A systematic review of marine macroalgal degradation: Toward a better understanding of macroalgal carbon sequestration potential

Abstract

Although macroalgae are gaining recognition for their potential role in marine carbon sequestration, critical knowledge gaps related to the fate of macroalgal carbon limit our capacity to quantify rates of macroalgal carbon sequestration. Understanding the degradation dynamics of macroalgal‐derived biomaterials—including tissue/wrack, particulate organic matter/carbon (POM/POC), and dissolved organic carbon (DOC)—as well as the environmental drivers of decomposition are critical for assessing the longevity of macroalgal carbon and the potential storage capacity of macroalgae. Thus, a systematic literature review of macroalgal degradation studies was conducted to compile data, estimate the relative recalcitrance (i.e., relative stability) of macroalgal biomaterials, and elucidate key drivers of macroalgal decomposition dynamics. We found that macroalgal decay trajectories are highly variable and not always best described by the often‐cited exponential decay models. Our analysis demonstrated that temperature was a notable driver of decomposition, with higher temperatures eliciting faster rates of decomposition. Furthermore, we found that brown algae had significantly higher proportions of recalcitrant biomaterials when compared to red algae. The impact of other factors, including biomaterial type, degradation environment, and tissue carbon and nitrogen content on macroalgal degradation, is variable across contexts, warranting further study. These results help to provide a foundation from which to plan and assess future studies on macroalgal degradation, which will improve our understanding of how macroalgae contribute to marine carbon cycles, trophic subsidies, and, potentially, marine carbon sequestration.

Abbreviations

AIC

Akaike information criterion

C:N

carbon to nitrogen ratio

CSP

carbon sequestration capacity

DOC

dissolved organic carbon

DOM

dissolved organic matter

eDNA

environmental DNA

GLMM

generalized linear mixed effects model

LMM

linear mixed effects model

NPP

net primary productivity

POC

particulate organic carbon

POM

particulate organic matter

PRISMA

preferred reporting items for systematic reviews and meta‐analyses

RDOC

recalcitrant dissolved organic carbon

INTRODUCTION

One of our most valuable assets in the fight against climate change is ecosystems that naturally sequester carbon (i.e., remove carbon from the atmosphere and store that carbon for 100 years or more; Lal, ²⁰⁰⁷). The ocean, which absorbs up to one third of anthropogenic CO₂ emissions annually, has recently become widely recognized for its substantial role in global carbon cycles (Gruber et al., 2019; Intergovernmental Panel on Climate Change, ²⁰¹³). Vegetated coastal ecosystems, including seagrass meadows, tidal marshes, and mangrove forests, are carbon sequestration hotspots and have, therefore, been deemed blue carbon ecosystems (Lovelock & Duarte, 2019; Macreadie et al., ²⁰²¹). Yet, despite being highly productive ecosystems with expansive global distributions (Duarte et al., 2022), macroalgal (i.e., seaweed) forests are often not considered as blue carbon ecosystems due to a current lack of understanding surrounding the ultimate fate of macroalgal‐derived carbon (Dolliver & O'Connor, ^2022a; Fujita et al., ²⁰²³; Gallagher et al., ²⁰²²; Macreadie et al., ²⁰¹⁹; Pessarrodona et al., ²⁰²³). Most types of recognized blue carbon ecosystems occur within soft sediment habitats (including seagrasses, tidal marshes, and mangroves), which are conducive to the rapid burial of organic matter (which is the primary mode through which carbon sequestration in these ecosystems occurs; Atwood et al., ²⁰²⁰; Mcleod et al., ²⁰¹¹). Macroalgae, however, commonly grow on rocky reefs where the burial of detritus does not happen readily, making assessments of the ultimate fate of macroalgal carbon a much more complex endeavor (Hurd et al., 2022; Krause‐Jensen et al., ²⁰¹⁸).

Macroalgae produce and release large amounts of organic matter to the surrounding environment via erosion, breakage, exudation, or total thalli dislodgement (de Bettignies et al., ²⁰¹³; Duarte et al., ²⁰²²; Pessarrodona et al., ²⁰¹⁸). Kelp forests, for instance, release more carbon via detritus/litterfall than most other vegetated ecosystems, at a rate comparable to that of tidal marshes (Pessarrodona et al., 2018). Macroalgae release carbon in many different forms, ranging from an entire algal thallus, tissue fragments, particulate organic matter (POM, which includes particulate organic carbon, POC), and dissolved organic matter (DOM, which includes dissolved organic carbon, DOC). Particulate organic matter is typically defined operationally as suspended organic matter collected on a filter (typically 0.2 or 0.7 μm pore size), while dissolved organic matter passes through a filter (again usually 0.2 or 0.7 μm pore size; see Glossary; Hansell, ²⁰⁰¹; Kharbush et al., ²⁰²⁰; Lee et al., ²⁰⁰⁴; Verdugo et al., ²⁰⁰⁴). On average, 30% of macroalgal net primary productivity (NPP) is released into the water column as DOC, and 60% of macroalgal NPP is eventually released as tissues and/or POM (Pessarrodona et al., 2023), although these rates vary substantially with species, environment, and season. This continual supply of organic matter means that macroalgae are recognized as “carbon donors,” contributing to allochthonous carbon subsidies to other ecosystems or areas (such as seagrass meadows, mangrove forests, coastal/intertidal areas, the open ocean, or the deep sea; Cartraud et al., ²⁰²¹; Filbee‐Dexter et al., ²⁰¹⁸; Hill et al., ²⁰¹⁵; Kennedy et al., ²⁰¹⁰; Krumhansl & Scheibling, ^2012b; Olson et al., ²⁰¹⁹).

After being released, macroalgal‐derived carbon can be sequestered in the ocean through three established pathways. First, macroalgal biomaterials may be exported to and buried in neighboring soft sediment habitats, enabling the stable storage of the carbon contained within buried macroalgal biomaterials (Braeckman et al., 2019; Erlania et al., ²⁰²³; Moreda et al., ²⁰²⁴; Ørberg et al., ²⁰²³; Queirós et al., ²⁰¹⁹; Wang et al., ²⁰¹⁸). However, although it is commonly assumed that the degradation of marine detritus slows almost to zero in anoxic sediments, there is evidence that degradation continues after macroalgal biomaterials are incorporated into marine sediments (Braeckman et al., 2019; Haram et al., ²⁰²⁰; Rossi, ²⁰⁰⁶). Also, if the sediments are disturbed, any stored carbon that is released becomes susceptible to reintroduction to the atmosphere (Lovelock, Atwood, et al., 2017; Pendleton et al., ²⁰¹²). Second, macroalgal biomaterials can be exported to the deep ocean, where extremely slow rates of mixing with surface waters mean that carbon is also likely to be sequestered (Chang et al., 2024; Filbee‐Dexter et al., ²⁰²⁴; Fischer & Wiencke, ¹⁹⁹²; Kokubu et al., ²⁰¹⁹; Schimani et al., ²⁰²²). Even so, quantifying sequestration of carbon in the deep sea comes with a number of uncertainties due to different rates and patterns of mixing that occur across ocean basins (Baker et al., 2022; Nowicki et al., ²⁰²²; Siegel et al., ²⁰²³). Last, macroalgal‐derived DOC may be sequestered within the ocean if that DOC avoids microbial‐ and photo‐degradation, persisting in seawater for timescales relevant to carbon sequestration (Chen et al., 2020; Feng et al., ²⁰²²; Hurd et al., ²⁰²²; Shank et al., ²⁰¹⁰; Wada et al., ²⁰¹⁵). Macroalgae may be important contributors to a substantial oceanic carbon reservoir of recalcitrant DOC (RDOC; Shen & Benner, ²⁰¹⁸), which remains relatively stable over time and is comparable in size to the atmospheric carbon pool (Dittmar et al., 2021; Hansell et al., ²⁰⁰⁹). Although this pool of DOC is derived from a variety of sources, macroalgae are known to be important contributors to the coastal DOC pool (contributing up to 20% of total coastal DOC concentrations; Paine et al., ²⁰²¹; Wada & Hama, ²⁰¹³), and vegetated coastal ecosystems are one of the largest sources of marine DOC (Barrón & Duarte, 2015; Bauer & Druffel, ¹⁹⁹⁸; Wagner et al., ²⁰²⁰). Therefore, although macroalgae are likely important contributors to the oceanic pool of RDOC, more study is needed to directly quantify the extent of their contribution. An important note is that microbes can also transform labile marine DOC into RDOC, under the microbial carbon pump hypothesis (Feng et al., 2022; Jiao et al., ²⁰¹⁰; Ogawa et al., ²⁰⁰¹). Thus, macroalgal‐produced DOC might not be inherently recalcitrant but may still ultimately be sequestered as RDOC after transformation by the microbial community (Jiao et al., 2010; Legendre et al., ²⁰¹⁵).

