A systematic review of floating and beach landing records of Sargassum beyond the Sargasso Sea

Sargassum algal blooms on ocean surfaces and landings of huge Sargassum mats on beaches is an emerging global environmental challenge with wide socio-economic and environmental implications. Literature on Sargassum growth cycles, travel patterns, species and morphotypes, and quantified impacts have tended to focus on a geographic region, or a specific event. Few, if any, publications document long term continuous monitoring of Sargassum blooms in large areas such as the Pacific, or the tropical Atlantic. To address this gap, this paper systematically reviews the global evidence of Sargassum bloom monitoring beyond the Sargasso Sea, and identifies gaps in the evidence base of floating and landing influxes. This systematic review uses combinations of two key terms relating to Sargassum and monitoring, and utilises the resources in ISI Web of Knowledge, Scopus and Google Scholar. The analysis moves us past a classic literature review, and produces an unbiased assessment of empirical research on Sargassum monitoring from 1960 to 2019. We find a significant research focus on open-ocean blooms and floating mats whereas research on beach landings and their associated impacts is comparatively limited. Research is focused within specific countries or water bodies (notably, the Gulf of Mexico, the Caribbean and North Atlantic Ocean) and tends not to comprehensively assess neighbouring or regional shorelines, for example, West Africa and Central America. There was a lack of consistency in the application of methods for quantifying Sargassum biomass volume (including dry/wet weight, unit of measurement, and spatial extent of calculations). Further, in many publications Sargassum species identification was omitted. Given current attempts to understand the drivers and impacts of the exponential growth in Sargassum in some parts of the world, a consistent and replicable research approach to monitoring Sargassum could support creation of a Sargassum evidence database. To move this agenda forwards, we propose a definition for a Sargassum ‘event’: a continuous bloom of any Sargassum in open oceans, or, an aggregation of landed Sargassum, with the potential to disrupt social, economic or ecosystem functioning, or to impact human health. This review highlights the importance of standardising Sargassum monitoring methods to facilitate improved documentation of temporal and spatial patterns of Sargassum blooms and beach landings.


