A Comparison of Characteristics of Infilling Sediments in Three Mud-Capped Dredge Pits on the Louisiana Continental Shelf

Abstract

Due to high sedimentation rate up to ~1 m/year, mud-capped dredge pits (MCDP) are often considered natural laboratories for studying sedimentary processes, slope stability and the impacts of dredging activities on marine environments. Although many studies have been performed on the Louisiana shelf, there is a lack of high spatial resolution research covering the eastern, central and western Louisiana shelf to comprehensively investigate sediment infilling. Eighteen vibracores were collected from the Peveto Channel dredge pit (PC), Raccoon Island dredge pit (RI) and Sandy Point dredge pit (SP), and more than 1300 samples were analyzed to study the spatial variation in surficial sediment using statistical analyses. Our results indicate that the inner Louisiana continental shelf is silt-dominated, and there was no consistent grain size variation when comparing the sediment within the pits with that outside the pits. Skewness emerged as a prominent factor in the RI and SP, while standard deviation was the most influential in the PC. Our analysis shows also that two principal components are confirmed and account for more than 95% of the total grain size variance.

1. Introduction

The Mississippi River system drains approximately 3,224,600 km², making it the largest drainage basin in the United States. One of its major distributaries, the Atchafalaya River, forms an extensive network of anastomosing channels, backwater swamps, freshwater marshes and wetland forests. In Louisiana, the Mississippi–Atchafalaya River system is the major supplier for suspended and dissolved continental materials to the coastal ocean, approximately 66% [1]. The sedimentation rate is around 7–8 cm·year⁻¹ in the Mississippi River mouth and is 2.2–5.6 cm·year⁻¹ in the Atchafalaya River mouth [2]. On the Mississippi River’s path to coastal ocean, the river changed courses and formed paleo-river channels and abandoned delta lobes composed of sand and mud over several thousands of years. Six larger deltaic lobes were formed by the Mississippi–Atchafalaya River system over the past 4600 years, including Salé-Cypremort, Cocodrie, Teche, St. Bernard, Lafourche, Plaquemine and Balize (modern).

In studies of the Mississippi River Delta and the northern Gulf of Mexico, dredge pits and delta lobe switching processes are regarded as natural laboratories for investigating sediment dynamics, river avulsion and sediment redistribution. Early work (e.g., [3]) established the geomorphic framework of lobe succession in the Mississippi Delta, and subsequent studies have explored how lobe switching influences sediment organization, deposition rates and coastal evolution [4,5]. Regarding the mechanisms of lobe switching, Refs. [6,7] emphasized the critical roles of channel evolution, aggradation and autogenic avulsion, while more recent studies have highlighted scale-dependent processes under high sedimentation rates and the comparability between physical and numerical modeling [8,9]. At local to nearshore scales, dredge pits and delta lobes can create pronounced spatial heterogeneity, affecting not only grain-size distribution and slope stability but also fine-scale transport processes [10,11]. Recent multi-scale observations and modeling efforts (e.g., [12,13]) further demonstrate that climate variability, human river engineering and sediment supply strongly alter the frequency and spatial distribution of lobe switching, thereby reshaping continental shelf and deep-sea depositional patterns. Building on this context, the present study employs high-resolution sampling from three dredge pits to compare sediment characteristics inside and outside the pits under conditions of rapid sedimentation (~1 m/year), aiming to fill a gap in spatially explicit sedimentological research and to explore how lobe switching interacts with local grain-size variability.

