1. Introduction
Saharan dust, a phenomenon characterized by the transport of fine mineral particles from the arid regions of the Sahara Desert in North Africa across vast distances, exerts substantial influences on ecosystems and human activities in the tropical North Atlantic and the greater Caribbean region [
1]. This natural phenomenon, also known as the Saharan Air Layer (SAL), has gained increasing attention due to its far-reaching impacts on air quality, weather patterns, and various sectors such as agriculture, health, and tourism [
2,
3,
4,
5,
6]. The Saharan dust episodes are often associated with elevated concentrations of particulate matter (
Figure 1) and the presence of various pollutants, which can lead to respiratory issues and reduced visibility [
7]. Additionally, these dust particles can interact with cloud formation and solar radiation, potentially influencing local weather and climate [
8].
As global temperatures rise due to climate change, the Sahara Desert experiences altered wind patterns and increased aridity, intensifying dust emissions [
9]. This higher dust load can affect the radiative balance of the Earth by absorbing or reflecting sunlight, potentially contributing to regional and even global climate shifts [
10,
11]. Furthermore, the nutrient content within the transported dust can fertilize marine ecosystems, including the pelagic
Sargassum in the tropical Atlantic [
1,
12,
13]. While the exact mechanisms are still being studied, it is suggested that iron and other dust minerals might enhance
Sargassum growth.
Saharan dust can have positive effects on
Sargassum. The deposition of Saharan dust on the ocean surface can provide nutrients, including iron and phosphorus, which are essential for the growth of
Sargassum [
14,
15]. These dust-derived nutrients act as fertilizers, stimulating primary productivity and potentially enhancing
Sargassum growth rates. Iron, in particular, is a micronutrient that can be limiting in some oceanic regions, and the input of iron-rich Saharan dust can alleviate this limitation for
Sargassum and other marine organisms [
16].
In recent years, unprecedented SAL events have coincided with
Sargassum blooms in the tropical waters of the North Atlantic, but little is known about whether Saharan dust deposition contributes to higher concentrations of the algae. Research on the relationship between SAL and pelagic
Sargassum in the tropical North Atlantic has shed light on their interactions and potential influences, with some studies showing that these dust particles can affect the growth and distribution of the algae (12), even triggering the appearance of large-scale blooms in the Caribbean Sea [
17,
18].
Some studies suggest that atmospheric dust, which contains essential nutrients such as iron, phosphorus, and nitrogen, can enhance the growth of marine macroalgae, including
Sargassum [
19]. Dust-borne phosphorus also plays a crucial role in marine productivity [
20]. There is empirical evidence supporting the hypothesis that nutrient-rich dust can stimulate algal blooms, identifying a correlation between dust transport events and increased
Sargassum growth in the Caribbean Sea [
21]. Such findings suggest that the deposition of nutrient-rich dust may be a significant factor in the proliferation of
Sargassum.
Other factors that are important drivers of
Sargassum blooms are sea surface temperature and high nutrient concentrations. Higher sea surface temperatures can promote the proliferation of
Sargassum as warmer waters provide optimal conditions for its growth and reproduction [
22]. On the other hand, higher chlorophyll concentrations in the water column can indicate regions of enhanced primary productivity and nutrients, which may be associated with favorable conditions for the growth and proliferation of
Sargassum populations [
23]. In combination with Saharan dust, sea surface temperatures and elevated nutrient concentrations could explain why some periods reflect higher
Sargassum blooms.
For that reason, the objective of this study is to further examine the relation between SAL events and Sargassum blooms to determine if higher dust deposition leads to larger algae growth in the area. This study examines temporal trends in Saharan dust events in the Eastern Caribbean to determine if high-concentration SAL occurrences have increased over time and if there is any connection with anomalous Sargassum blooms in the region. After examining trends in SAL events in the tropical waters of the North Atlantic, the focus shifts to examining whether those events coincide with Sargassum outbreaks in the region.