The rate of and magnitude at which macroalgal‐derived carbon is sequestered through these pathways is currently poorly resolved (Krause‐Jensen et al., 2018; Pessarrodona et al., ²⁰²³). One key knowledge gap that hinders our ability to estimate the carbon sequestration potential of macroalgae is a lack of understanding of the patterns and drivers of macroalgal biomaterial degradation (Dolliver & O'Connor, ^2022a; Paine et al., ²⁰²¹). Like all organic materials, macroalgal biomaterials, once released, are susceptible to degradation/decomposition through microbial degradation (Boldreel et al., 2023; Feng et al., ²⁰²²; Morrison et al., ²⁰¹⁷; Rieper‐Kirchner, ¹⁹⁸⁹) and/or photodecomposition (Huang et al., 2023; Shank et al., ²⁰¹⁰; Wada et al., ²⁰¹⁵). Organic matter degradation is defined as the conversion of complex organic molecules contained within organic matter into simpler, inorganic molecules (LaRowe & Van Cappellen, 2011; Middelburg, ¹⁹⁸⁹; Oades, ¹⁹⁸⁹). In the context of carbon sequestration, the conversion of organic carbon into inorganic carbon via degradation means that inorganic carbon is susceptible to re‐introduction to the atmosphere (Lal, 2007). Macroalgal‐derived biomaterials are also susceptible to ingestion by grazers; this process also interacts with degradation dynamics (Bedford & Moore, 1984; Catenazzi & Donnelly, ²⁰⁰⁷; Gómez et al., ²⁰¹⁸; Kotta et al., ²⁰¹⁰; Kristensen & Mikkelsen, ²⁰⁰³; Salathe & Riera, ²⁰¹²). Therefore, in order to assess the carbon sequestration potential of macroalgae, we need to understand whether macroalgal biomaterials can either resist degradation or ingestion for timescales relevant to climate change amelioration (i.e., 100+ years; Moreda et al., ²⁰²⁴) or for however long it takes for the biomaterials to reach areas conducive to carbon sequestration (e.g., soft sediment habitats or the deep ocean).

The degradation of organic matter is often best explained by an exponential decay model, which is characterized by an initial period of rapid biomaterial loss sometimes referred to as the leaching period, wherein labile, easily accessible portions of biomaterial (such as carbohydrates and amino acids) are quickly utilized by microbes (Berg & Laskowski, 2005; Wider & Lang, ¹⁹⁸²). Subsequently, the degradation rate rapidly declines after the highly labile components are digested, leaving behind recalcitrant fractions of biomaterial that are harder for microbes to attack (de Bettignies et al., ²⁰²⁰; Jenny et al., ¹⁹⁴⁹; Yang & Janssen, ²⁰⁰¹). Recalcitrant organic matter is defined as organic matter which persists in a system, although the length of time for which the organic matter must resist degradation for it to be deemed recalcitrant varies substantially across and within different disciplines (Hansell, 2013; Kleber, ²⁰¹⁰; McCann & Carpita, ²⁰¹⁵). This recalcitrant portion of biomaterial may either have molecular characteristics that hinder microbial degradation—for example, by requiring many enzymes and/or a high activation energy to break down—or the recalcitrant biomaterial may be protected from microbial attack due to physical separation from microbes, environmental drivers, microbial food saturation, or inadequate microbial population and/or diversity (Dungait et al., 2012; Kleber, ²⁰¹⁰; Kleber et al., ²⁰¹¹; Marschner et al., ²⁰⁰⁸; Schmidt et al., ²⁰¹¹). Research has shown that long‐term recalcitrance of organic matter in terrestrial soil systems is predominately controlled by extrinsic factors rather than by the molecular structure of the organic matter itself (Amelung et al., 2008; Dungait et al., ²⁰¹²; Schmidt et al., ²⁰¹¹). Thus, in many instances across marine and terrestrial contexts, organic matter only resists degradation under certain biotic and abiotic conditions and will continue to degrade in novel conditions (Chabbi et al., 2009; Jiao et al., ²⁰¹⁴; Pedersen et al., ²⁰²¹; Shen & Benner, ²⁰¹⁸).

In macroalgae, recalcitrant portions of biomaterial are often made up of structural cell wall compounds which may be difficult for microbes to digest; these compounds include polyphenolics, ulvans, alginates, carrageenan, and fucoidan (Adair et al., 2008; Becker et al., ²⁰²⁰; Bligh et al., ²⁰²²; Deniaud‐Bouët et al., ²⁰¹⁷; Domozych, ²⁰¹⁹; Imran et al., ²⁰¹⁷; Shukla et al., ²⁰¹⁶; Sichert et al., ²⁰²⁰). These compounds are absent from the cell walls of terrestrial plants, which are instead made up of cellulose, hemicellulose, pectin, and lignin (Domozych et al., 2012; Niklas et al., ²⁰¹⁷; Popper et al., ²⁰¹¹). These unique algal polymers have a variety of functions, including protection against grazing and/or microbial attack, osmotic and/or ionic regulation, and protection from UV damage (Deniaud‐Bouët et al., 2017; Holzinger et al., ²⁰¹⁵; Kloareg & Quatrano, ¹⁹⁸⁸; Oren, ²⁰⁰⁷; Shtein et al., ²⁰¹⁸). Brown, green, and red macroalgae have distinct cell‐wall polymers, likely attributable to their distinct evolutionary histories (Ciancia et al., 2020; Lee & Ho, ²⁰²²; Popper et al., ²⁰¹¹; Synytsya et al., ²⁰¹⁵). Brown algae, in particular, contain many unique compounds that deter microbial attack and grazing, meaning that they are thought to be harder to degrade and, thus, may be more important contributors to marine carbon sequestration compared to other algal groups (Arnosti, 2011; Deniaud‐Bouët et al., ²⁰¹⁷; Littler et al., ¹⁹⁸³). These compounds include polyphenolics (including phlorotannins), laminarin, fucoidan and alginate, which are all highly resistant to microbial degradation, requiring complex and highly specialized suites of enzymes to degrade (Alderkamp et al., 2007; Armstrong & Patel, ¹⁹⁹⁴; Bligh et al., ²⁰²²; Buck‐Wiese et al., ²⁰²³; Sichert et al., ²⁰²⁰; Targett et al., ¹⁹⁹²; Zhang et al., ²⁰²¹).

Assessing the relative contributions of various groups of macroalgae to marine sediments and/or the deep sea (through eDNA and/or stable isotope modeling approaches) can also give insights into the relative stability of different types of macroalgae. One study that characterized the prevalence of macroalgal DNA across a range of ocean depths around the world observed that red algae were the most prevalent overall (Ortega et al., 2019). However, other studies with more narrow geographic scopes have observed that brown algae are typically more prominent in sediments and/or the deep sea, as compared with red and green algae (Erlania et al., 2023; Kokubu et al., ²⁰¹²; Ørberg et al., ²⁰²³; Queirós et al., ²⁰²³), and one study noted that green algae were the most common type of macroalgae in seagrass‐associated sediments (Arina et al., 2023). These trends, however, are likely somewhat influenced by the abundance of algal groups in certain areas and are likely not solely indicative of the relative stability of various algal groups (Erlania et al., 2023; Ortega et al., ²⁰¹⁹).

There are many additional reasons why understanding macroalgal degradation dynamics is an important pursuit, apart from the carbon sequestration potential of macroalgae. First, the rate at which macroalgal biomaterial degrades has important implications for ecosystem functioning, as it impacts the flow of carbon, nitrogen, and other nutrients contained within the biomaterial both within and beyond macroalgal‐dominated ecosystems (Hanisak, 1993; Luo, Dai, et al., ²⁰²²; Mellbrand et al., ²⁰¹¹; Pedersen & Johnsen, ²⁰¹⁷; Renaud et al., ²⁰¹⁵). Macroalgal detritus is an important food source for various meio‐ and macrofaunal communities (Alkemade & Van Rijswijk, 1993; Chown, ¹⁹⁹⁶; Hosen et al., ²⁰¹¹; Kristensen & Mikkelsen, ²⁰⁰³; Nedzarek & Rakusa‐Suszczewski, ²⁰⁰⁴), and decaying macroalgae can often harbor and support a diverse array of meiofauna (Duggins et al., 2016; Hwang et al., ²⁰²³; Olabarria et al., ²⁰⁰⁷; Rieper‐Kirchner, ¹⁹⁹⁰; Urban‐Malinga & Burska, ²⁰⁰⁹). For instance, the input of beach‐cast macroalgal wrack to sandy beaches is a crucial source of nutrients and habitat for these ecosystems, since sandy beaches usually have little in situ primary production and little three‐dimensional habitat (Colombini et al., 2000; Griffiths et al., ¹⁹⁸³; Hyndes et al., ²⁰²²; Ince et al., ²⁰⁰⁷; Salathe & Riera, ²⁰¹²).