Sands in the submarine shoals and paleo-river channels of Mississippi–Atchafalaya River system have been widely used for barrier island restoration projects in coastal Louisiana, but there is still limited knowledge on how dredged pits evolve and what kind of sediment infills which is significant to ensuring long-term coastal restoration success, minimizing environmental impacts and supporting sustainable sediment resource management. Past studies were mainly focused on sandy dredge pits [14,15,16,17] or on the muddy dredge pits what are partially infilled by muddy sediments [18]. Sand buried in paleo-river channels is often covered by a few meters thick mud caps in the northern Gulf due to abundant mud supply from the muddy Mississippi River. Mud-capped dredge pits (MCDPs) are defined as a pit formed after the removal of a muddy cap at seafloor surface and the excavation of sand from paleo-river channels. Three large MCDPs in coastal Louisiana are Peveto Channel, Raccoon Island and Sandy Point (from west to east; Figure 1). The Peveto Channel dredge pit, which is offshore Holly Beach of Louisiana, was excavated in the year 2003. This pit has dimensions of 400 m (shore-parallel) by 600 m (shore-normal) and was about 8 m deep after dredging [19,20]. Robichaux et al. [21] conducted a geophysical study on this pit in 2016 and found that the pit was 100% filled up. They concluded the continuing dewatering and long consolidation processes at this pit after a complete infilling. However, they did not collect any sediment samples to study the consolidation processes. Raccoon Island is the westernmost barrier island in the Isles Dernieres chain in central coastal Louisiana [22]. This dredge pit was excavated in 2013 offshore Raccoon Island (Figure 1). The Raccoon Island pit experienced a fast-infilling rate of 50–100 cm year⁻¹ in a few years after the dredging in 2013 and was 100% filled up in 2018. However, three years later, this pit experienced a “new subsidence” of 0.5 m below ambient sea floor, possible due to fast degassing and consolidation in response to energetic events [23]. This is the first time that a filled-up dredge pit in Louisiana experiences a new subsidence/collapse, indicating that long-term degassing and continuing consolation should be considered in long term pit morphology prediction. The Sandy Point dredge pit was excavated in 2012 to provide sand sources for restoration of the Barataria Bay coastline [23]. This pit is located 20 km west of the Balize Mississippi Delta in a water depth of 11 m and experienced a sediment infilling of about 54 cm year⁻¹ [24] (Figure 1).

Compared with natural deposition rates of 1–8 cm year⁻¹ at river mouths, the high sediment deposition rates of 54–100 cm year⁻¹ in three MCDPs provide a unique opportunity to study high temporal variations in depositional materials. Three MCDPs serve as ideal natural “laboratories” to study the transport and deposition of muds, sands, organic matter and carbonate on the Louisiana continental shelf, from east to west. Our study aims to address the following three scientific objectives that will improve our understanding of MCDPs’ evolution after dredging: (1) compare the means and variations in sediment grain size, organic matter and carbonate content in three pits; (2) investigate the correlations and relative importance among grain size, organic matter and carbonate content; and (3) discuss sediment sources and physical processes controlling sediment infilling. Our results will help the scientific community, sand resource managers and policymakers better understand the sediment dynamics inside and outside dredge pits and minimize the impacts of dredging activities on the ambient geological and biological environments.

2. Materials and Methods

2.1. Field Data Acquisition

All the cores used in this study were collected using a 5 m-long aluminum core barrel aboard Louisiana State University (LSU) Coastal Studies Institute’s R/V Coastal Profiler (Louisiana State University, Baton Rouge, LA, USA). In Peveto Channel (here defined as PC) dredge pit (west), 7 vibracores were collected on 28 July 2021 (Figure 1;

Table 1. Sediment cores collected during field cruises. See Figure 1 for coring locations.