2. Data & Methods
Average dust mass column density (DMCD) data—which refers to the total mass per unit area of a specific substance, such as aerosols or gasses, extending vertically from the Earth’s surface to the top of the atmosphere—were obtained from MERRA-2 (Modern-Era Retrospective Analysis for Research and Applications, version 2), a state-of-the-art atmospheric reanalysis model developed by NASA. This variable is calculated by integrating the atmospheric concentration of the aerosols over the entire vertical column above a given point on the Earth’s surface. The integration involves accounting for factors such as the vertical distribution of the dust and aerosols, the density of the air, and the altitude variations within the column. The monthly average MCD data were extracted for the Eastern Caribbean region of the tropical North Atlantic (−80.5, 10, −60, 23.5) for the entire period 1980–2022 (
Figure 2A). The analysis was limited to that section of the study area to compare DMCD concentrations with
Sargassum coverage areas (
Figure 2B). The DMCD data have been used by previous researchers examining trends and patterns in Saharan dust concentrations [
24,
25,
26].
Monthly
Sargassum concentrations (SCT) in tons for a section of the Eastern Caribbean (−80.5, 10, −60, 23.5) were obtained from the Optical Oceanography Lab at the University of South Florida [
27]. The monthly average
Sargassum in tons was extracted for June, July, and August (JJA), which have been documented to be the period with higher concentrations of the algae [
1,
12]. The temporal resolution of the
Sargassum data is from 2010 through 2022. Mann–Kendall (MK) tests for trends were performed on the DMCD and SCT to determine if the dust and
Sargassum concentrations during the JJA period or individual months were increasing over time. The MK test is a widely used statistical test in the atmospheric sciences that allows for robust trend detection, and several studies have implemented the method to analyze trends in dust aerosols [
28,
29,
30].
Mean sea surface temperature (SST) and chlorophyll (CHLO) data were also extracted for the study area to determine if those factors, in combination with DMCD, could also explain SCT variability. Mean SST data for June, July, and August were obtained from the NOAA Extended Reconstruction SSTs Version 5. Average chlorophyll concentrations for the same months and region were extracted from the Moderate-Resolution Imaging Spectroradiometer (MODIS).
The DMCD and SCT data were analyzed using simple linear regression models to determine if higher dust amounts coincided with higher Sargassum concentrations in the study area. The regression models between Saharan dust and Sargassum concentrations were performed on the individual months and the JJA period. The simple linear regression models were also performed with a lag, this was done to consider the possibility that previous Saharan dust deposition in the ocean would later enhance Sargassum growth in a later period. This type of regression is particularly useful when you suspect that the current value of a variable depends on its past values; for example, in this case, we hypothesize that higher Saharan dust deposition leads to higher Sargassum concentrations.
A one-month lag between dust arrival and
Sargassum concentration is justified as nutrients from dust, like iron and phosphorus, take several weeks to be absorbed and utilized. This aligns with observations that phytoplankton blooms occur weeks after nutrient deposition.
Sargassum mats can persist in regions, allowing time to utilize these nutrients [
18,
31].
Forward stepwise regression (FSR) procedures were used to examine how important DMCD was to SCT variability when compared to other factors like SST and CHLO. FSR is a statistical method that sequentially adds predictor variables to a regression model based on their contribution to improving the model’s fit, aiming to identify the most relevant factors for explaining the variability in the response variable. In this case, SCT is the response variable that we are interested in predicting using the predictors DMCD, SST, and CHLO.
3. Results and Discussion
3.1. Trends in Saharan Dust and Sargassum Concentrations
When trends in monthly and seasonal average DMCD concentrations are examined, we find that Saharan layer dust quantities in the atmosphere of the tropical North Atlantic have been increasing over time (
Figure 3A). When the summer months of JJA are analyzed together, the results show a steadily increasing trend of mean DMCD for the tropical North Atlantic for the 1980–2022 period (
Figure 3A). The years with the highest mean DMCD concentrations were 2018 and 2020. Of the three summer months (JJA) with higher mean DMCD concentrations, June was found to have the most significant increasing trend in Saharan dust from 1980 through 2022 (
Figure 3A). For the June series, 2019 was found to be the year with the highest mean DMCD. The individual months of July and August show increasing trends in average DMCD concentrations (
Figure 3A), yet those were not as significant as the positive trend in Saharan dust quantities that June reflected. Of all of the years analyzed, July 2018 was found to be the period with the highest average DMCD. These results confirm what some studies have been suggesting, that Saharan layer dust concentrations have been increasing over time in the tropical atmosphere of the North Atlantic [
9].