Other timely reasons to study marine macroalgal degradation dynamics include understanding the ecological ramifications of macroalgal blooms (Castaldelli et al., 2003; Conover et al., ²⁰¹⁶; Lanari, Copertino, et al., ²⁰¹⁸), non‐native macroalgal invasions (Haram et al., 2020; Krumhansl & Scheibling, ^2012a; Lozada et al., ²⁰²²; Pedersen et al., ²⁰⁰⁵; Rodil et al., ²⁰⁰⁸), and large scale seaweed farming (Dolliver & O'Connor, 2022; Li et al., ²⁰²²; Luo et al., ^{2021, 2024}), all of which are increasing in occurrence and intensity worldwide (Cai et al., 2021; Seebens et al., ²⁰¹⁷; Xing et al., ²⁰¹⁵; Zhang, Liao, et al., ²⁰²²). Furthermore, the replacement of kelp and fucoid‐dominated ecosystems with turf algae‐ and crustose coralline algae‐dominated systems is becoming increasingly common due to anthropogenic stressors (including eutrophication, ocean warming, and epiphytism; Filbee‐Dexter & Wernberg, ²⁰¹⁸; Harrold & Reed, ¹⁹⁸⁵; Krumhansl et al., ²⁰¹⁶; Matsunaga et al., ¹⁹⁹⁹; Rogers‐Bennett & Catton, ²⁰¹⁹; Smith et al., ²⁰²⁴). Thus, understanding if there are differences in degradation dynamics across different macroalgal functional groups will give insight into the ecological ramifications of current and potential future changes to macroalgal community structures.

Despite being an ecologically important and relatively well‐studied process, there has yet to be a comprehensive, systematic review published on marine macroalgal degradation dynamics. This review synthesizes trends within the body of peer‐reviewed literature on marine macroalgal degradation to ascertain drivers of macroalgal biomaterial degradation and better understand the longevity and potential recalcitrance of macroalgal carbon. To this end, the rate of degradation was verified for 504 observations of macroalgal biomaterial degradation across 105 studies, and the trajectory of degradation was verified wherever possible. This enabled assessments of macroalgal half‐lives (i.e., the time it takes for 50% of initial biomaterial to degrade) and percent recalcitrance (i.e., relative stability; see Glossary). We then used our half‐life and percent recalcitrance estimates to statistically assess the influence of methodological and environmental factors on macroalgal degradation. Importantly, we have also summarized the results of studies that independently investigated the roles of various drivers on macroalgal degradation. We have also reported trends in study motivation, study location, macroalgal type, biomaterial type, and various experimental design choices across the body of literature. Lastly, we have highlighted key remaining knowledge gaps, suggested future directions, and recommended methodologies and approaches for future study, with a particular focus on macroalgal carbon sequestration potential. This review will improve our understanding of the ultimate fate of macroalgal‐derived carbon in the ocean, understanding that is needed in order to ascertain the contribution of macroalgae to marine carbon sequestration. In addition, understanding the drivers of degradation will provide important insights into the factors that influence the flow of macroalgal biomaterials (and the nutrients contained within those biomaterials) to other ecosystems and organisms, many of which rely on macroalgal trophic subsidies (Krumhansl & Scheibling, ^2012b; Norderhaug et al., ²⁰⁰³; Raut & Capone, ²⁰²¹).

METHODS

Literature search and inclusion criteria

A systematic search of published literature was conducted using the databases Web of Science Core Collection and Scopus (using a title‐abstract‐keyword search) using the following search string: “macroalga* OR seaweed OR kelp OR fucoid AND detrit* OR litter OR wrack OR POM OR DOM OR POC OR DOC AND decompos* OR degrad* OR decay.” We also conducted a secondary search in Google Scholar in which we screened the first 500 search results, out of 11,600 results. (This returned 20 unique studies we deemed appropriate for inclusion in our review.)

The titles, abstracts, and results were screened to determine if the study met our inclusion criteria (outlined in Figure S1). To be included in this study, publications needed to be peer‐reviewed, primary research on marine or estuarine macroalgal degradation, and include data on the loss of macroalgal biomaterials (in the form of tissues, POM, POC, DOM, or DOC) over time. Studies were discarded for numerous reasons, outlined in the PRISMA flowchart (Figure S1). Many studies were discarded as data pertaining to algal degradation (including an initial assessment biomaterial quantity) were not available in text, within figures, or within supplemental materials. We excluded studies that only analyzed degradation through thermogravimetric analysis or only measured changes in chromophoric dissolved organic matter concentration. We also excluded unpublished theses, preprints, and gray literature. Databases were screened by C.B. and J.K. independently, with the final literature search occurring in August 2024. (See Table S1 for details on when each search was conducted and by whom.) A full list of studies screened, included, and excluded is in Supplemental Datafiles 1, 2, and 5 (available at https://osf.io/vmtfd/files/osfstorage).

Data extraction

We extracted macroalgal degradation data from 504 observations contained within 105 publications selected for inclusion in this review. We define an observation of macroalgal degradation as a data set that describes change in biomaterial mass or concentration, or percent loss of biomaterial, over time (i.e., an individual degradation trajectory). Where degradation data was not presented in text form, data was extracted from figures using the online software PlotDigitizer (PlotDigitizer, 2024). In most cases, the raw data were not available, so average values for each time point were extracted. Many papers did not include estimates of standard error or standard deviation and so were omitted. Data extraction was conducted by C.B. and J.K. independently. Extracted degradation data can be found in Supplemental Datafile 3.

All other relevant details from each study were extracted, including methodological details and biophysical parameters, and are available in Supplemental Datafile 1. We recorded what type of macroalgal‐derived biomaterial the authors of each study measured the decay trajectory of, based on whether study authors stated that they measured the degradation of: macroalgal tissue, carbon content within macroalgal tissue, POM, POC, or DOC. See Glossary for a definition of the various types of biomaterials. Since the definition of POM/POC and DOM/DOC can vary, for observations of POM/POC degradation, we recorded what size range of particles a study classified as being POM/POC, and for studies interested in DOC degradation, we recorded the filter pore size used by study authors to quantify DOC in water samples. Next, we classified studies on the basis of what the authors stated their motivation was for conducting their study, as described either in the abstract and/or discussion, binning studies into six categories of study motivations outlined in Table 1. Also, we noted if the authors chose to pre‐treat or pre‐kill macroalgae prior to the start of the degradation experiment. If so, we noted how they pre‐treated the algae. (Pre‐treatment methods included freezing the algae, placing the algae in a hot water bath, or drying the tissue out.) Other relevant study parameters we extracted included but were not limited to: study location (ex situ or in situ, latitude and longitude, study location characteristics) and methodological details (macroalgal species, study duration, number of replicates, temperature, light level, and others). We also recorded data on changes in carbon content, nitrogen content, and/or carbon: nitrogen ratios in macroalgal detritus throughout degradation, if authors measured and reported these data (Supplemental Datafile 4).

TABLE 1.

Proportions of studies included in this review that fit into six broad categories of study motivation, along with the definition for each study motivation category.

Study motivation category	Category definition	Percentage of studies belonging to this category
Nutrient and/or heavy metal release and cycling	Many studies stated that understanding nutrient release rates (including carbon, nitrogen, phosphorous and others) during degradation was their main motive. This also includes authors interested in the rates of heavy metal release during macroalgal decay	29%
Meiofaunal community dynamics and/or impact of meiofauna on degradation	Authors were primarily interested in either understanding how wrack acts as a habitat for small invertebrates and/or how this meiofaunal community impacts algal degradation dynamics	19%
Carbon sequestration potential (CSP) of macroalgae	Studies' with the primary motivation to better understand the CSP of macroalgae by characterizing degradation dynamics (26% of studies identify macroalgal CSP at least once throughout their paper, often in the discussion to contextualize their results)	18%
Cross‐species comparison	Studies' with the central aim to compare how degradation rates changed across various species of macroalgae and/or aquatic plants. For example, comparing the degradation of invasive algal species to native species or bloom‐forming algae to longer‐lived algae	17%
Microbial community dynamics and/or impact of microbes on degradation	Studies that primarily focused on characterizing the microbial community throughout degradation and/or describing how microbes impact degradation	12%
Impact of abiotic factors on degradation	These abiotic factors include temperature, depth, oxygen availability, salinity, and others	6%

Note: Study motivation refers to what the authors state in their abstract and/or introduction as being their motive for conducting their research.