Introduction
Pelagic Sargassum seaweed was first reported in the Sargasso Sea in the 15th Century and has since been documented in this area (Fine 1970, Lapointe 1986, Wang et al 2019. In recent years, research has described significant pelagic Sargassum seaweed blooms (free-floating brown seaweeds) across water bodies and beach landings globally, presenting a new environmental challenge (Langin 2018). Sargassum blooming events in the Atlantic are thought to be initiated and influenced by a combination of factors including: nutrient discharge from the Amazon River, changes in ocean upwelling, higher sea surface temperatures and Ocean pattern changes, such as an unusually strong North Atlantic Oscillation patterns, Atlantic Multidecadal Oscillation and El Niño-Southern Oscillation events (Sissini et al 2017, Sanchez-Rubio et al 2018, Oviatt et al 2019, Wang et al 2019, Johns et al 2020. In East Asia, blooms are thought to be caused when Sargassum is detached from beds due to strong waves and currents (Komatsu et al 2014). Some publications have assessed annual or seasonal cycles and transport patterns of Sargassum in the Atlantic Ocean and indicated potential origin sources of the blooms (examples include Gower and King 2011, Wang and Hu 2016, Brooks et al 2018, Putman et al 2018, Sanchez-Rubio et al 2018. However, there is no comprehensive study assessing the spatial distribution of Sargassum bloom events. Since 2011, Sargassum blooms appear to have increased in frequency and magnitude, notably in the tropical Atlantic and the Caribbean region (Wang and Hu 2017). In 2018, Wang et al (2019) estimated there was over 20 million metric tons of Sargassum across the Tropical Atlantic in the summer months.
The socio-economic impacts of Sargassum blooms and beach landings are notable on the aquaculture and tourist industries; for example, Sargassum clogs fishing gear and limits fishing ground, resulting in a reduction in revenue and income and an increase in maintenance costs (Solarin et al 2014, Ramlogan et al 2017. Xing et al (2017) estimated that, in China, seaweed damage cost the aquaculture industry 73 million USD. Additionally, tourism has decreased due to the visual impact and odour of Sargassum (Chávez et al 2020a). There are claims that decomposition of Sargassum releases toxic gases and can cause potentially fatal health problems in humans (ANSES 2017, Resiere et al 2018). Environmental impacts of Sargassum blooms have also been observed; for example, turtles looking to nest and neonate hatchlings accessing the sea can be hindered by Sargassum beach landings (Maurer et al 2015). Additionally, surface blooms restrict light penetration through the water column which affects benthic communities (McGlathery 2001). Despite the negative impacts on communities, Sargassum influxes also present opportunities for economic benefit as it has a variety of potential uses including for biofuel energy, soil fertiliser and animal feed, construction blocks, bioplastics and pharmaceutical products (Milledge et al 2016, Chávez et al 2020b, Thompson et al 2020. Large Sargassum influxes are generating high levels of concern among policy makers due to their impacts on economies, health, and society. Internationally, the United Nations Environmental Programme (UNEP) has created a Working Group on Sargassum within the Joint Group of Experts on the Scientific Aspects of Marine Environmental Protection (GESAMP) to identify key Sargassum challenges and responses. Regionally, conventions have been developed to acknowledge the issue and highlight the need for solutions; in the Tropical Atlantic this can be seen within the UN's Abidjan and Cartagena Conventions (UNEP 2018). However, critical information that could help policy makers and communities cope better with Sargassum is missing. For example, there is a lack of data on: the drivers of Sargassum-specifically input sources for specific regions; the temporal and spatial spread of significant Sargassum bloom events; the quantity of Sargassum in oceans and landing on beaches; and the distribution of species and morphotypes of Sargassum within bloom events (which can influence re-use opportunities).
This paper aims to contribute to our understanding of the spatial and temporal distribution of Sargassum by exploring the spectrum and prevalence of use of methods employed to monitor blooms and beach landings. Methods to document Sargassum are categorised herein as either remote sensing-based or in situ. Remote sensing methods include applying algorithms to identify Sargassum blooms in airborne and spaceborne imagery, largely focusing on surface blooms in open ocean areas. In-situ methods encompass site visits, often by boat, to survey or take samples of Sargassum blooms or deposits. Understanding which methods are being used, how effective they are, where and how they are considered, can help support the development of higher quality monitoring globally.
The paper is structured as follows: section 2 introduces the systematic review method-a methodology designed to reduce the potential for bias in a traditional literature review. Results are presented in section 3, focussed around the key questions asked in the systematic review, notably: Is the research related to floating or landing events, and if so where did it occur (beach, near-shore or open-ocean)?; What is the spatial extent of Sargassum research? Does the literature indicate a source of Sargassum?; What was the temporal scale of analysis?; Which species of Sargassum were identified?; What data has been used to document the occurrence of Sargassum blooms and beach landings? What was the volume of Sargassum (if calculated)? The key findings are discussed in section 4, around the main themes found in the systematic review. Section 5 concludes the paper, drawing out remaining research gaps. . The systematic review method offers a robust methodology to identify and analyse empirical published evidence, however, it has limitations. Some publications will be omitted due to: the way in which the search engines index their results (see (Beel and Gipp 2010)); because the journals appear only in print form and not online; or because some search engines, e.g. ISI WOK, quality controls their collections and only includes longestablished journals. Nonetheless, it remains a highly used and robust research tool (Moher et al 2016).