The time series plots for
Sargassum concentrations also suggest that the algae have been experiencing an increasing trend in total quantity and area in the region (
Figure 3B). The results for the JJA period show that
Sargassum concentrations have been increasing from 2010 through 2022, with 2018 being the year with the highest agglomerations detected (
Figure 3B). It is important to note that 2018 was also the year with the highest DMCD concentrations in the region, so both the highest
Sargassum agglomerations and Saharan dust quantities coincided during that period. When the individual months are examined, we find that all of them exhibited significant increasing trends in
Sargassum concentration (
Figure 3B). June and July of 2018 and August of 2021 were the periods with the highest mean concentrations of
Sargassum in the region (
Figure 3B). These results coincide with the findings of other studies that have detected a recent increase in
Sargassum concentrations and total area covered [
18], which might be associated with higher dust deposition from the Sahara [
12].
When both DMCD and SCT time series are analyzed using Mann–Kendall tests for trends, we find that Saharan dust and
Sargassum concentrations have been increasing over time in the tropical atmosphere of the North Atlantic (
Table 1). The month of June was found to have the most statistically significant increase in DMCD and SCT concentrations. When the three months are examined together, the results show that both Saharan dust and
Sargassum concentrations reflect statistically significant trends. Even though July and August show weaker increasing trends in mean DMCD when compared to June and the JJA period, they still exhibit statistically significant trends (
Table 1). The Mann–Kendall results for SCT show that all of the individual months and the JJA period exhibited statistically significant increasing trends in
Sargassum concentrations, with June showing the strongest trend.
3.2. Statistical Modeling Results
The results of the simple linear regression models between DMCD and SCT show an overall positive correlation between the two variables (
Figure 4); as Saharan dust concentrations increase, so do
Sargassum concentrations. When the analysis is done for the JJA (
Figure 4A), the results show a very strong positive correlation (R2 0.429) between DMCD and SCT. The results for the individual months show the month of July (
Figure 4C) with the strongest correlation, while weaker correlations were found for June and August (
Figure 4B,D). These findings suggest that the relationship between Saharan dust and
Sargassum concentrations has a particular temporal component, with weaker correlations in June and August, and stronger correlations in July and the JJA period overall.
The results of the lagged regression models also show positive correlations between DMCD and SCT (
Figure 4E,D). When the lagged regression was done using DMCD for June and SCT for July, we found a positive yet weak correlation between Saharan dust and
Sargassum (
Figure 4E). However, when the analysis was performed using DMCD for July and SCT for August, the results yielded a stronger positive correlation between the two factors (
Figure 4D). These findings suggest that higher dust concentrations in June do not necessarily materialize into higher
Sargassum quantities in July. Yet, greater amounts of dust in July show a stronger statistical relationship with elevated
Sargassum concentrations. These results are similar to those found in previous studies [
12,
14], showing that higher Saharan dust concentrations likely enhanced some of the recent
Sargassum outbreaks in the region.
The results of the forward stepwise regression (FSR) models suggest that DMCD is one of the most important predictors of SCT in the Eastern Caribbean region that was analyzed in this study (
Table 2). The FSR results for June selected average CHLO in May as the most important factor behind higher mean SCT, suggesting that elevated chlorophyll concentrations often indicate increased phytoplankton growth. While phytoplankton and Sargassum are both primary producers and compete for nutrients, the presence of phytoplankton, indicated by elevated chlorophyll levels, suggests nutrient-rich conditions that can support Sargassum growth [
18]. FSR results for July and August show DMCD as the most important predictor of SCT, suggesting that higher Saharan dust deposition during those months is the key driver of higher
Sargassum concentrations (
Table 2). The FSR results for the JJA period identified both CHLO and DMCD as the most important factors that were examined, explaining higher SCT in the region. CHLO alone accounted for a third of the variability in SCT, while the addition of DMCD into the model increased the adjusted R-square to 0.563. These results suggest that Saharan dust and chlorophyll are the most important predictors of SCT in the Eastern Caribbean region. It is important to note that SST was not selected as an important factor in the FSR models, suggesting that ocean temperatures in this region might not be as critical to higher
Sargassum concentrations as Saharan dust and chlorophyll.