All taxonomy was confirmed and updated using the World Register of Marine Species (https://www.marinespecies.org/index.php). For analytical purposes, macroalgae were partitioned into functional groups including “kelps” from the order Laminariales (Jayathilake & Costello, 2020); “fucoids” from the order Fucales; “turf” including low‐lying, ephemeral, densely aggregating macroalgae (see Connell et al., ²⁰¹⁴); and “foliose” macroalgae (i.e., macroalgae larger than turf and not within the kelp or fucoid functional groups; Steneck & Dethier, ¹⁹⁹⁴). We also recorded macroalgal classification, which refers to whether the species is a brown algae (Class Phaeophyceae), green algae (Phylum Chlorophyta), or red algae (Phylum Rhodophyta). See Supplemental Datafile 6 for a full list of the 99 species of macroalgae included in this review and the corresponding functional group and classification we reported each species as.

Data analysis

Verifying the trajectory of degradation within algal degradation studies

Across macroalgal degradation studies, two common models were used to describe biomaterial decomposition rate: exponential decay and linear loss. Exponential decay provides a decay constant or k‐value (Equations 1 and 2), and linear loss provides a constant degradation rate that estimates the percent loss per day (Equation 3). When presenting degradation data, some authors verified the best model fit for their data and then presented the corresponding summary statistic, while others presented both values, or just one value without verification of the best fit. For this reason, the trajectory of degradation was re‐analyzed for all observations collected in this study to determine what model best fit the data.

Degradation trajectories could not be calculated in some cases. First, in some cases macroalgal growth (i.e., increase in biomass/concentration) outweighed/negated any signs of decay, meaning that degradation trajectory could not be easily described, as in these instances teasing apart growth and degradation was not possible (Figure 1d). These observations were designated as being “confounded by growth.” Furthermore, decay trajectories could not be verified if degradation was only assessed from an initial and final value. In these cases, the trajectory of degradation was assumed to be linear. In cases where study authors only included degradation rates and/or k‐values without including any raw degradation data, degradation trajectory verification was not possible, so only reported degradation rates and/or k‐values were used in our analyses.

FIGURE 1.

Example data plots which illustrate the types of models used to describe macroalgal decay in this review paper. (a) Simple (two‐parameter) exponential decay, (b) three‐parameter exponential decay, (c) linear loss, (d) confounded by growth (meaning no model could be fit to the data), (e) three‐parameter sigmoidal model, and (f) four‐parameter sigmoidal model.

For studies with at least two sampling events (plus an initial sample), we were able to characterize the degradation trajectory. First, units for biomass/concentration were converted into percent of the initial value remaining, if not done already. Using Sigmaplot's regression Wizard, we tested various curves for each data set to determine which curve had the highest R ² value and, therefore, best described the data. We tested linear, sigmoidal, and exponential decay models for each data set (see Figure 1 for example data and fitted model types). We endeavored to avoid overfitting the data by avoiding adding unnecessary parameters, particularly for data sets with a limited number of sampling events. All extracted data, model types, equations, and associated R ² values can be found in Supplemental Datafile 3.

Five model types were used to describe degradation data in this review:

1. Two‐parameter exponential decay model (Equation 1; Figure 1a): characterized by an initial period of rapid degradation, followed by a period of slower degradation until the curve reaches zero, where y ₀ = initial value, k = decay constant, t = degradation time in days.

$$y=y_0{e}^{-\mathit{kt}}$$ (1)

• 2Three‐parameter exponential decay model (Equation 2; Figure 1b): an exponential decay function (as described above) where the curve never reaches zero and instead levels off at a horizontal asymptote (representing a portion of biomass that is considered recalcitrant). Here R = the asymptote (i.e., percent which is recalcitrant, if the asymptote is positive), a = initial value‐R, k = decay constant, t = degradation time in days.

$$y={\mathit{ae}}^{-\mathit{kt}}+R$$ (2)

• 3Linear model (Equation 3; Figure 1c): characterized by a relatively constant degradation rate throughout the experiment duration. Here y ₀ = initial value, d = degradation rate (in percent change per day), t = degradation time in days.

$$y=y_0+\mathit{dt}$$ (3)

• 4Three‐parameter sigmoidal model (Equation 4; Figure 1e): characterized by a period of negligible degradation at the beginning of the experiment, then the degradation rate increases to an inflection point, after which degradation rate decreases. Here a = maximum degradation rate, c = time point of maximum degradation rate, s = scaling value, t = degradation time in days. Note: For data described by sigmoidal models, we calculated the “next best fit” by testing what model has the next‐highest R ² value, so that we could estimate half‐life from these data.

$$y=\frac{a}{1+e^{\left(t-c\right)/s}}$$ (4)

• 5Four‐parameter sigmoidal model (Equation 5; Figure 1f): a sigmoidal curve (as described above) but with a positive horizontal asymptote, representing a portion of recalcitrant material. Here a = maximum degradation rate, c = time point of maximum degradation rate, s = scaling value, t = degradation time in days, R = the asymptote, or portion of recalcitrant material.

$$y=\frac{a}{1+e^{\left(t-c\right)/s}}+R$$ (5)

Calculating macroalgal half‐life and percent recalcitrance

We estimated the half‐life of macroalgal biomaterial (i.e., time it takes in days for 50% of initial biomaterial to disappear) either using the formula half‐life = 50/degradation rate for linear relationships, or half‐life = ln(2)/k for two‐parameter exponential decay models. For data best described by three‐parameter exponential decay models, y = 50 was used in the best‐fit equation and then t was solved for. For sigmoidal curves, we used the next best fit model (i.e., next highest R ² value) to calculate the half‐life. Where the asymptote value was greater than half of the initial material, the half‐life could not be calculated.

Determining what portion of macroalgal biomaterial is recalcitrant can be difficult since it is highly dependent on the duration and methodology of the experiment. Although researchers might report that a portion of initial biomaterial remains after their experiment concludes, if monitored for longer or exposed to different environmental conditions, the remaining biomaterial may continue to degrade (Figure S2). However, if the data follow a three‐parameter exponential decay model or a four‐parameter sigmoidal model with a positive R value (i.e., positive horizontal asymptote; Equations 2 and 5), there is likely a recalcitrant portion of biomaterial since the degradation rate has leveled off. Thus, we estimated percent recalcitrance as the value of the horizontal asymptote in cases where degradation data were best described by a model with a horizontal asymptote.

We assessed that there was zero percent recalcitrance if (a) there was no biomaterial left when the experiment ended or (b) the experiment duration was longer than 1 month and the degradation data were not best described by a model with a horizontal asymptote. We chose 1 month as a cut‐off after visually inspecting all degradation curves that followed a three‐parameter exponential decay model and observed that the overall average approximate time it took for the degradation rate to approach zero was 1 month. We were unable to confidently assess whether there was a portion of recalcitrant biomaterial in the following cases: (a) if the study did not provide degradation data and only provided an estimate of degradation rate and/or k value, (b) if the study only measured initial and final biomass/concentration and there was still remaining biomaterial when the experiment ended, and (c) if the degradation trajectory was not best described by a model with a positive horizontal asymptote and the experiment duration was less than 1 month and there was still remaining biomaterial at the end of the experiment.

Statistically analyzing trends across studies

All statistical analyses were performed in R (v4.2.3, R Core Team, ²⁰²³). To assess the potential influence of environmental and methodological factors on macroalgal half‐life (in days), we ran a multivariate linear mixed model (LMM) with study included as a random effect, using the lme4 package (Bates et al., 2015). Predictor variable selection was performed using the best subset regression modeling approach, based on Akaike information criterion (AIC) and Mallow's Cp values, using the package olsrr (Hebbali, 2024). We only considered the inclusion of predictor variables for which we had an a priori hypothesis as to why these factors may influence macroalgal degradation (see Table 2). We also ensured that none of the potential predictor variables were significantly correlated with each other to avoid multicollinearity in our model. This resulted in the selection of the following variables for inclusion in the half‐life LMM: algal functional group, algal class, experimental temperature (°C), pre‐treatment, litterbag mesh diameter (mm), degradation environment, and light availability (see Table 2 for a specification of categorical variable levels, groups means, and sample sizes). Half‐life and litterbag mesh size data were both log₁₀ transformed prior to analysis. Four studies which analyzed the degradation of algal polycultures (i.e., a mixture of various algal species) were excluded from the analyses of functional group and algal class (by coding these cells as NAs in our data frame). Also note that since very few studies reported light intensity, we analyzed the effect of light as a binary variable (i.e., light was available or the experiment was done in darkness). Additionally, despite the fact that there are many different methods of pre‐treating macroalgae, we chose to analyze pre‐treatment as a binary variable since most pre‐treatment types were represented by very few studies. To understand the impacts of categorical variables on half‐life and percent recalcitrance, the emmeans package was used to extract pairwise contrasts from our models (Lenth, 2024). Model diagnostics for all models were assessed using the DHARMa package (Hartig, 2022).