Method
For this paper, the systematic review method was used to identify all empirical research on Sargassum monitoring from 1960 to 2019. Multiple systematic review researchers recommend using trusted high-quality academic data bases, such as Scopus, or ISI Web of Knowledge (ISI WOK), see (Webb et al 2015). For example, Berrang-Ford et al (2011) justify the use of ISI WOK as it is powerful and comprehensive; Falagas et al (2008) use Scopus as it offers a wider journal range. Our search was supplemented with Google Scholar. The search was undertaken using all combinations of synonyms for 'Sargassum' and 'monitoring' (see supplementary materials for details (available online at stacks.iop.org/ERC/2/122001/mmedia)). A total of 106 571 results were returned from all three search engines and a reference manager was used to organise the results. A two-step filtering framework was applied to publications. The first step was to read the abstracts and titles of all publications to determine if they satisfied the inclusion criteria (i.e. empirical study, within our research focal area, pelagic sargassum, see supplementary material for details), as well as to remove duplicate results. A total of 283 publications were then taken forward to the second filtering step, which involved reading the papers in full to determine whether they still satisfied the inclusion criteria. Reasons for exclusion were: non-English language, not accessible, non-empirical research; a focus only on the Sargasso Sea; or, a focus on benthic species of Sargassum (see supplementary material for details). Once all the papers had been identified, the metadata of the literature was collated, and the findings within the publications were analysed to answer seven research queries covering aspects related to Sargassum identification, location, distribution, quantity and sources: 1. Is the research related to floating or landing events, and if so where did it occur (beach, near-shore or openocean)?
2. What is the spatial extent of Sargassum research?
3. Does the literature indicate a source of Sargassum?
4. What was the temporal scale of analysis?
5. Which species of Sargassum were identified?
6. What data has been used to document the occurrence of Sargassum blooms and beach landings?
7. What was the volume of Sargassum (if calculated)? Finally, a paper quality review was undertaken by subjectively rating each publication on a scale of 1-5 (1=low; 5=high) based on clarity of methods (justification and replicability), comprehensivity of presentation of results and relevance of themes to the research queries (based on criteria adapted from (Porter et al 2014)). The aim of the quality review was to understand any quality patterns within publications documenting Sargassum floating or landing events. The final dataset contains 76 publications spanning 60 years, which were mostly judged to have a quality rating of 3 or 4; for ease of reading, the empirical publications are numbered in square brackets [1]- [76] and are provided in the annex.

Results
The number of publications documenting Sargassum has grown in recent years; of the 65 publications in the past decade, 29 (45%) were published in 2018 and 2019 (figure 1). Despite the boom in publications in recent years, there remain several identifiable knowledge gaps where research is limited.

Floating and landing publications
Sargassum is reported to have a more significant impact on the coast and near-shore and on the communities whose livelihoods depend on access to the coast (Louime et al 2017). Yet only 5% of publications focussed only on landing Sargassum; a significant proportion focussed on floating Sargassum (83%), and more recently a combined analysis of both. Only four publications undertook work on landed Sargassum and these were based in Brazil (Atalaia beach, north-eastern Amazonian coast), Germany (island of Heligoland), San Andres Island

Spatial extent of floating and landing Sargassum research
Research has been undertaken in Africa, the Americas, Asia, and Europe; in the Atlantic, Pacific and Indian Oceans, with three publications undertaking a global survey of Sargassum [18, 21, 59] (figure 2). We hypothesise that coastlines adjacent to water bodies experiencing significant impacts of Sargassum would be the focus of more research than those experiencing fewer and less severe influxes; however, there are gaps, for example there is no research with a focus on Belize, Colombia and Japan. West Africa has two pockets of research in Nigeria and Ghana, which considers their own and neighbouring coastlines [2]. In the Caribbean region, there is a focus on floating Sargassum on monitoring the general area, only four publications examined individual islands within the Caribbean-Virgin Islands, San Andres Island, Puerto Rico and Barbados [7,16,48,66].

Input sources for the Sargassum
Of the 76 publications, only 13 (17%) speculated or indicated a theory on the origin of the Sargassum in their respective study areas (table 1). However, it is important to note that most publications speculate the origins of the Sargassum and conclude that the source is uncertain, with two concluding that it was unknown [23,61].
There is a significant number of publications focusing research on bloom origins around the Atlantic (n=10/13). From a deeper analysis of Atlantic based publications, it can be seen there is an emphasis on the tropical West Atlantic and Caribbean area. The Gulf of Guinea and West Africa are not always studied as a separate region but are often encompassed in publications focusing on the Tropical or Equatorial Atlantic. Three of the publications present research on the Pacific region around East Asia. There are no publications focussing on other regions of the world which explore the origin of Sargassum blooms.