4. Conclusions
In this study, the average Saharan dust mass concentration densities (DMCD) and mean Sargassum concentrations (SCT) in the tropical North Atlantic were analyzed to identify any coinciding increasing trends over the examined period. Mann–Kendall tests for trends and lagged regression models were used to examine the relationship between DMCD and SCT to determine if higher Saharan dust concentrations were associated with Sargassum outbreaks in the region. The analysis aimed to understand the potential impact of Saharan dust on the proliferation of Sargassum.
The analysis revealed a significant upward trend in Saharan dust quantities from 1980 to 2022, particularly during the summer months of June, July, and August (JJA). Notably, the years 2018 and 2020 recorded the highest levels of DMCD, with June exhibiting the most significant increasing trend, peaking in 2019. These findings are consistent with existing studies that indicate a continuous rise in Saharan dust concentrations in the tropical atmosphere of the North Atlantic, suggesting an ongoing elevation of dust levels in the region.
Simultaneously, Sargassum concentrations showed a parallel increasing trend, with 2018 marked by both peak SCT agglomerations and heightened DMCD concentrations. The Mann–Kendall tests confirmed statistically significant increasing trends in both Saharan dust and Sargassum concentrations. Additionally, simple linear regression models and lagged regression analyses established positive correlations between DMCD and SCT, revealing a temporal component with stronger correlations observed in July and the overall JJA period. These results support the hypothesis that Saharan dust contributes to Sargassum outbreaks in the tropical North Atlantic.
While this study provides valuable insights into the relationship between Saharan dust and Sargassum blooms, several limitations should be considered. The use of monthly average data may obscure short-term variations and specific events that could offer a more nuanced understanding of the relationship between dust deposition and Sargassum growth. The study’s focus on a specific section of the tropical North Atlantic may not capture the full spatial variability of both dust deposition and Sargassum blooms across the entire region.
Moreover, while correlations between DMCD and SCT were identified, the specific nutrient dynamics and biological mechanisms by which Saharan dust influences Sargassum growth remain complex and are not fully explored. Other factors, such as ocean currents, local nutrient sources, and anthropogenic impacts, were not comprehensively analyzed and could also play significant roles.
Future research should address these limitations by incorporating higher temporal resolution data, expanding the spatial scope of analysis, and investigating the detailed nutrient dynamics and biological mechanisms involved. Additionally, exploring the interplay between Saharan dust, local nutrient sources, and other environmental factors will provide a more comprehensive understanding of the drivers behind Sargassum blooms.
Author Contributions
Conceptualization, J.J.H.A. and R.M.-T.; methodology, J.J.H.A.; software, J.J.H.A.; validation, J.J.H.A. and R.M.-T.; formal analysis, J.J.H.A.; investigation, J.J.H.A. and R.M.-T.; resources, R.M.-T.; data curation, J.J.H.A.; writing—original draft preparation, J.J.H.A. and R.M.-T.; writing—review and editing, J.J.H.A. and R.M.-T.; visualization, J.J.H.A.; supervision, R.M.-T.; project administration, R.M.-T.; funding acquisition, R.M.-T. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Institutional Review Board Statement
Not applicable.
Data Availability Statement
Conflicts of Interest
The authors declare no conflict of interest.