Core Name	Date	Latitude (Degree N)	Longitude (Degree W)	Length (m)
RI1	19 August 2020	28.9831	90.9212	1.97
RI2	19 August 2020	28.9784	90.9225	2.63
RI3	19 August 2020	28.9774	90.9223	3.01
RI4	19 August 2020	28.9762	90.9221	3.31
RI5 *	19 August 2020	28.9814	90.9235	3.78
RI9 *	19 August 2020	28.9814	90.9241	3.28
PC3	15 July 2021	29.6959	93.54651	3.14
PC4	15 July 2021	29.6955	93.5469	3.25
PC5	15 July 2021	29.6952	93.5472	3.21
PC6	15 July 2021	29.6950	93.5476	3.19
PC8 *	15 July 2021	29.6919	93.5500	2.32
PC9	15 July 2021	29.6939	93.5435	2.54
PC10	15 July 2021	29.6933	93.5439	2.42
SP2	9 December 2022	29.1042	89.51	2.51
SP11 *	9 December 2022	29.1384	89.519	4.17
SP12	9 December 2022	29.0979	89.5106	3.13
SP22	9 December 2022	29.1031	89.512	3.9
SP23	9 December 2022	29.099	89.5105	2.43

Note: * cores taken outside the dredge pit.

). In Raccoon Island (RI) dredge pit (middle), 6 vibracores were collected on 19 August 2020, with 4 and 2 vibracores inside and outside the pit, respectively (Figure 1). In Sandy Point (SP) dredge pit (east), 5 vibracores were collected on 12 September 2022, with 4 and 1 vibracores inside and outside the pit, respectively. Vibracore methods have been successfully used in mud-capped dredge pits on the Louisiana shelf in the past [16,17]. All the vibracores were capped and sealed with electrical tape in the field and shipped to a lab at LSU and stored in a cold room. The cores were then extruded and split lengthwise every 20 cm. The top 1 cm of every 20 cm of sediment segment was then subsampled for grain size, loss-on-ignition (LOI) and carbonate analyses, as described below. All sediment samples were sealed in Whirl-Pak sampling bags before further analysis.

2.2. Grain Size, Organic Matter and Carbonate

A Beckman–Coulter laser diffraction particle size analyzer with an Aqueous Liquid Module, having measurement values ranging from 0.02 to 2000 µm, was used to analyze the vibracore samples. Sediment samples were prepared by mixing ~2 g of wet sample with 5~7 drops of 30% hydrogen peroxide in the test tubes. These tubes containing samples were put in water bath beakers at a temperature of 70 °C for about 3 days. After organic matter was digested completely, testing tubes were placed in a centrifuge for 4 min at a speed of 4000 RPM. Each sample was then mixed and sonicated prior to measurement. Grain size distribution plots were then generated using MATLAB R2024asoftware. Then the fractions of sand (>63 µm; phi < 4), silt (4–63 µm; phi is 4–8) and clay (<4 µm; phi > 8) were determined. More details of the grain size method are in [25].

Loss-on-ignition (LOI) analysis was performed on sediment samples to estimate both organic matter and carbonate contents [26]. Sediment samples of ~1.5 g were dried at 100 °C for 24 h, ground to powder, weighed and heated in a ceramic crucible using a muffle furnace for 3 h at 550 °C. Remaining samples were weighed, and organic matter (LOI_550°) was calculated using dry weights (DW) in Equation (1) [25].

Organic Matter% = (DW_100° − DW_550°)/DW_100°) × 100,

(1)

where DW_100° and DW_550° are the samples’ dry weight at 100 °C and 550 °C, respectively (both in g).

The remaining sediment samples were combusted again in a ceramic crucible using a muffle furnace for 2 h at 950 °C. The remaining sediments were then weighed, and the carbonate content (LOI_950°) was determined using Equation (2) below [24]:

Carbonate% = (DW550° − DW950°)/DW100°) × 100,

(2)

where DW550° and DW950° indicate the samples’ dry weight at 550 °C and 950 °C, respectively (both in g). All LOI samples were tested twice and the LOI of each sediment sample was an average of two duplicates.

The Geotek Multi-Sensor Core Logger (MSCL) (Geotek Ltd., Daventry, UK) was utilized to obtain the gamma density on the sediment vibracores collected for this study. Strong local variations in the density observed in a core often correlate to the variations in grain size. For instance, sand layers correspond with densities higher than those of mud layers. More details of the MSCL method are in [2].