TABLE 2.

List of environmental and methodological factors that were investigated using statistical and qualitative approaches in this review.

Factor	A priori hypothesis	Half‐life (in days) group means ± standard error (sample size)	Percent recalcitrance (%) group mean ± standard error (sample size)
Biomaterial type a	Many studies have shown marine DOC from a variety of sources to be considerably recalcitrant, we might expect macroalgal derived DOC to be more recalcitrant and/or degrade slower than tissues or particulate matter	POC = 43.86 ± 12.92 (n = 16) POM = 37.45 ± 10.31 (n = 15) DOC = 11.8 ± 2.95 (n = 34) Tissue = 59.31 ± 6.70 (n = 349)	POC = 40.72 ± 18.62 (n = 4) POM = 42.77 ± 5.88 (n = 18) DOC = 30.88 ± 4.280 (n = 32) Tissue = 15.52 ± 1.85 (n = 166)
Functional group	Different functional groups of macroalgae differ in their cellular composition which may impact how they degrade	Foliose = 58.89 ± 13.88 (n = 49) Fucoid = 35.97 ± 9.78 (n = 76) Turf = 27.14 ± 3.04 (n = 118) Kelp = 74.74 ± 1.02 (n = 167) Removed: 14 “Mixed” observations	Foliose = 14.39 ± 4.072 (n = 32) Fucoid = 33.98 ± 4.469 (n = 38) Turf = 18.82 ± 2.906 (n = 80) Kelp = 18.46 ± 2.552 (n = 83) Removed: 1 “Mixed” observation
Classification	Taxonomic similarities within groups may drive differences in susceptibility to degradation	Red = 23.5 ± 3.99 (n = 62) Green = 26.7 ± 3.85 (n = 73) Brown = 64.73 ± 8.26 (n = 276) Removed: 14 “Mixed” observations	Red = 14.29 ± 4.218 (n = 29) Green = 19.83 ± 3.501 (n = 53) Brown = 22.01 ± 2.114 (n = 151) Removed: 1 “Mixed” observation
Average reported temperature across experiment duration (°C)	Higher temperatures accelerate decomposition in terrestrial systems, so we expect that macroalgae will degrade quicker at higher temperatures	n = 319	n = 172
Was the detritus pre‐treated?	Pre‐killing algae is expected to kickstart the degradation process, resulting in faster degradation rates	Yes = 25.31 ± 2.77 (n = 190) No = 71.3 ± 9.48 (n = 236)	Yes = 25.49 ± 2.852 (n = 78) No = 17.95 ± 2.028 (n = 156)
Degradation environment	The environment where degradation takes place affects the type of microbes available to degrade algae, as well as various abiotic factors, which may impact degradation rate	Suspended = 55.78 ± 15.57 (n = 52) Intertidal = 33.4 ± 5.23 (n = 64) Seafloor = 68.31 ± 15.13 (n = 120) Lab = 44.2 ± 6.24 (n = 190)	Suspended = 9.72 ± 3.487 (n = 37) Intertidal = 30.22 ± 4.523 (n = 37) Seafloor = 8.06 ± 2.492 (n = 60) Lab = 28.27 ± 2.519 (n = 100)
Mesh size of detritus litterbag (in mm)	Litterbags made from smaller diameter mesh (i.e., more tightly woven mesh) may reduce the amount of contract that the detritus has with grazers and/or microbes and may reduce the amount of light which can penetrate the litterbag. Both of these factors may impact degradation	n = 274	n = 148
Experiment duration (in days) a	Since organic matter typically exhibits a rapid rate of degradation during the initial phases of degradation, shorter studies are expected to report faster degradation rates as compared to studies with longer durations which are expected to capture the final, slower stages of degradation	n = 426	n = 234
Light availability	If light is available, detritus is able to continue to photosynthesize, delaying degradation	Dark = 33.18 ± 3.39 (n = 158) Lit = 66.28 ± 9.23 (n = 244)	Dark = 26.63 ± 5.364 (n = 82) Lit = 13.16 ± 2.270 (n = 135)

Note: We also present an a priori hypothesis as to why each factor might impact macroalgal degradation. For categorical variables, we list the group mean values ± standard error (sample size) for both half‐life and percent recalcitrance.

^a The impact of biomaterial type on degradation could not be analyzed statistically, and the effect of experiment duration was only analyzed in terms of its effect on macroalgal percent recalcitrance, not half‐life.

We performed a similar model‐selection process (best subset regression modeling based on AIC and C(p) values using the package olsrr; Hebbali, ²⁰²⁴) to analyze the effects of various factors on macroalgal percent recalcitrance (%). Since the percent recalcitrance data were highly zero‐inflated, we analyzed these data using a multivariate generalized linear mixed effects model (GLMM) with a zero‐inflated negative binomial distribution using the glmmTMB package (Brooks et al., 2017). The following variables were selected for inclusion in the percent recalcitrance GLMM: experimental temperature (°C), litterbag mesh diameter (mm; log₁₀ transformed), experiment duration (in days), light availability, pre‐treatment, algal functional group, algal class, and degradation environment (see Table 2). Study was also included as a random effect in this test.

Finally, because only select studies measured macroalgal carbon content, nitrogen content, and/or carbon:nitrogen ratios, we analyzed these variables in a separate test to determine if they predicted changes in macroalgal half‐life. We used a linear mixed effects model (using the lme4 package; Bates et al., ²⁰¹⁵) to analyze the impact of initial percent carbon (%), initial percent nitrogen (%), and initial carbon:nitrogen ratio on macroalgal half‐life, with study included as a random effect.

RESULTS AND DISCUSSION

Motivations for studying macroalgal decay

There has been sustained interest in the topic of macroalgal decomposition since 1976. Throughout this time, authors have described a variety of motives for studying macroalgal degradation dynamics, with the most common reasons for studying this topic being investigating nutrient and/or heavy metal release and cycling, the impact of meiofauna/changes to meiofaunal community throughout degradation, and the carbon sequestration potential of macroalgae (Table 1). The recent spike in publications on this topic is largely attributable to the current surge in interest in understanding the carbon sequestration potential of macroalgae (Figure 2).

FIGURE 2.

Histogram demonstrating the number of macroalgal degradation studies through time. Color‐coding represents the study motivation, meaning what the study authors state as their motive for conducting their study. Although interest in the topic of marine macroalgal degradation has been quite steady throughout time, there has been a recent uptick in studies that state that understanding macroalgal carbon sequestration potential (CSP) was a main aim of their study (N = 105 studies). Studies are binned into 5‐year periods based on publication date.

Methodological trends

Although the body of knowledge on the topic of macroalgal degradation is large, research efforts have not been evenly spread across geographic regions, macroalgal groups, or biomaterial types. Geographically, although macroalgal degradation studies have been conducted on all continents, the majority of work has occurred in the Northern hemisphere (82% of all observations), with Europe being particularly over‐represented (54% of all observations; Figure 3). Kelps, perhaps the most enigmatic group of macroalgae, have been the most commonly studied macroalgal group for degradation experiments (42% of all observations; Figure 4a), followed by turf algae (27% of observations). Additionally, data regarding macroalgal tissue degradation have made up the majority of observations included in this review (81% of observations), while other biomaterials have been more commonly overlooked—only 9.5% of observations are from DOC degradation experiments and 9% of observations are from POC or POM degradation experiments.

FIGURE 3.

Map showing the global distribution of macroalgal degradation studies. Studies on marine macroalgal degradation have been conducted on all continents. Although there is a bias toward Europe and the East coast of North America (N = 504 observations). Larger sized points indicate a higher number of overlapping observations at that location. Algal class is depicted as differently shaped and colored points.

FIGURE 4.

Frequency distributions of characteristics of macroalgal degradation studies. (a) macroalgal functional group, (b) general model type which best describes the degradation data, where gray coloring indicates models with a positive horizontal asymptote (indicating a recalcitrant portion of biomaterial), (c) habitat where degradation took place; in a lab or in situ either submerged on the seafloor, in the intertidal zone, or suspended in the water column, (d) mesh size (in mm) of bags containing macroalgal tissues (if applicable), (e) experiment duration in days, and (f) number of sampling events (not including the initial sample).