References
- Méndez-Tejeda, R.; Rosado-Jimenez, G.A. Influence of climatic factors on Sargassum arrivals to the coasts of the Dominican Republic. J. Oceanogr. Mar. Sci. 2019, 10, 22–32. [Google Scholar] [CrossRef]
- Prospero, J.M.; Ginoux, P.; Torres, O.; Nicholson, S.E.; Gill, T.E. Environmental characterization of global sources of atmospheric soil dust identified with the Nimbus 7 Total Ozone Mapping Spectrometer (TOMS) absorbing aerosol product. Rev. Geophys. 2002, 40, 1002. [Google Scholar] [CrossRef]
- Bergametti, G.; Liousse, C.; Dulac, F. Composition of aerosol collected in the free troposphere during the ACE 2 project. Tellus B. Chem. Phys. Meteorol. 2007, 59, 368–377. [Google Scholar]
- Monteil, M.A. Saharan dust clouds and human health in the English-speaking Caribbean: What we know and don’t know. Environ. Geochem. Health 2008, 30, 339–343. [Google Scholar] [CrossRef]
- Akpinar-Elci, M.; Martin, F.E.; Behr, J.G.; Diaz, R. Saharan dust, climate variability, and asthma in Grenada, the Caribbean. Int. J. Biometeorol. 2015, 59, 1667–1671. [Google Scholar] [CrossRef]
- Xu-Yang, Y.; Dessert, C.; Losno, R. Atmospheric deposition over the Caribbean region: Sea salt and Saharan dust are sources of essential elements on the island of Guadeloupe. J. Geophys. Res. Atmos. 2022, 127, E2022jd037175. [Google Scholar] [CrossRef]
- Tagliabue, A.; Bowie, A.R.; Boyd, P.W.; Buck, K.N.; Johnson, K.S.; Saito, M.A. The integral role of iron in ocean biogeochemistry. Nature 2017, 543, 51–59. [Google Scholar] [CrossRef]
- Gyan, K.; Henry, W.; Lacaille, S.; Laloo, A.; Lamsee-Ebanks, C.; McKay, S.; Monteil, M.A. African dust clouds are associated with increased pediatric asthma accident and emergency admissions on the Caribbean island of Trinidad. Int. J. Biometeorol. 2005, 49, 371–376. [Google Scholar] [CrossRef]
- Sakhamuri, S.; Cummings, S. Increasing trans-Atlantic intrusion of Sahara dust: A cause of concern? Lancet Planet. Health 2019, 3, e242–e243. [Google Scholar] [CrossRef]
- Tanré, D.; Haywood, J.; Pelon, J.; Léon, J.F.; Chatenet, B.; Formenti, P.; Myhre, G. Measurement and modeling of the Saharan dust radiative impact: Overview of the Saharan Dust Experiment (SHADE). J. Geophys. Res. Atmos. 2003, 108. [Google Scholar] [CrossRef]
- Slingo, A.; Ackerman, T.P.; Allan, R.P.; Kassianov, E.I.; McFarlane, S.A.; Robinson, G.J.; Dewitte, S. Observations of the impact of a major Saharan dust storm on the atmospheric radiation balance. Geophys. Res. Lett. 2006, 33. [Google Scholar] [CrossRef]
- Lapointe, B.E.; Brewton, R.A.; Herren, L.W.; Wang, M.; Hu, C.; McGillicuddy Jr, D.J.; Morton, P.L. Nutrient content and stoichiometry of pelagic Sargassum reflects increasing nitrogen availability in the Atlantic Basin. Nat. Commun. 2021, 12, 3060. [Google Scholar] [CrossRef]
- Skliris, N.; Marsh, R.; Appeaning Addo, K.; Oxenford, H. Physical drivers of pelagic Sargassum bloom interannual variability in the Central West Atlantic over 2010–2020. Ocean. Dyn. 2022, 72, 383–404. [Google Scholar] [CrossRef]
- Baker, A.R.; Jickells, T.D.; Witt, M. Atmospheric deposition of soluble iron to the global oceans: Implications for iron limitation of phytoplankton. Annu. Rev. Mar. Sci. 2020, 12, 523–550. [Google Scholar]
- Meskhidze, N.; Xu, J.; Gantt, B. Iron in atmospheric mineral dust: Biogeochemical cycling and impact on ocean productivity. Glob. Biogeochem. Cycles 2021, 35, e2020GB006864. [Google Scholar]
- Liriano-Gómez, R.E.; Gómez, A.C.; Bolívar, M.G.; González-Hernández, Y.; Jauregui-Haza, U.J. Characterization of Sargassum accumulated on Dominican beaches in 2021: Analysis of heavy, alkaline, and alkaline-earth metals, proteins, and fats. Mar. Pollut. Bull. 2023, 193, 115120. [Google Scholar] [CrossRef]