A large surficial seabed sediment dataset for the northern Gulf of Mexico were downloaded from USGS’s usSEABED website [27]. Sand percent was extracted to generate a categorized sand percent map in ArcGIS.

2.3. Statistical Analysis

The lengths of vibracores collected in the field varied from site to site. The shortest core used in this study was 180 cm long. For minimizing bias and easier comparison among all the cores in three pits (PC, RI and SP), only the top 180 cm of each vibracore was used to perform the following statistical analyses. Instead of looking at the life cycle of pits, our research is focused on ‘recent’ sediment infilling in the topmost part of pits. A multiple linear regression model was implemented to predict grain size from depth, standard deviation (SD, also called sorting, a measure of the range of grain sizes and the magnitude of the spread), skewness (the level of symmetry), kurtosis (the ratio between the spread in the middle part of the distribution and the spread in the tails), organic matter, carbonate and MSCL gamma density for each pit. Moreover, standardized coefficients are computed and rounded with two decimals to achieve more accessible comparison among explanatory variables for each model. PROC REG procedure in SAS Enterprise Guide 7.1 with Stepwise Selection option was utilized to generate the multiple linear regression model results (https://support.sas.com/documentation/onlinedoc/guide/, accessed on 9 June 2024).

Furthermore, the generalized linear model was implemented to compare the grain size difference between paired positions. The significant level has been chosen as 0.05. If the p value is less than 0.05, there will be sufficient evidence to conclude that the grain size difference between the paired positions is statistically significant. PROC REG procedure with CLASS and LSMEANS statements in SAS Enterprise Guide 7.1 was utilized to generate the generalized linear model results.

Paired t-test on grain size data was performed to compare differences between Raccoon Island Inside and outside of pits, Peveto Channel Inside and outside of pits, and Sandy Point Inside and outside of pits vibracores. Correlation coefficients of the relationship between grain size carbonate content and organic matter.

Centered log-ratio (clr) transformations were performed on the closed grain size distribution (GSD) prior to multivariate statistical analysis. The centered log-ratio transformation divides each compositional part geometrically by the mean of all parts. It is written as

$$clr\left(x\right)i=\mathit{log}{\displaystyle \frac{xi}{g\left(x\right)}},i=1,\dots ,D,$$

$$g\left(x\right)={(\prod _{i=1}^{D}xi)}^{1/D}=\mathit{exp}({\displaystyle \frac{1}{D}}\sum _{i=1}^D\mathit{log}xi).$$

More details about centered log-ratio transformations are discussed in [28].

PCA was conducted in 12 grain size classes that were present in the sediment samples. They are listed in

Table 2. Grain size classes defined by phi (φ) intervals.

Class	Grain Size Interval (φ)
A	12.02–10.99
B	10.99–9.99
C	9.99–8.97
D	8.97–8.00
E	8.00–7.00
F	7.00–6.00
G	6.00–5.00
H	5.00–4.76
I	4.76–4.61
J	4.61–4.24
K	4.24–4.00
L	4.00–1.00

. This data-grouping method is referenced from [28,29]. All figure generation and statistical analyses were conducted using Python 3.11.

3. Results

The interquartile range (IQR) for grain size at the Sandy Point dredge pit is 3.1 phi inside the pit and 1.9 phi outside of it, indicating a broader spread and greater variability within the pit. Both inside and outside the pit, the grain size data exhibit positive skewness. In contrast, the Peveto Channel dredge pit shows IQRs of 0.4 phi inside and 0.9 phi outside, while the Raccoon Island dredge pit has IQRs of 0.8 phi inside and 1.5 phi outside. Notably, the grain size data outside both the Raccoon Island and Peveto Channel pits are positively skewed, whereas inside these pits, the data are negatively skewed. Overall, the Sandy Point dredge pit displays the largest IQR, suggesting that its grain size is more poorly sorted and highly variable. On Figure 2, only the grain size data from the Sandy Point dredge pit include sand and clay portion within IQR. In addition, Peveto Channel dredge pit grain size data have the smallest IOR spreading. The median grain size values of these six data groups are inside the range between 6 and 8 phi, corresponding to fine silt. Specifically, the grain size inside the Raccoon Island dredge pit has the smallest median value, while that inside the Peveto Channel pit has the biggest median value (Figure 2). The grain size data inside the Peveto Channel dredge pit have the most potential outliers.