All macroalgal exudate (DOC, POC, and POM) degradation studies included in this review were conducted in a laboratory. This reflects the logistical constraints of measuring exudate degradation in situ as constant water movements prevent experimenters from being able to realistically measure exudate degradation over time in the open water column. Containing a portion of seawater in a closed chamber in situ and measuring exudate degradation is possible and would expose the samples to natural daily light and temperature fluxes (Wada et al., 2007; Watanabe et al., ²⁰²⁰); however, this approach prevents water circulation, which likely influences degradation dynamics due to a lack of introduction of new microbes and/or grazers as well as the potential depletion of oxygen and/or nutrients by microbes within the chamber. Although tracer molecules such as radiolabeled isotopes, stable isotopes, or eDNA can be used to track the movement of macroalgal exudates in situ (Hill & McQuaid, 2009; Ørberg et al., ²⁰²³; Queirós et al., ²⁰¹⁹; Sosik & Simenstad, ²⁰¹³), many of these approaches are currently limited in quantitative scope and are more useful for qualitative assessments (Beng & Corlett, 2020; Fonseca, ²⁰¹⁸; Roussel et al., ²⁰¹⁵). For example, these methods can identify a pathway from source to sink but are limited in their ability to accurately quantify the amount of macroalgal exudates that reach a given sink (Dolliver & O'Connor, ^2022a). Although the quantitative aspects of methodologies such as eDNA are improving (Lacoursière‐Roussel et al., 2016; Yates et al., ²⁰¹⁹), there are still many logistical restraints surrounding analysis of macroalgal exudate degradation in situ.

It is also important to note that there are two general approaches experimenters have utilized to study macroalgal‐derived exudate degradation. The experimenters can assess the changes in exudate concentration as macroalgal tissue degrades in the same container, or the experimenters can take a sample of seawater containing macroalgal‐derived exudates and measure how the exudate degrades (i.e., without macroalgal tissue also present). The problem with measuring degradation of macroalgal tissue and exudates in the same container is that the leaching of exudates from the macroalgal tissue confounds the degradation of the exudates. A researcher's choice of methodology will depend on the research question asked, since isolated exudate studies will provide high‐quality data pertaining to exudate degradation; however, combined leaching and degradation studies may give a more holistic understanding of the patterns of macroalgal biomaterial release and longevity.

Trajectory of degradation

For 37% of all observations (i.e., individual degradation trajectories), we were unable to assess the trajectory of macroalgal degradation, either because authors only assessed initial and final biomass/carbon change (25%) or because authors only included an estimate of degradation rate(s) and/or k value(s) without providing raw degradation data (12%). For observations wherein assessing degradation trajectory was possible, 23% followed a two‐parameter exponential decay model, 41% followed a three‐parameter exponential decay model (i.e., exponential decay with a horizontal asymptote), 11% of observations were confounded by algal growth, 9% were sigmoidal, and 16% followed a negative linear trend (Figure 4b). See Supplemental Datafiles 1 and 3 for all R‐squared values and model equations for each fitted model.

In the majority of cases, it was observed that an exponential decay model was the best fit for decomposition, affirming that exponential decay is usually suitable to describe macroalgal decay (Figure 4b). However, for 36% of all observations for which model fitting was possible, exponential decay was not the best fit. This demonstrates the substantial variation in the trajectories of macroalgal degradation and highlights the complexity of understanding drivers of macroalgal biomaterial longevity. This may be because, unlike most other forms of organic detritus, macroalgal biomaterial does not necessarily senesce once it is dislodged from the substrate or the main plant and can continue to photosynthesize and grow, which creates unpredictable degradation dynamics (Frontier et al., 2021; Wright & Foggo, ²⁰²¹; Wright & Kregting, ²⁰²³). Indeed, we categorized 11% of observations as being confounded by growth, meaning that macroalgal biomaterials increased in biomass during the experiment, creating difficulties with data analysis (Figure 1d). The findings of these studies are still important in the context of carbon sequestration, as resisting degradation by continuing to grow may be a pathway through which algae can eventually become sequestered (Frontier et al., 2021, 2022; Wright & Foggo, ²⁰²¹). More specifically, if macroalgal tissue or detached whole plants can continue to photosynthesize and grow, thereby avoiding degradation for a significant amount of time, this may increase the transport distance that those macroalgal biomaterials are able to cover, increasing the likelihood that those biomaterials end up in an area where sequestration is possible (i.e., soft sediments or the deep sea; Ager et al., ²⁰²³; Johnson & Richardson, ¹⁹⁷⁷; Kokubu et al., ²⁰¹⁹; Krause‐Jensen & Duarte, ²⁰¹⁶). This is particularly true if the macroalgal biomaterials remain relatively buoyant (Wernberg & Filbee‐Dexter, 2018).

The prevalence of sigmoidal and linear degradation trajectories (accounting for 25% of observations for which model fitting was possible) was surprising (Figure 4b). A sigmoidal curve was the best fit for 9% of all observations, despite the fact that it is not commonly used to describe organic matter decomposition (Figure 1e,f). Here, perhaps a mixture of degradation and growth was occurring simultaneously during the first portion of the experiment, which may be why there was negligible apparent change in biomass through time for the initial period of the experiment. Then, as a greater portion of the macroalgal matter senesces, the degradation rate increases as labile components are utilized, as in exponential decay models. Alternatively, microbial colonization may be delayed in these instances, and once a critical number of microbes colonize the macroalgal matter, degradation progresses more quickly. In cases where a linear model best described the data (16%), this might indicate that microbes and consumers were able to steadily degrade macroalgal biomaterial over time, perhaps due to high microbial density and/or diversity. However, this data may only appear to be linear because there was a low number of sampling events and/or a short experiment duration, which makes it difficult to discern the “true” trajectory of degradation (Figure S2).

Factors impacting macroalgal degradation

Overall, mean macroalgal half‐life was 50.79 ± 5.5 days (N = 426; Table S2), and mean percent recalcitrance (i.e., relative stability; see Glossary) was 20.44 ± 1.66% (N = 235). For 15% of observations, we could not estimate half‐life since more than 50% of the initial biomaterials remained once the experiment ended. Also, for the majority of observations (54%), we were not able to accurately assess recalcitrance of macroalgal biomaterials, which was due to either a lack of data or an experiment duration that was less than 1 month. Percent recalcitrance data were highly zero‐inflated; out of the 235 observations for which calculating percent recalcitrance was possible, 45% reported 0% recalcitrance. It is important to note that many studies included in this review (88% of in situ tissue degradation studies) utilized the litterbag technique (allowing macroalgal tissue to decay in a bag made of mesh); however, a shortfall of this approach is that it cannot account for the degradation of tissue fragments or exudates that escape through the gaps in the mesh, and these biomaterials may or may not, in fact, persist in the water column (Barrón et al., 2014; Krause‐Jensen & Duarte, ²⁰¹⁶; Paine et al., ²⁰²¹).

Our statistical analysis of macroalgal half‐life determined that increased temperatures predict shorter macroalgal half‐lives and that larger litterbag mesh sizes predict longer half‐lives (Table S3). Our analysis of macroalgal percent recalcitrance showed that brown algae had significantly higher mean percent recalcitrance than red algae and that pre‐treating macroalgae leads to a larger percentage of recalcitrant biomaterials (Table S4). However, the unequal and/or low sample sizes across certain groups may limit our statistical power in these tests (see Table 2). Also, the random effect of study was highly significant for the half‐life analysis (p < 0.00001), but there was no significant random effect of study for the percent recalcitrance model (p = 1). This indicates that although individual study characteristics often have a significant impact on a study's half‐life estimates, percent recalcitrance estimates do not seem to be impacted by distinct study methodologies/contexts.

Biomaterial type (DOC, POC, POM, or tissues)

Macroalgal‐derived DOC generally decays faster than tissues, POM, and POC (as evidenced by shorter mean half‐life; Table 2; Figure 5). Although DOC, POC, and POM generally had higher mean proportions of recalcitrant biomaterial compared to tissues (Table 2) the starkly uneven sample sizes across groups and heterogeneity of variances in this data prevented us from being able to statistically analyze these data.

FIGURE 5.

Violin plot displaying how macroalgal half‐life in days (panel a; data are log₁₀ transformed) and percent recalcitrance (panel b) vary across biomaterial types (dissolved organic carbon, DOC, particulate organic carbon, POC, particulate organic matter, POM, or macroalgal tissues).