- Wang, M.; Hu, C.; Barnes, B.B.; Mitchum, G.T. The expanding Sargassum horizons: Mechanisms, detection, and management. Remote Sens. Environ. 2019, 231, 111215. [Google Scholar] [CrossRef]
- Johnson, M.T.; Hu, C.; Schollaert Uz, S.; Zhao, J.; Muller-Karger, F.E. The Ecology of Harmful Algal Blooms in the Caribbean and Western Atlantic. Front. Mar. Sci. 2020, 7, 342. [Google Scholar]
- Measures, C.I.; Vink, S. On the use of dissolved aluminum in surface waters to estimate dust deposition to the ocean. Glob. Biogeochem. Cycles 2000, 14, 317–327. [Google Scholar] [CrossRef]
- Mahowald, N.M.; Jickells, T.D.; Baker, A.R.; Artaxo, P.; Benitez-Nelson, C.R.; Bergametti, G.; Bond, T.C.; Chen, Y.; Cohen, D.D.; Herut, B.; et al. Global distribution of atmospheric phosphorus sources, concentrations and deposition rates, and anthropogenic impacts. Glob. Biogeochem. Cycles 2008, 22, GB4026. [Google Scholar] [CrossRef]
- Gower, J.; Young, E.; King, S. Satellite Images Suggest a New Sargassum Source Region in 2011. Remote Sens. Lett. 2013, 4, 764–773. [Google Scholar] [CrossRef]
- Djakouré, S.; Araujo, M.; Hounsou-Gbo, A.; Noriega, C.; Bourlès, B. On the potential causes of the recent Pelagic Sargassum blooms events in the tropical NorthAtlantic Ocean. Biogeosciences Discuss. 2017, 1–20. [Google Scholar]
- Oviatt, C.A.; Huizenga, K.; Rogers, C.S.; Miller, W.J. What nutrient sources support anomalous growth and the recent Sargassum mass stranding on Caribbean beaches? A review. Mar. Pollut. Bull. 2019, 145, 517–525. [Google Scholar] [CrossRef]
- Asutosh, A.; Vinoj, V.; Murukesh, N.; Ramisetty, R.; Mittal, N. Investigation of June 2020 giant Saharan dust storm using remote sensing observations and model reanalysis. Sci. Rep. 2022, 12, 6114. [Google Scholar] [CrossRef] [PubMed]
- Bencherif, H.; Bounhir, A.; Bègue, N.; Millet, T.; Benkhaldoun, Z.; Lamy, K.; Gadouali, F. Aerosol Distributions and Sahara Dust Transport in Southern Morocco from Ground-Based and Satellite Observations. Remote Sens. 2022, 14, 2454. [Google Scholar] [CrossRef]
- Hu, C.; Zhang, S.; Barnes, B.B.; Xie, Y.; Wang, M.; Cannizzaro, J.P.; English, D.C. Mapping and quantifying pelagic Sargassum in the Atlantic Ocean using multi-band medium-resolution satellite data and deep learning. Remote Sens. Environ. 2023, 289, 113515. [Google Scholar] [CrossRef]
- Kang, L.; Huang, J.; Chen, S.; Wang, X. Long-term trends of dust events over the Tibetan Plateau during 1961–2010. Atmos. Environ. 2016, 125, 188–198. [Google Scholar] [CrossRef]
- Gavrouzou, M.; Hatzianastassiou, N.; Gkikas, A.; Korras-Carraca, M.B.; Mihalopoulos, N. A global climatology of dust aerosols based on satellite data: Spatial, seasonal and inter-annual patterns over the period 2005–2019. Remote Sens. 2021, 13, 359. [Google Scholar] [CrossRef]
- Jafari, R.; Amiri, M.; Asgari, F.; Tarkesh, M. Dust source susceptibility mapping based on remote sensing and machine learning techniques. Ecol. Inform. 2022, 72, 101872. [Google Scholar] [CrossRef]
- Jickells, T.D.; An, Z.S.; Andersen, K.K.; Baker, A.R.; Bergametti, G.; Brooks, N.; Cao, J.; Boyd, P.W.; Duce, R.A.; Hunter, K.A.; et al. Global Iron Connections Between Desert Dust, Ocean Biogeochemistry, and Climate. Science 2005, 308, 67–71. [Google Scholar] [CrossRef]
- Lapointe, B.E. A Comparison of Nutrient-Limited Productivity in Sargassum natans from Different Water Masses in the Western North Atlantic Ocean. Limnol. Oceanogr. 1995, 40, 625–633. [Google Scholar] [CrossRef]
| Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).