The lagoon area II is located at the far eastern part of the the Ébrié system ( Figure 1 ), specifically between 3˚40' and 3˚50' West longitudes and 5˚20' and 5˚21' North latitudes. Its mean length from East to West is 17.143 km, while its mean width is 5.714 km [19] . This lagoon ecosystem covers an area of 87 km² [16] . The hydrology of this lagoon area is primarily influenced by Comoé River, Mé River, and Atlantic Ocean. The influence of the latter on the zone has increased considerably with the reopening of the Grand-Bassam inlet. Its continental watershed is mainly that of Comoé River, which drains the entire Côte d’Ivoire from Northeast to South over an area of 78,000 km² [20] , given that Mé River watershed covers only 4300 km² [21] . Mé River contributes to this lagoon site, as do the Aghien and Potou lagoons. The altitude of the Ébrié system in general, and the lagoon area II in particular, relative to Atlantic Ocean, favors permanent inputs of water and sediments from Comoé River towards Atlantic Ocean [18] . This has long been the cause of the silting and subsequent closure of the Grand-Bassam inlet [16] .

The strong marine intrusion, caused by the reopening of the Grand-Bassam inlet, is likely to alter its mode of functioning, particularly its hydrochemistry [18] . Such a situation has already been observed by Albaret and Ecoutin [22] , who noted a modification of the floristic composition and the mortality of certain macrophytes, such as Eichhornia crassipes, during the penultimate reopening of this pass in 1987.

Considering the hydrological regime of Mé River, which depends on the subequatorial climate in the South, and that of Comoé River, which is a function of the subequatorial climate in the South, the humid tropical climate in the Center, and the dry savanna tropical climate in the North [23] , on the one hand, and the hydrological seasons of Atlantic Ocean off Abidjan on the other, Drida and Yao [18] defined the hydrological seasons of this lagoon ecosystem as follows: Hot Season (HS), from December to April; Rainy Season (RS), from May to July; Cold Season (SF), from August to September and; Flood Season (FS), from October to November. The continental watershed is dominated by the intensive practice of agriculture with chemical inputs, industrial and mining activities, as well as the practice of illegal gold mining [19] .

Drida and Yao [18] observed that, just after the reopening of the inlet, the surface sediments were dominated by very coarse, coarse, and medium sands, originating mainly from Comoé River. A similar finding was reported by Adopo et al. [15] and Irié et al. [24] before the pass was reopened, with the latter noting that these sands were deposited in a fluviatile, beach-type environment. Their mineralogy is diverse, including the presence of monazites, kaolinites, magnetites, ilmenites, and feldspars, as reported by Irié et al. [24] .

This study spanned a full year, from January to December 2024. Samples were collected from five sampling sites. These sites are identical to those used by Mahi et al. [19] just prior to the reopening of the inlet, as well as those employed by Drida and Yao [18] just after this reopening ( Figure 2 ).

Samples were collected from the study sites using a Van-Veen grab, taking the top 0 - 5 cm of sediment in strict compliance with the ISO 5667-19 standard [25] . Upon collection, the samples were immediately transferred to polyethylene bottles, as recommended by the ISO 5667-15 standard [26] . Monthly collections were conducted, yielding a total of 60 sediment samples over the entire study period.

Samples were processed in the laboratory by first removing coarse materials such as plant debris, rubble, and shells. They were then homogenized and oven-dried in accordance with the ISO 17892-1 A1 standard [27] before being ground to a particle diameter less than 0.63 µm. After processing, the samples were hermetically sealed in polyethylene bottles and stored in a moisture-free environment, as recommended by the ISO 5667-15 standard [25] .

Total phosphorus concentrations in the various samples were determined by the acid persulfate digestion method, in accordance with the ISO 6878 standard [28] . All experiments were performed in triplicate (n = 3).

The various phosphorus fractions were determined using the sequential extraction protocols of Bonzongo [9] , Hieltjes and Lijklema [10] , Paludan and Jensen [11] , and Azzouz [12] . The detailed implementation of these protocols is presented in Figure 3 . The concentrations of the different phosphorus fractions were obtained by the ammonium molybdate spectrometric method, according to the ISO 6878 standard [28] . All experiments were also performed in triplicate (n = 3).

Based on the results obtained, the seasonal phosphorus recovery rates (R) in these substrates were evaluated using the following relationship:

$\text{R}\left(\%\right)=\frac{{\displaystyle \sum \text{Phosphorus fraction concnetration}}}{\text{Total phosphorus concentration}}\times 100$

A protocol demonstrates satisfactory efficiency and precision in a given season when its recovery rates fall within the 80% to 120% range [29] .