Macroalgal taxonomy

Given that different types of macroalgae have quite distinct molecular constitutions, which potentially impact susceptibility to microbial attack (Lee & Ho, 2022; Popper et al., ²⁰¹¹; Shtein et al., ²⁰¹⁸), we had expected that macroalgal taxonomy would be an important predictor of degradation dynamics. However, although select studies included in this review investigated how algal taxonomy impacts algal degradation within their individual studies, their results were contradictory. One study reported that four turf algal tissue species (Ulva compressa, U. rigida, Gracilaria vermiculophylla, and Agardhiella subulata) all decayed faster than the fucoid Fucus vesiculosus (Conover et al., 2016). Conversely, two other studies reported that brown algae (F. vesiculosus and Undaria pinnatifida) decayed faster than turf algae (Ulva spp. in Hwang et al., ²⁰²³; Corallina officinalis and Ulva intestinalis in Kim et al., ²⁰²¹; Table S5).

Importantly, both Buchsbaum et al. (1991) and Hwang et al. (2023) have reported that brown algae had higher proportions of recalcitrant biomaterial, as compared with turf algae (Table S5). Furthermore, our analysis uncovered that brown algae had significantly higher proportions of recalcitrant biomaterials, but only when compared to red algae (p = 0.043; brown algae – green algae p = 0.2639; green algae – red algae p = 0.6424; Table S4; Figure 6). This finding is supported by the prevalence of particularly recalcitrant compounds in brown algae, such as polyphenolics, fucoidan, and alginate (Bligh et al., 2022; Buck‐Wiese et al., ²⁰²³; Deniaud‐Bouët et al., ²⁰¹⁴; Zhang et al., ²⁰²¹). However, we did not detect a significant effect of macroalgal taxonomy on macroalgal half‐life through our analysis (Table S3; Figure S3).

FIGURE 6.

Violin plot displaying how macroalgal percent recalcitrance varies with macroalgal class. Brown algae are significantly more recalcitrant, as compared to red algae.

Temperature

Increased temperature caused macroalgal degradation to progress faster, both within and across studies. Eleven studies included in this review independently reported that macroalgal decay quickened at higher temperatures (Table 3). For instance, Boldreel et al. (2023) observed that the decay constant (k value) was five times lower at 8°C than at 15°C for Saccharina latissima and 9.5 times lower for Alaria esculenta. Additionally, Conover et al. (2016) reported a three‐fold increase in the decay constant of Ulva rigida with an increase of only 2°C. However, four studies noted that temperature did not play a significant role in macroalgal degradation dynamics (Gao et al., 2021; Kotta et al., ²⁰¹⁰; Pedersen et al., ²⁰²¹; Smale et al., ²⁰²²), with Kotta et al. (2010) being the only study to observe that macroalgae decayed significantly faster in cooler water temperatures than warmer water temperatures (only for Pilayella littoralis; Cladophora glomerata decay was unaffected by temperature). In some cases (Smale et al., 2022), a narrow range of experimental temperatures (9.8–12.4°C) may have limited the detection of a difference in degradation caused by variation in temperature. Across all studies that reported experimental temperature, studies that exposed algae to higher mean temperatures resulted in significantly smaller half‐life values, meaning that macroalgal biomaterial degraded faster in warmer temperatures (p < 0.00001; Table S3; Figure 7). More specifically, our model predicted that a 1°C increase in temperature resulted in a 0.078 day decrease in macroalgal half‐life (Table S3). This may mean that ocean warming may lower the carbon sequestration potential of macroalgae (Filbee‐Dexter et al., 2022), although we did not find that temperature significantly influenced percent recalcitrance in our analysis (p = 0.3528; Table S4).

TABLE 3.

Summary of studies which investigated how continuous abiotic variables affect macroalgal degradation dynamics.

Abiotic factor	Positively correlated with degradation	Not correlated with degradation	Negatively correlated with degradation
Temperature	11 studies (Boldreel et al., 2023*; Conover et al., 2016; Dufour et al., 2012; Filbee‐Dexter et al., 2022; Frontier et al., 2022; Hanisak, 1993*; Kristensen et al., 1992; Litchfield et al., 2020; Paalme et al., 2002*; Pedersen & Johnsen, 2017*; Zielinski, 1981*) 15 species tested (9 browns, 3 reds and 3 greens)	4 studies (Gao et al., 2021; Kotta et al., 2010; Pedersen et al., 2021; Smale et al., 2022) 3 species tested (2 browns and 1 green)	1 study (Kotta et al., 2010) 1 species tested (1 brown)
Depth	3 studies (Duggins et al., 2016*; Frontier et al., 2022; Pedersen et al., 2021*) 3 species tested (3 browns)	1 study (Duggins et al., 2016*) 1 species tested (1 brown)	1 study (Salovius & Bonsdorff, 2004) 1 species tested (1 green)
Light	No studies	1 study (Filbee‐Dexter et al., 2022) 2 species tested (2 browns)	2 studies (Frontier et al., 2021, 2022) 2 species tested (2 browns)
Oxygen availability	2 studies (Paalme et al., 2002*; Pedersen et al., 2021) 2 species tested (2 browns)	1 study (Paalme et al., 2002*) 1 species tested (1 brown)	No studies
Salinity	2 studies (Franzitta et al., 2015; Lopes et al., 2011) 1 species tested (1 brown)	2 studies (Conover et al., 2016; Josselyn & Mathieson, 1980*) 2 species tested (1 brown and 1 green)	1 study (Josselyn & Mathieson, 1980*) 1 species tested (1 brown)
pH	No studies	1 study (Litchfield et al., 2020) 1 species tested (1 brown)	No studies
Microplastic concentration	No studies	No studies	1 study (Litchfield et al., 2020) 1 species tested (1 brown)
Water movement	No studies	1 study (Filbee‐Dexter et al., 2022) 2 species tested (2 browns)	No studies
Intertidal shore height	No studies	2 studies (Rodil et al., 2008*; Urban‐Malinga et al., 2008, a ) 3 species tested (3 browns)	No studies
Nutrient enrichment	2 studies (Robinson et al., 1982; Yang et al., 2021) 2 species tested (1 brown and 1 green)	No studies	No studies

Note: Studies which found that degradation rate was positively correlated, negatively correlated, or not correlated with an increase in the factor listed in that row are pooled. Asterisks denote studies which did not verify statistically whether the factor had a significant effect on degradation rate. Note that some studies reported differing results based on algal species, which is why some studies are listed in multiple columns within the same row.

^a This study observed that algal biomaterial degraded fastest at mid‐intertidal shore heights and degraded slowest at the lowest and highest intertidal shore locations.

FIGURE 7.

Scatterplot displaying how macroalgal half‐life (in days) varies with experimental temperature (n = 319). Macroalgal half‐life significantly decreases with increasing experimental mean temperature; therefore, higher temperatures accelerate macroalgal decay.

It has been well established in the terrestrial literature that increased temperatures accelerate organic matter decomposition, since higher temperatures increase the metabolic rate of decomposing organisms (Burke et al., 2004; Conant et al., ²⁰¹¹; Cornwell et al., ²⁰⁰⁸). Similar review papers on other types of marine macrophyte degradation also identified that higher temperatures increased degradation rates across various studies included in these reviews (mangroves, seagrasses, and tidal marshes in Ouyang et al., ²⁰²³; seagrasses in Trevathan‐Tackett et al., ²⁰²⁰). This contrasts with the findings of another review on mangrove degradation in which the authors did not report a significant effect of absolute latitude on degradation rates, which may be due to the narrower range in latitudes and, therefore, temperatures that mangroves grow in (Simpson et al., 2023). It is important to note that in addition to higher temperatures accelerating macroalgal degradation, higher temperatures also often increase the rate at which macroalgae release biomaterials into the water column via erosion and/or DOC release (Bennett et al., 2024; Endo et al., ²⁰²⁰; Krumhansl & Scheibling, ^2012b; Pessarrodona et al., ²⁰¹⁸). Ocean warming is therefore expected to increase the supply of macroalgal biomaterials to the surrounding environment while also increasing the rate at which those biomaterials degrade.

Depth, light, and oxygen availability

The water depth at which macroalgal degradation takes place may impact macroalgal degradation dynamics in situ, since light, temperature, and oxygen availability all typically decline with depth, and microbial and meiofaunal communities can also change with depth (Duggins et al., 2016; Matear & Hirst, ²⁰⁰³; Pawlowicz, ²⁰¹³; Reynolds & Lutz, ²⁰⁰¹; Zhang et al., ²⁰¹⁸). Three individual studies included in this review reported that macroalgal degradation accelerated at deeper depths (Duggins et al., 2016; Frontier et al., ²⁰²²; Pedersen et al., ²⁰²¹). One study documented that the effect of depth was dependent on algal species (Duggins et al., 2016) and another study observed that degradation slowed with increasing depth, which the authors attributed to the lower temperatures (~10°C cooler) to which deeper algal detritus was exposed (Salovius & Bonsdorff, 2004).