PCA was conducted on 12 grain size classes present in the sediment samples from six datasets: inside and RI_out, RI_in and PC_out, and PC_in and SP_in and SP_out. Across the six data groups, the first and second principal components (PCs) capture varying proportions of the total variance. For the Raccoon Island inside group (RI_in), they account for 89.04% and 6.11%, respectively. In the Raccoon Island outside group (RI_out), these components explain 83.76% and 14.55% of the variance. The Peveto Channel inside group (PC_in) shows 76.44% and 21.81%, while the outside group (PC_out) has 76.21% and 20.01%. For the Sandy Point inside group (SP_in), the PCs cover 78.24% and 18.44%, and for the outside group (SP_out), they account for 84.33% and 12.42% of the variance (Figure 3).

Three multiple linear regression models with standardized coefficients have been generated for three pits (PC, RI and SP). The formula for the multiple linear regression model is

y = β₀ + β₁x₁ + β₂x₂ + β₃x₃ + β₄x₄ + β₅x₅ + β₆x₆ + β₇x₇ + ε.

In the established models, Y, X₁, X₂, X₃, X₄, X₅, X₆ and X₇ stands for grain size, sediment core depth, standard deviation, skewness, kurtosis, organic matter, carbonate and MSCL gamma density, respectively. Moreover, since the coefficients have been standardized, the estimated intercepts are all zeros. The models are presented as below for each position:

PC: Y = 0X₁ − 0.81X₂ − 0.63X₃ − 0.15X₄ + 0X₅ + 0X₆ − 0.10X₇,       R² = 0.9861,

RI: Y = 0X₁ − 0.33X₂ − 0.60X₃ − 0.53X₄ + 0.18X₅ + 0.10X₆ + 0X₇,   R² = 0.8706,

SP: Y = 0X₁ + 0.26X₂ − 0.76X₃ + 0X₄ + 0X₅ + 0X₆ + 0X₇,                  R² = 0.7798.

From the results of multiple linear regression models, the grain size can be predicted from depth, SD, skewness, kurtosis, organic matter, carbonate and density for each pit. In the PC, SD, skewness, kurtosis and density are negatively correlated to grain size. Meanwhile, the SD is the most influential explanatory variable based on the absolute value of the standardized coefficients (

Table 3. Standardized coefficients for each multiple linear regression model.

Pit	Grain Size	Depth	SD	Skew	Kurt	Organic Matter	Carbonate	Density	Intercept	RMSE	R-Square
PC			−0.81	−0.63	−0.15	0	0	−0.10	0	0.0822	0.9861
RI			−0.33	−0.60	−0.53	0.18	0.10	0	0	0.3678	0.8706
SP			0.26	−0.76	0	0	0	0	0	0.9242	0.7798

Note: SD = standard deviation; Skew = skewness; Kurt = kurtosis; RMSE = root mean square error.

). In the RI, SD, skewness and kurtosis are negatively correlated to grain size, while organic matter and carbonate are positively correlated to grain size. Meanwhile, skewness is the most influential explanatory variable based on the absolute value of the standardized coefficients (Table 3). In the SP, skewness is negatively correlated to grain size, while SD is positively correlated to grain size. Meanwhile, skewness is the most influential explanatory variable based on the absolute value of the standardized coefficients (Table 3). From the results of the generalized linear model, the grain size difference between paired positions can be qualitatively determined. The PC_in position has significantly higher grain size than the PC_out, RI_in, RI_out, SP_in and SP_out positions. The PC_out position has significantly higher grain size than the RI_out and SP_in positions. The SP_in position has significantly lower grain size than the SP_out position.