The development of a sequential phosphorus extraction protocol adapted to these sediments will be based on the structure of the protocol, among the four studied, that exhibits the highest seasonal recovery rates. If several protocols yield similar results, the one capable of determining the greatest number of phosphorus fractions will be selected. Subsequently, distinct phosphorus fractions measured by the other three protocols will be integrated into the chosen methodology, while carefully assessing the impact of the extractants on the subsequent analytical steps to avoid any interference.

To optimize this protocol, the volumes of the extractants will be adjusted from 10 to 50 mL. The selection of this range of extractant volumes is not arbitrary; rather, it results from a careful balance between chemical efficiency and analytical precision. Extractant volume primary role is to ensure thermodynamic equilibrium by maintaining a sufficient reagent concentration to shift the equilibrium toward the dissolution of sedimentary phosphorus. This volume is directly related to the solid-to-liquid ratio (S/L), a critical parameter that influences kinetics. An optimal volume ensures efficient contact between the solid and liquid phases. Furthermore, a sufficient volume is crucial for preserving the extractant’s buffering capacity, thereby maintaining a constant pH or ionic strength. This stability enables the selective solubilization of phosphorus (particularly that bound to iron and aluminium oxides). Ultimately, this specific range represents an indispensable analytical compromise: it maximizes the extraction yield while ensuring the final phosphorus concentration in the solution remains high enough for accurate measurement using standard laboratory techniques (UV/Vis) [12] .

The optimal volume will be the one that yields the highest concentration for the specific phosphorus form under consideration. In the case where a single extractant is used to determine the concentrations of two phosphorus forms, the best volume will be the one that simultaneously provides the highest concentrations for both forms. If two volumes of a single extractant lead to maximum concentrations for a phosphorus form that are not statistically different, the smaller volume will be retained in the interest of reagent economy.

A protocol demonstrates satisfactory efficiency and precision in a given season, therefore on all the study period, when its recovery rates fall within the 80% to 120% range [29] .

In addition to standard univariate statistical techniques (namely the mean (m), standard deviation (s), coefficient of variation (VC), minimum value (Min), and maximum value (Max)), a one-way ANOVA was used. The one-way ANOVA aimed to determine if there was a statistically significant difference between the various seasonal phosphorus recovery rates obtained with the four protocols. This statistical test was also employed during protocol optimization to determine if the concentrations of the different phosphorus forms obtained with various volumes of an extractant were statistically different. When such a difference was significant, three post-hoc tests (Fisher’s LSD, Tukey’s HSD, and Dunnett’s test) were implemented to identify distinct seasonal means and homogeneous subgroups that did not differ from one another. The results of the one-way ANOVA and post-hoc tests were considered statistically significant at p < 0.05.

The concentration of total phosphorus in these sediments ranged from 849.9 ± 109.9 (RS) to 1199.9 ± 104.7 (HS) ( Figure 4 ). A one-way ANOVA revealed a significant overall difference between the seasonal total phosphorus concentrations in these substrates (p < 0.05). Subsequent post-hoc tests showed that only the total phosphorus concentration in the HS was statistically different from that in the RS (p < 0.05). Thus, the seasonal total phosphorus concentrations in these sediments, over the study period, followed the following increasing order:

Ptot (HS) > Ptot (RS) = Ptot (CS) = Ptot (FS).

The seasonal phosphorus recovery rates in these substrates, obtained with the Paludan and Jensen protocol [11] , were found to be the highest, except during the RS, where the highest rate was reportedly determined using the Bonzongo protocol [9] . The Azzouz protocol [12] , in contrast, showed the lowest seasonal recovery rates ( Figure 5 ).

A one-way ANOVA revealed overall significant differences between the mean seasonal phosphorus recovery rates of these protocols. Post-hoc tests demonstrated that only the phosphorus recovery rates determined with the Bonzongo protocol [9] were not statistically different from those obtained with the Paludan and Jensen protocol [11] . Consequently, in virtually every season, the phosphorus recovery rates followed the following order of importance:

TR (Bonzongo) = TR (Paludan et Jensen) > TR (Hieltjes et Lijklema) > TR (Azzouz)

Considering the efficiency requirement for a sequential extraction protocol, which necessitates a recovery rate between 80% and 120%, it is clear that only the protocols of Bonzongo [9] and Paludan and Jensen [11] are suitable for studying phosphorus mobility in these substrates during the RS and CS, while they are unsuitable during the HS and FS.