Temporal distribution of Sargassum research
As expected, most of the research to date (78% of all papers) document the experience of individual locations, such as an area of sea, or an island, rather than a specific event, e.g. the 2018 bloom event. Location-based research provides either recurrent or one-off data for a specific area based on an expectation of potential Sargassum presence. Regular location monitoring (such as [3,11,47,50]) is useful for a variety of reasons such as assessing presence, extent and frequency. Event-based monitoring (in response to the occurrence of a blooming or landing event) was present in 16% of the publications. The notion of a Sargassum 'event' is rarely and inconsistently defined. Whether a publication collected evidence by location or in response to an 'event' was often inferred for this analysis, but not stated explicitly in the research. Research which appeared to focus on specific events generally collected evidence of Sargassum immediately after or in response to the emergence of a bloom over water bodies or the appearance of Sargassum mats in coastal or beach areas. For example, in response to a bloom off the coast of Florida, Marmorino et al (2011) [39] used airborne imagery to collect evidence of the Sargassum raft. Similarly, Oyesiku and Egunyomi (2014) [46] responded to reports of Sargassum in Nigeria by visiting the site and collecting samples. Some publications collected data by both monitoring locations and responding to Sargassum events; for example, Hu et al (2015) [28] utilised remote sensing to regularly monitor the Gulf of Mexico and Atlantic area and the AVIRIS sensor for event response.

Prevalence of Sargassum species in research outputs
To effectively valorise Sargassum biomass, a critical piece of information is the biochemical composition of the landing seaweed. There are more than 300 Sargassum species globally, and several morphotypes within some species, each of which potentially has a different chemical signature (Hardouin et al 2014). Composition analysis of many Sargassum species has not been investigated yet and, in relation to this, key questions remain: what is the abundance of the different Sargassum species and/or morphotypes in the seaweed mats? Are some Sargassum species more typically found in some locations than in others? Interestingly 32% (n=24) of publications did not distinguish between different species of Sargassum in their research ( figure 3).
Publications such as [9,31,34,35] show that S. horneri can be found as floating in the East China and Yellow Sea. However, Liu et al (2018) showed that these were detached Sargassum (i.e. the force of waves and currents cause the seaweed to disconnect from the bottom and they are buoyant due to having gas vesicles) rather than pelagic. It is possible that 'other' species of Sargassum are also detached and not pelagic, although more research is needed to confirm this. It is apparent that the dominant holopelagic species are S. fluitans and S. natans and their respective morphotypes, particularly in and around the Atlantic region. Only three publications, identify morphotypes of these Sargassum species [3,17,58]. This further heightens the uncertainty around Sargassum nomenclature and identification. Schell et al (2015) [58] identified that the dominant species in 2014/2015 was the morphotype S. natans VIII in the Caribbean. However, by the end of 2019, no other publications have compared morphotype dominance in any study area or time periods. If Sargassum monitoring publications, investigated the species and morphotypes of Sargassum, this could improve understanding of past trends, as well as improving prediction of future events.

Types of methods and data used to document the occurrence of Sargassum blooms and beach landings
The methods used to detect and monitor Sargassum were varied, with 46% employing a remote sensing based approach, 28% in situ (i.e. direct surveys and sampling of Sargassum) and 26% employed a combination of both remote sensing and in situ methods (table 2). Table 2 shows that for the Atlantic region most publications utilise remote sensing based methods, whereas for the Pacific region, in situ approaches have been more commonly used. For global evidence collection, understandably only remote sensing based methods are used due to the scale of the study area. Remote sensing publications are most commonly based on satellite data sources including Moderate Resolution Imaging Spectroradiometer (MODIS), Medium Resolution Imaging Spectrometer (MERIS) and Landsat, which were often accompanied by ground truth data or higher spatial resolution dataset for a small subset of the study area, examples include [4, 28, 30, 47]. A minority of publications used unmanned aerial vehicles (e.g. drones) or other alternatives for aerial photography [39,64]. Open-ocean Sargassum detection methods were most commonly based on the red-edge concept, such as floating algal index (FAI) [27]. In contrast, there was less clarity about the sampling methods used within in situ research. Often sampling methods were not clearly stated (and hence were considered to be lower quality research, see supplementary materials). Those publications that documented their methods most commonly used boats to access Sargassum rafts, examples include [35,42,55].