4. Discussion

4.1. Comparisons of Sediment Grain Sizes

A paired t-test, also known as a dependent samples t-test, is a statistical method used to determine if there is a significant difference between the means of two groups. The mutual p-values are greater than 0.05 for the grain size data from outside the Peveto Channel, Raccoon Island, and Sandy Point dredge pits (

Table 4. p value from paired t-test on grain size data from the Raccoon Island, Peveto Channel and Sandy Point dredge pits.

Grain Size	RI_in	RI_out	PC_in	PC_out	SP_in	SP_out	SP_edge
RI_in	0.3072	0.4814	<0.0001	0.0035	0.0881	0.0188	0.2223
RI_out	0.4814	0.7744	0.0004	0.2967	0.0306	0.138	0.1422
PC_in	<0.0001	0.0004	0.0046 or 0.6835	2.2184	1.199	0.0492	0.0009
PC_out	0.0035	0.2967	2.2184	/	0.0141	0.6808	0.0047
SP_in	0.13712	0.0306	1.199	0.0141	k/	0.0573	0.818
SP_out	0.0188	0.138	0.0492	0.6808	0.0573	/	0.0012
SP_edge	0.2223	0.1422	0.0009	0.0047	0.818	0.0012	0.8905

Note: Ps: ‘in’ means inside the pit, ‘out’ means outside the dredge pit. The p-values shaded are smaller than 0.05.

). It indicates that the seafloor grain size from Holly Beach in the west to the Mississippi River Delta in the east is not statistically different. This result supports the conclusion that the inner Louisiana continental shelf is mainly silt-dominated sediment. The p-values between grain size data inside the Raccoon Island pit and Peveto Channel pit are smaller than 0.05 which means the newly infilling sediments to these two dredge pits are statistically different (Table 4). However, the p-values between grain size data inside Raccoon Island and Sandy Point dredge pits are bigger than 0.05 which means their difference is not statistically significant. The p-values between grain size data inside the Peveto Channel and Sandy Point dredge pits are greater than 0.05 too.

4.2. Correlations Among Grain Size, Organic Matter and Carbonate

Grain size and organic matter from the Sandy Point dredge pit has a high correlation relationship, where R² = 0.834 (Figure 4. As discussed above, the IQR of grain size from the Sandy Point dredge pit has the biggest spreading among the other two dredge pits (Figure 2). In addition, the organic matter from the Peveto Channel dredge pit has the smallest IQR variation. At the same time, the grain size and organic matter from the Peveto Channel dredge pit have a medium–high-correlation relationship, R² = 0.527 (Figure 4). This could explain the smallest IQR spreading from the Peveto Channel dredge pit partially. In contrast to the sediment from the Sandy Point dredge pit, the correlation relationship between grain size, organic matter and carbonate from the Raccoon Island dredge pit have the smallest correlation relationship; specifically, the carbonate content has the weakest correlation with grain size and organic matter (Figure 4). This likely reflects variations in sediment sorting, resuspension and post-depositional reworking among the sites.

4.3. Sources and Transport of Sediment

The Peveto Channel, Raccoon Island and Sandy Point dredge pits span the inner Louisiana continental shelf, from Holly Beach in the west to the Mississippi bird-foot delta in the east (Figure 1). In addition to sediment transport on the Louisiana continental shelf serving as a primary control on regional sediment dynamics, factors such as surficial sand and mud distribution, proximity to river mouths, paleo-river channel presence, seafloor topography and extreme weather events may also influence sedimentation patterns within the three dredged pits.