Given the advantages offered by the Paludan and Jensen protocol [11] compared to that of Bonzongo [9] , particularly its ability to determine a greater number of phosphorus fractions, it is more appropriate to develop an optimal extraction protocol for these sediments based on this methodology.

1) Protocol Development

A new protocol was devised to integrate nearly all the phosphorus fractions proposed by the four methods used in this study. The flowchart for this protocol is presented in Figure 6 .

2) Protocol Optimization

a) Selection of Optimal Extractant Volumes

Table 1 presents the mean concentrations of the different phosphorus fractions obtained using volumes ranging from 10 to 50 mL.

The mean concentration of SO-P in these sediments was observed to increase progressively with increasing the volume of 30% H₂O₂. A one-way ANOVA

SO-P: soluble organic phosphorus; labile-P: labile phosphorus; Fe-P: iron bound to phosphorus; Al-P: Aluminium bound to aluminium; HA-P: humic acid bound to phosphorus; oxy-P: metal (oxy)hydroxides bound to phosphorus; auth-P: authigenic phosphorus; Ca-P: detrital phosphorus; RO-P: residual organic phosphorus.

revealed statistically significant overall differences among these concentrations as a function of the volume (p < 0.05). Post-hoc tests showed that only the mean SO-P concentration obtained with a 50 mL volume was statistically different from those obtained with the other volumes (p < 0.05). Consequently, the 50 mL volume of 30% H₂O₂, which yielded the greatest POS concentration in these substrates, is considered optimal for this study.

The mean concentration of labile-P also increased progressively with the increasing volume of 1 M NH₄Cl. A one-way ANOVA highlighted overall significant differences among the mean P-labile concentrations (p < 0.05). Post-hoc tests revealed that only the labile-P concentrations obtained with 40 mL and 50 mL were not statistically different (p < 0.05). Therefore, in the interest of both efficiency and reagent economy, it is advisable to retain the 40 mL volume for the extraction of this phosphorus fraction from these substrates.

Just as with the mean concentrations of SO-P and labile-P, the concentration of Fe-P increased progressively with the increasing volume of the combined 0.1 M Na₂S₂O₄/0.1 M NaHCO₃ reagent. A one-way ANOVA highlighted overall significant differences among the mean Fe-P concentrations obtained with the different volumes (p < 0.05). Post-hoc tests revealed that only the Fe-P concentrations obtained with the 20 mL and 30 mL volumes were not statistically distinct (p > 0.05). Furthermore, the highest Fe-P concentrations in these sediments were obtained with 50 mL. Consequently, this volume was retained for the extraction of this phosphorus fraction in the remainder of the study.

The mean concentration of Al-P increased progressively with the 0.1 M NaOH volume from 10 to 40 mL, then slightly decreased at a volume of 50 mL. Conversely, the mean concentration of HA-P increased consistently with the increasing volume of this extractant. A one-way ANOVA revealed overall significant differences between the mean concentrations of these two phosphorus forms obtained with the different volumes of 0.1 M NaOH (p < 0.05). Post-hoc tests revealed the following:

The difference between the mean Al-P concentrations obtained with 40 mL and 50 mL of 0.1 M NaOH is less significant than with the other volumes (10, 20, and 30 mL). To simultaneously obtain high concentrations for both Al-P and HA-P in these substrates, it is therefore judicious to select a 40 mL volume.

The highest mean concentration of oxy-P was observed at a 20 mL volume of 1 M NaOH, while the lowest was observed with 10 mL and 30 mL of this extracting agent. A one-way ANOVA highlighted significant overall differences among the oxy-P concentrations obtained with these volumes (p < 0.05). Post-hoc tests revealed that only the oxy-P concentrations obtained with the 20 mL volume of 1 M NaOH were statistically different from those obtained with all other volumes (10, 30, 40, and 50 mL) (p < 0.05). Consequently, the 20 mL volume is deemed optimal for the extraction of this phosphorus fraction in this context.

The highest mean concentrations of auth-P were observed with 20, 40, and 50 mL volumes of the CH₃COONa/CH₃COOH buffer solution (pH = 4). The lowest was noted with a 10 mL volume. A one-way ANOVA highlighted overall significant differences among the mean auth-P concentrations obtained with these different volumes (p < 0.05). Post-hoc tests revealed that:

Given these observations and in the interest of reagent economy, it is more judicious to select the 20 mL volume for the extraction of this fraction.