Discussion
The systematic documentation of locations affected by floating and landings of Sargassum presented in this work is the first analysis of its kind and identifies some unexpected results relating to the distribution and quality of Sargassum research outputs. Despite the global distribution of pelagic Sargassum research, there are surprising gaps. For example, Belize shares the same Caribbean coastline as Mexico, and could be equally prone to Sargassum influxes, yet there is currently an absence of literature documenting Sargassum in Belize. Unless all areas in a possible Sargassum impact area are monitored for Sargassum landings, it will not be possible to fully appreciate the extent of Sargassum in the Caribbean. Similarly, there is no published research on Sargassum in West African countries such as Cameroon and Gabon, yet based on their geographic location, it would make sense to assume that Sargassum is likely to be landing there. Figure 2 illustrates other areas where few publications have been undertaken and where research could be intensified, such as South America, East Africa, and North and West Africa. To fully understand and prepare for the impacts of Sargassum on stretches of coastline, spatial gaps in research need to be filled.
It is worth noting that only 17 publications undertake empirical research on monitoring Sargassum in the wider Atlantic Ocean (excluding Gulf of Mexico, Sargasso Sea and the Caribbean), which speaks to the scarcity of research on pelagic Sargassum over the course of 60 years. It further appears that the proportion of Sargassum research in a country may correlate with its relative wealth. For example, countries with a higher gross national income per capita (GNI)-using (The World Bank 2018)-such as the USA (which has n=6 publications), undertook substantially more research than those with a lower GNI, such as Togo (n=1), Liberia (n=1), Guyana (n=1) and others in West Africa (n=2), Central America (n=0) and South America(n=3). The relative levels of economic development of countries could be reflected in their investment in research into pelagic Sargassum and offer an explanation for spatial research gaps globally. Development of local empirical evidence bases on Sargassum is very important as the potential solutions to manage the influx of Sargassum in the future would require location specific information, and local strategies.
It was surprising to find that one fifth of the publications did not specify the species of Sargassum, and only three publications reported on the morphotypes of the specimen. This exemplifies the challenges in correctly identifying some of the Sargassum species, and some issues with their nomenclature. As an example, S. aquifolium appears to have a variety of synonyms according to Algae Base (Guiry 2020), which could contribute to hesitation in identifying Sargassum species in research publications. However, for the three morphotypes that affect the Caribbean and Western Africa clear morphological criteria and molecular markers have been established to identify them (Amaral-Zettler et al 2017). Another important aspect to consider is that Sargassum species may possess such similar qualities, such as their biochemical composition, that there is little need to distinguish between them. A limited number of publications have investigated aspects of the biochemical composition of holopelagic Sargassum biomass (Oyesiku and Egunyomi 2014, Addico and deGraft-Johnson 2016, Baker et al 2018, and more recently of the three individual morphotypes (Davis et al 2020, Milledge et al 2020. However, wider and more comprehensive research into composition of holopelagic Sargassum species and of their morphotypes would offer transparency of differences and could unite species that are currently thought to be distinguished. Further research on this topic for specific species and morphotypes would enable this issue to be addressed, and it may also offer clarity on taxonomy. Additionally, detailed knowledge of Sargassum composition would facilitate understanding of Sargassum uses, impacts and management options. The limited number of publications estimating quantities of Sargassum and their methodological inconsistencies prevent construction of a long-term record of Sargassum influxes and spatial-temporal analysis. Although, more recent publications are starting to do this (García-Sánchez et al 2020); there are management implications of this as it generates uncertainty. For example, Sargassum landing on beaches has occurred regularly in the past; however, in years (or seasons) with significantly high Sargassum influx, such as 2015 (which had 20 times the historical amount (Wang and Hu 2016)), management strategies are imperative to prevent socio-economic and environmental losses. Inconsistencies in estimating volumes of Sargassum prevent effective management as authorities cannot accurately anticipate and prepare for Sargassum influxes. It can be speculated that this is especially true for developing countries which have less to invest in monitoring and management. Therefore, to facilitate effective management, estimations of volume should be provided in a standardised manner, ideally alongside landing forecasts.
There are many Sargassum management questions that remain outstanding, including how long do beaching events last?, where does Sargassum occur most regularly?, what are the local socio-economic impacts and how can they be mitigated?, what are the environmental impacts on specific areas/habitats?. None of these can be assessed or quantified when there is little research on beach landings of Sargassum. Focussing research on floating and openocean Sargassum and overlooking analysis of the magnitude and severity of beach impacts leaves management queries unanswered. A further under-researched area which hinders management capacity is a lack of research on event response. With the majority of publications focussing on regular monitoring of open-ocean areas, event response research is limited. These research gaps hamper detailed analysis on how Sargassum interacts with communities and its impacts on livelihoods and economies. In Mexico, management plans have been put into place, as reported by print media, such as installing 'trial and error' hard engineering solutions including barriers (Mexico News Daily 2019). Although attempts at management are possible in the absence of detailed impact data, it can be argued that with more robust research on beach landings of Sargassum and on Sargassum events, more reliable solutions can be introduced with a higher potential for success.