It is well established that both sand and mud are present on Louisiana continental shelf (Figure 5). Mud is generally widely distributed in marshes, delta fronts and on the inner-middle shelf (Figure 5). Sand is mainly found near barrier islands, such as Isle Dernieres (Figure 5). Submarine shoals like Ship Shoal, Tiger Shoal and Trinity Shoal contain high quality sand suitable for use as borrow areas in coastal restoration projects. Both modeling and observational studies indicate that overall sediment transport along the inner to middle Louisianan shelf is directed westward (and [30]). The area east (and thus upstream) of the Peveto Channel pit is mainly the muddy inner shelf, without many nearby sandy sources. The area east of the Raccoon Island pit lies within a muddy “trough” situated between Ship Shoal and Isle Dernieres, both of which are sandy (Figure 5). In contrast, the area east of the Sandy Point pit is the Mississippi bird-foot delta, which contains muddy marshes, river sand bars, sandy barrier islands and a muddy delta front.

According to our grain-size distribution map (Figure 5), the inner Louisiana continental shelf is generally silt-dominated, with sand composition primarily controlled by proximity to sandy shoal and paleo-river channels ([23]; Figure 6), the two main sand sources on the inner shelf. This pattern reflects the spatial distribution of surficial sands and muds on the seafloor.

In Peveto Channel dredge pit, the composition of sand, silt and clay has fewer variations than the Sandy Point and Raccoon Island dredge pits (Figure 6). In addition, its IQR in grain size is the smallest (Figure 2). This demonstrates that the infilling sediment at the Peveto Channel dredge pit is from a stable and well-sorted source, i.e., the westward mud stream downdrift of the Atchafalaya Bay, controlled by current-wave-enhanced sediment gravity flows [32]. The Peveto Channel dredge pit sediment vibracores were collected in July 2021 (Table 1). Two major hurricanes, Delta (October 2020) and Laura (August 2020) passed through this area before the sediment sampling. However, there is not much sand portion variation. Therefore, it is likely that extreme weather events, such as hurricanes and cold fronts, did not play a defining role in altering the grain-size composition at the Peveto Channel pit due to the extensive mud surrounding this pit.

However, sand portion outside the Raccoon Island dredge pit ranges from 0% to 79% and the sand portion inside the Sandy Point dredge pit ranges from 5% to 97% (Figure 6). In Raccoon Island pit, the high sandy portion was found from the sediment cores outside the dredge pit. One possible reason is that the sandy portion comes from the nearby paleo-river channels. Another possible explanation is that the sandy portion comes from the submarine Ship Shoal. Reference [33] reported that multicores close to Ship Shoal have a higher sandy portion and vibracores have sandy layers which is possibly a result of hurricane induced sediment transport.

Different from the Raccoon Island pit, high sand portion was found inside the Sandy Point pit. Finer sediment accumulated inside the pit, as the topographic low created by dredging acts as an effective sediment trap. At the Sandy Point pit, sand was identified in two vibracores collected from the collapsed pit wall (Figure 1). These sand deposits likely originated from the collapse of an ambient paleo-river channel, while the rest of the fine-grained sediments were probably supplied by hydrodynamic activities. Therefore, the variation in sand content within the Sandy Point dredge pit is likely attributable to a combination of dredging activities and hurricane-induced transport. Notably, Hurricanes Laura (August 2020), Delta (October 2020), Zeta (October 2020) and Ida (August 2021) all passed near the Sandy Point pit in recent years.