The mean concentrations of Ca-P progressively increased with the 1 M HCl volume from 10 to 20 mL, and then from 30 to 50 mL. A one-way ANOVA revealed overall significant differences among the mean concentrations of Ca-P obtained with the different volumes of this acid. Post-hoc tests showed no statistically significant difference between:

From these observations, it’s appropriate to select 20 mL as the optimal volume of this extracting agent for this phosphorus fraction in these sediments. This choice balances the need for high extraction efficiency with reagent economy.

The optimized sequential extraction scheme for the different phosphorus fractions considered in this study is presented in Figure 7 .

b) Protocol Efficiency

The seasonal phosphorus recovery rates in the surface sediments of the study area ranged from 85.7% in the HS to 114.7% in the RS, with low intra-seasonal variations. All these rates were found to be higher than those obtained with the Paludan and Jensen protocol [11] , with the exception of the CS rates ( Table 2 ).

These seasonal rates, all of which fall between 80% and 120%, confirm the efficiency of this method for studying phosphorus mobility in these substrates. Furthermore, this protocol enables the assessment of SO-P, auth-P, and oxy-P concentrations, in addition to those accounted for by the Paludan and Jensen protocol [11] . As a result, it ensures high efficiency for the study of phosphorus mobility, regardless of the season.

Total phosphorus concentrations in the superficial sediments of this estuary were exceptionally high throughout the study period, surpassing values documented in comparable environments, such as the Al-Shaibah and Al-Budhia lagoons (Red Sea, Saudi Arabia) [30] , the northern Adriatic Sea [31] , and the Shamei Lagoon [32] . This enrichment is attributed to the combined influence of the sandy sediment texture, the estuary’s hydrology, and its overall sedimentological processes [33] - [35] . This is driven by the presence of detrital phosphorus associated with weathering-resistant primary minerals, coupled with the significant accumulation of organic matter in the sedimentary environments [33] [35] . The strong intrusion of Atlantic oceanic waters during the HS drives the relatively high phosphorus content observed in these substrates. The elevated salinity and ionic composition of the oceanic input, coupled with increased oxygen levels, catalyze phosphorus-trapping processes (flocculation and precipitation) that sequester phosphorus originating from the Comoé River. This leads to the accumulation of phosphorus-rich, fine-grained material within the sandy matrix [29] [35] [36] . These stabilizing phenomena are significantly reduced during high-precipitation seasons (from RS to FS) due to inhibited oceanic influence by Comoé River. Consequently, the high total phosphorus concentration measured during the HS was statistically greater than those observed during the other three seasons.

Analysis of the seasonal phosphorus recovery rates demonstrates that sequential extraction protocols utilizing both a strong oxidant and large reagent volumes (the protocols of Bonzongo [9] and Paludan and Jensen [11] ) achieved significantly higher rates. These results are in sharp contrast to those from protocols that use no oxidizing agent (the Hieltjes and Lijklema protocol [10] ) and those that employ low volumes of extraction reagents (the Azzouz protocol [12] ). The Hieltjes and Lijklema protocol [10] has known limitations for phosphorus speciation, especially in complex mineral matrices like rock fragments. These issues stem from reagent inadequacy [37] [38] . The Bonzongo protocol [9] , developed as a modification of the Hieltjes and Lijklema protocol [10] , overcomes these challenges in geological matrices through the critical preliminary step of using concentrated H₂O₂ (30% - 35%). This powerful oxidizing agent targets organic matter and sulphides (major immobilizers of phosphorus), catalyzing the degradation of organic bonds and releasing phosphorus as soluble phosphate ions [7] [39] [40] . By eliminating organic matrix interferences and improving access to occluded phosphorus, H₂O₂ serves as an essential pretreatment, significantly enhancing the efficiency and precision of subsequent extraction steps [9] . This mechanistic improvement ultimately justifies the significantly higher seasonal phosphorus recovery rates observed with the Bonzongo protocol [9] compared to the Hieltjes and Lijklema protocol [10] in this study. The Paludan and Jensen protocol [11] notably improves speciation by providing access to the RO-P. This is achieved using Na₂S₂O₈, a stronger oxidizing agent than H₂O₂ used in the Bonzongo protocol [9] . Due to its comprehensive methodology, the Paludan and Jensen protocol yielded seasonal total phosphorus recovery rates in the studied sediments that were statistically identical to those obtained with the Bonzongo protocol [9] . In contrast, the Azzouz protocol [10] exhibited low seasonal P recovery rates due primarily to two factors: insufficient extractant volumes and the inadequacy of its digestion method for these complex sediments [41] - [44] .

