lateral roots (diameter <1.5 mm).The "holes" or patches in the image are randomly colored to distinguish them.(C,F,I,L) Segmented image of the root.

Figure 3 .
Figure 3. Two-dimensional root images of Coari × La Mé and Deli × La Mé under conditions of deficiency (−) and application (+) of nitrogen after 14 weeks using the RhizoVision Explorer.(A,D,G,J) Original image of the roots.(B,E,H,K) Processed images.The image shows a blue convex polygon wrapping around the entire root system for extracting the convex area.Lines in dark blue indicate the primary roots (diameter > 1.5 mm), and lines in red indicate the lateral roots (diameter < 1.5 mm).The "holes" or patches in the image are randomly colored to distinguish them.(C,F,I,L) Segmented image of the root.

Figure 4 .
Figure 4. Features extracted from root image processing of Coari × La Mé and Deli × La Mé under both nitrogen-deficient and nitrogen-applied conditions after 14 weeks.In cases where there is no