4.4. Sources and Transport of Organic Matter and Carbonate

Organic matter (OM) in coastal sediments varies significantly depending on proximity to terrestrial sources, hydrodynamic energy and depositional environment. Terrestrial OM tends to dominate nearshore and marsh environments due to riverine input and wetland plant decay, often exceeding 50% in marsh sediments [33,34]. By contrast, sediments on the Louisiana continental shelf typically contain 1–5% OM as a result of offshore dilution, mineral input and reduced OM preservation [35,36]. The data from this study fall within, or slightly above, this shelf range but show significant spatial variability among dredge pits and between inside- vs. outside-pit locations (Figure S1). For instance, median OM concentrations are highest in the Sandy Point and Peveto Channel dredge pits (~8%) and lowest in Raccoon Island (~3–5%). Interestingly, the inside of the Raccoon Island pit shows slightly higher OM than the outside, possibly due to OM accumulation in a low-energy setting. Conversely, Sandy Point exhibit higher OM outside the pit, likely due to inputs from the Mississippi River plume and eroding marsh edges. The Peveto Channel shows relatively uniform OM between inside and outside, consistent with its proximity to the mainland (~7 km from Holly Beach) and possibly steady input of terrestrial OM. In addition, the Sandy Point dredge pit is about 9.6 km away from the coastal land, and the Raccoon Island dredge pit is about 10 km from the Isles Dernieres and about 18 km from the nearest coastal line (Figure 1). Therefore, the proximity to the nearby land could be a factor impacting organic matter content at the Peveto Channel dredge pit. Although the Sandy Point dredge pit has a similar relative distance to the coastal land nearby, it is the nearest distance to the Mississippi River mouth, about 30 km (Figure 1). The river flow with high dissolved organic matter (DOM) and particulate organic matter (POM) from various sources such as decaying plant material, animal waste and other organic debris, could be the reason for the high organic matter around the Sandy Point dredge pit.

Carbonate content across all pits is uniformly low (mostly 1–2%) and often approaches analytical detection limits, suggesting it plays a minimal role in sediment geochemistry. These low values are consistent with previous studies indicating that the Louisiana shelf is dominated by terrigenous input, with limited shells or modern biogenic carbonate production [37]. Visual inspection of cores did not reveal significant shell material, suggesting that any carbonate present is likely in the form of fine shell hash or reworked particles rather than in situ biological production. Variability in carbonate is the highest at Sandy Point (especially inside the pit), which may be linked to broader grain size distribution caused by dredging and slope failure events (Figure 2 and Figure S2). Meanwhile, carbonate shows weak correlation with both OM and grain size at Raccoon Island, further supporting its limited environmental relevance to the parameters analyzed in this study. Previous work in the region suggests that carbonate saturation states are more strongly influenced by seasonal productivity and stratification than by sediment input [38], which may explain some of the subtle variations observed here. Overall, carbonate appears to be a small component of these sediment systems on the Louisiana inner shelf.

5. Conclusions

A total of eighteen vibracores from the Raccoon Island dredge pit, Peveto Channel dredge pit and Sandy Point dredge pit were used in this paper to study their grain size, density, organic matter and carbonate content. The major conclusions are as follows:

(1)

The inner Louisiana continental shelf is silt-dominated. Its sand composition is controlled by its proximity to sandy shoals and paleo-river channels and the extreme weather conditions are the driving forces for sand portion transport. Hurricanes transport the sandy sediment to its ambient muddy environment, while the dredging activities create a trap for sediment infilling.

(2)

Organic matter and grain size (phi) have a strong correlation in the sediment from the Sandy Point dredge pit (R² = 0.834). Since grain size (phi) and grain size (um) are inversely related, the finer the grain size in um, the higher the organic matter content. The gamma density data from outside these three dredge pits are similar; however, the gamma density data inside the pits are different.

(3)

The PCA results confirm two principal components that account for more than 95% of the total grain size variance. Overall, the regression analysis demonstrated that specific variables—namely standard deviation and skewness—consistently played key roles in predicting grain size, with variations in their strength and direction across positions. Skewness emerged as a prominent factor in the RI and SP, while standard deviation was the most influential factor in the PC. Density was found to be particularly important in the PC, highlighting the varying influence of density on grain size in different environments. Moreover, organic matter and carbonate play a crucial role in increasing the grain size in RI, corresponding to specific geological settings.

(4)

Our results enhance the scientific community’s ability to make informed plans and decisions that support the safe, environmentally responsible and efficient management of offshore mud and sand resources in and around dredge pits, while also protecting environmental assets and minimizing potential conflicts with energy infrastructure.