The efficiency of the Bonzongo [9] and Paludan and Jensen [11] protocols for studying phosphorus mobility in the sediments of this estuary varies seasonally. Their inadequacy during certain periods is explained by both the sediment mineralogy and the volumes of extractants used. Although these sediments predominantly originate from Comoé River, their mineralogical composition changes seasonally, as the river’s watershed is sedimentary in the south and rests on a geological basement from the center to the north [18] [20] . The results obtained with these two protocols indicate that the sediments of this estuary contain more easily extractable phosphorus forms during the CS and RS, compared to the other two seasons. Furthermore, the excess volume of extractants used during sequential phosphorus extractions may have skewed the concentrations due to their interference with the reagents employed for orthophosphate determination via colorimetry, particularly with sediments sampled during the HS and FS. These interferences have been widely documented, showing that extractant residues can compromise analytical accuracy, especially for samples with high mineral phosphorus content [45] .

These observations highlight that phosphorus sequential extraction in these sediments is intrinsically impacted by their sandy nature, the estuary’s hydrology, and its sedimentological processes. These factors determine how phosphorus is stored, remobilized, and, consequently, the quantity recovered in each operational fraction (labile, Fe/Al oxide-bound, organic, and residual) [33] - [35] [46] . The water regime controls the sedimentary phosphorus stock through two mechanisms. First, external inputs (meteorological runoff, Comoé River during the FS, and Mé River via the Potou and Me lagoons during the RS) introduce an increased quantity of suspended matter rich in organic phosphorus and primary mineral phosphorus. These inputs are likely to increase sedimentary phosphorus because the material is freshly deposited and less chemically transformed [37] [46] . Second, the estuary’s shallow depth and water residence time control oxygenation [46] . During seasons of high meteorological inputs (notably in the FS), poor water ventilation leads to significant organic matter degradation, resulting in anoxic conditions at the water-sediment interface. This anoxia triggers the reduction of iron (Fe(III) to Fe(II)), which dissolves oxide matrices and releases the Fe-bound to phosphorus (Fe-P) [5] [33] [36] , thereby significantly influencing the measured speciation profile. Furthermore, during the HS, strong oceanic water intrusion can increase water salinity, promoting phosphorus desorption via ionic competition and potentially affecting the most labile fractions [35] . The sandy nature of these sediments translates into a low overall phosphorus adsorption capacity. In this sediment type, phosphorus is not trapped by the surface of the grains themselves but is strongly bound to Fe/Al oxide coatings or to fine organic matter associated with the sand [36] [47] . Consequently, any sequential extraction performed on these sediments must effectively target these coatings (notably via NaOH or H₂O₂), as the phosphorus is not homogeneously distributed [39] [40] . Furthermore, both the quantity and nature of the organic matter directly influence the organic phosphorus fraction [33] [35] . A sedimentology that favours the accumulation of autochthonous organic matter leads to a more labile phosphorus organic fraction compared to phosphorus organic originating from more refractory allochthonous organic matter [34] . This process significantly impacts the measured speciation profile.

The suitability of the implemented protocol for studying phosphorus mobility in these substrates, regardless of the season, is based on two major methodological adaptations. On the one hand, the use of a H₂O₂ pretreatment to oxidize organic matter increased the precision and efficiency of the subsequent extraction methods. On the other hand, adapting the volumes of the extractants was essential to minimize interference from their residues with the reagents used for orthophosphate determination via colorimetry. Despite the robustness of this approach, the low phosphorus recovery rate observed in the sediments of the HS, compared to that of other seasons, illustrates a high degree of crystallization of this nutrient during this period. This contrasts with the presence of Atlantic oceanic waters in this estuary, which normally promotes phosphorus trapping on the sediments through flocculation and precipitation [33] [36] . The high crystallization of these substrates considerably limits phosphorus extraction, highlighting its low mobility and geochemically stable nature [37] [38] during this season. The identification of multiple organic phosphorus fractions is essential for assessing the lagoon’s potential for delayed P release. The mobility and bioavailability of organic phosphorus typically decrease in the order: SO-P > HA-P > RO-P [4] . By quantifying these specific fractions, this protocol allows researchers to distinguish the immediately threatening phosphorus (SO-P) from the stabilized stock (RO-P), thereby providing a more precise picture of the sediment’s actual trophic potential.