Conclusions
The outstanding Sargassum research gaps relate to input sources, locations and species identification, as well as the quantity in the oceans and the amount landing on beaches. The rapid growth in publications on Sargassum over recent years is a welcome step towards understanding Sargassum blooms and its geographic spread. However, there is a need to improve the robustness and extent of research to ensure in-depth understanding of the complex issue and support a comprehensive management plan for all affected communities.
First, the spatial coverage of research should be expanded to represent many missing countries and coastlines, notably in Central America and West Africa. This will better support Sargassum management within integrated coastline management across geopolitical boundaries. The spatial gaps in research likely contribute to the lack of agreement on where the blooms emerge and the potential cycles and triggers.
Second, our analysis shows that most Sargassum publications have focused on longer-term regular monitoring of specific locations and not addressed 'event' response effectively. To aid the production of comparable research on Sargassum events, the notion of a Sargassum event needs to be more clearly defined. Longer term records of Sargassum are needed to monitor temporal changes in frequency of events and quantities landed-both of which are required to better understand how to reuse or manage the events. Research focussing on influx and blooming events should generate a longer and more detailed temporal record. To address the lack of a definition of a 'Sargassum event', we propose the following: A Sargassum 'event' is a continuous bloom of any Sargassum in open oceans, or, an aggregation of landed Sargassum with the potential to disrupt local social, economic or ecosystem functioning, or to impact human health. An event can affect one country, or several contiguous countries.
Third, at present Sargassum species identification and reporting is not standardised, creating incomparability issues in the Sargassum evidence base. Standard species identification should clarify which species are dominant in different regions and enable determination of the variation of dominant species seasonally and annually. Important questions need to be answered, such as do the mats in the tropical Atlantic have the same species mix throughout the season or does it differ? Are different species dominant in different regions of the Atlantic Ocean? Are mats of one type of species invading other coastal areas or restricted to one location? These cannot be addressed if the species and morphotypes of Sargassum are not always recorded in publications. Similarly, a robust and standardised method for estimating volume needs to be developed and implemented to enable research gaps to be addressed and effective management strategies to be implemented.
Finally, there is a key research gap in understanding the nature and extent of the impact of Sargassum landing on the socio-economic activities of the affected communities. Further research on short-and longer-term impacts of Sargassum landing on coastal communities would be crucial to developing any Sargassum risk mitigation strategy.