The acceleration of macroalgal degradation at deeper depths observed in prior studies may be attributed to the decrease in light availability deeper in the water column. This is because detached macroalgal tissues (i.e., detritus) can continue to photosynthesize and grow when there is adequate light (Frontier et al., 2021; Rothäusler et al., ²⁰¹⁸; Tala et al., ²⁰¹⁹; van Hees et al., ²⁰¹⁸; Wright et al., ²⁰²⁴; Wright & Foggo, ²⁰²¹; Wright & Kregting, ²⁰²³). Three studies included in this review assessed the impact of light on macroalgal degradation, and two of the three studies observed that increased light increased the longevity of macroalgal tissues (Table 3; Frontier et al., ^{2021, 2022}). In contrast, Filbee‐Dexter et al. (2022) did not find that light was not a significant predictor of macroalgal degradation in their field study. When we tested directly if light availability significantly impacted macroalgal half‐life or percent recalcitrance across studies (either dark or lit), a significant effect of light was not observed (p = 0.4077 for half‐life; p = 0.6163 for precent recalcitrance). However, we were not able to include observations that were confounded by growth (i.e., where growth negated degradation) in these analyses, since we were unable to extract half‐life or percent recalcitrance estimates from these data. Interestingly, all observations of macroalgal tissue degradation that were classified as being confounded by growth occurred in experiments wherein adequate light was available: Either experiments were performed at relatively shallow subtidal depths in situ (<10 m) or in a laboratory where artificial light sources were used. It is, therefore, evident that sustained detrital photosynthesis is an important aspect of macroalgal tissue degradation dynamics that should not be overlooked.

Oxygen availability also generally decreases with ocean depth (Lalli & Parsons, 1993). Two individual studies identified that decreased oxygen availability decreased macroalgal degradation rates, although this effect may be species‐dependent (Table 3; Paalme et al., ²⁰⁰²; Pedersen et al., ²⁰²¹). The slowing of degradation in low or no oxygen conditions is likely due to how restricting oxygen excludes microbes that cannot tolerate hypoxic or anoxic conditions, reducing the diversity and abundance of microbes available to degrade macroalgae (Bertagnolli & Stewart, 2018; Fenchel & Finlay, ²⁰⁰⁸; Glud, ²⁰⁰⁸). Trevathan‐Tackett et al. (2020) also documented that increasing oxygen availability increases seagrass degradation rates in their review.

Degradation environment

Macroalgal degradation has been investigated in a wide variety of conditions and contexts. Laboratory studies have been equally common as in situ studies. (Forty‐seven percent of studies were laboratory or mesocosm studies, 44% were conducted in situ, and 7% of studies included data from both in situ and laboratory experiments.) Macroalgal degradation can occur intertidally (after macroalgal tissues wash up on shore) or it can occur subtidally (either on the seafloor or while macroalgal biomaterials are suspended in the water column), and authors have investigated macroalgal degradation in all of these contexts (Figure 4c). Nedzarek and Rakusa‐Suszczewski (2004) and Zielinski (1981) both observed that intertidally placed algal biomaterial degraded faster than submerged biomaterial, but Josselyn and Mathieson (1980) documented that continuously submerged detritus degraded faster than detritus exposed to air (Table S6). Perhaps exposure to the air accelerates the senescence process, since this desiccates the tissue (Luo, Xie, et al., 2022). However, our statistical analysis did not indicate that degradation environment had a significant impact on degradation dynamics (Tables S3 and S4). Simpson et al. (2023) also did not find that the placement of litterbags containing mangrove litter (submerged, buried, or exposed to air) had a significant impact on degradation dynamics in their review.

Select studies included in this review observed that macroalgal biomaterial that is in direct contact with the seafloor sediment degrades faster compared to macroalgal biomaterial that is suspended in the water column (Table S6; Salovius & Bonsdorff, ²⁰⁰⁴; Williams, ¹⁹⁸⁴). This may be attributable to the biomaterial being in closer contact with the microbial and benthic grazer communities in the sediment, although this mechanism has yet to be validated. Furthermore, three studies included in this review investigated how burying macroalgal detritus versus placing detritus on the sediment surface impacts degradation dynamics (Table S6). Boldreel et al. (2023) observed that macroalgae degraded faster on the sediment surface, while Haram et al. (2020) and Kristensen and Mikkelsen (2003) both observed that macroalgae degraded slower when buried in sediment. Haram et al. (2020) and Kristensen and Mikkelsen (2003) both conducted their studies in the presence of grazers, whereas Boldreel et al. (2023) conducted their study in a lab in the absence of grazers. Thus, the presence of grazers is likely an important confounding factor in this context, since burying detritus presumably limits its availability to grazers that inhabit the sediment surface. Relatedly, in the one study directly investigating substratum type (rocky reef compared to soft sediment) on degradation rate, substratum did not impact degradation rate (Hunter, 1976).

Litterbag mesh diameter

Eighty‐eight percent of in situ macroalgal tissue degradation studies chose to use mesh bags (litterbags) or mesh cages to contain macroalgal biomaterial in the field, with litterbag mesh size ranging from 53 μm to 5 cm in diameter (median = 1 mm, mean = 4.4 mm; Figure 4d). Litterbag mesh size may influence degradation rate since extremely small, tight mesh can reduce the amount of water and/or sediment in contact with the biomaterial, limiting the exposure of biomaterials to microbes and/or grazers, and reducing light availability (Handa et al., 2014; Lecerf, ²⁰¹⁷; Xie, ²⁰²⁰). Larger sized mesh may increase the biomaterial's susceptibility to grazing and microbial attack, since it increases the amount of contact the biomaterial has with the sediment and/or water column, as well as potentially allowing more small/particulate detrital matter to escape through the mesh (Chassain et al., 2021; Harrison, ¹⁹⁸⁹). Hence, the litterbag mesh diameter chosen by a researcher may impact that researchers' subsequent measurements of degradation dynamics.

Three studies included in this review independently assessed if litterbag mesh size impacted macroalgal degradation dynamics. Two of the studies observed that mesh size had no impact on degradation dynamics (Catenazzi & Donnelly, 2007; Wright & Kregting, ²⁰²³), but Bedford and Moore (1984) observed that macroalgae degraded faster in 1‐mm diameter mesh bags, compared to 50‐mm diameter mesh bags (Table S6). Our analysis showed that increasing litterbag mesh diameter leads to slower macroalgal decay (p = 0.0411), wherein a 1‐mm increase in mesh size predicts a 0.44 day increase in macroalgal half‐life (Table S3; Figure 8). This may be because larger‐sized mesh allows more light to reach macroalgal detritus, which allows detrital tissue to continue to photosynthesize and avoid degradation (Frontier et al., 2021). However, litterbag mesh diameter was not a significant predictor of macroalgal percent recalcitrance (p = 0.0776). Contrastingly, a systematic review of terrestrial organic matter degradation studies observed that increasing litterbag mesh size significantly accelerates terrestrial leaf litter degradation, as evidenced by higher k‐values (Xie, 2020).

FIGURE 8.

Scatter plot demonstrating the relationship between litterbag mesh size (mm) and macroalgal half‐life (in days); half‐life increases (meaning degradation slows) with increasing mesh size.

Experiment duration

Most studies included in this review had relatively short durations (mean = 54 days, median = 35 days; Figure 4e). Short experiment durations lower the confidence with which we can incorporate macroalgal degradation data into predictive models, which could estimate the ultimate fate of macroalgal biomaterials. This is because there is wide variation in coastal residence time (i.e., time it takes for coastal waters to mix with the open ocean), ranging from just a few days to weeks, sometimes up to years (Filbee‐Dexter et al., 2024; Liu et al., ²⁰¹⁹; Safak et al., ²⁰¹⁵), so understanding the long‐term degradation dynamics of macroalgae is crucial in the context of carbon sequestration. Furthermore, experiment duration may impact our perception of degradation dynamics since short experiments may only capture the initial stages of degradation, excluding the posterior stages of degradation which are typically characterized by slower degradation rates (Figure S2). Longer experiment durations also enable researchers to make more accurate estimates of how much macroalgal biomaterial is recalcitrant. However, our analysis did not uncover a significant effect of experiment duration on macroalgal percent recalcitrance (p = 0.3993; Table S4; half‐life data could not be analyzed). Trevathan‐Tackett et al. (2020) noted in their review that studies on seagrass degradation with durations of <120 days had a two‐fold increase in average half‐life, compared to studies >120 days in duration. Our macroalgal data show a similar, albeit much less pronounced trend, wherein studies that were longer than 120 days reported a mean half‐life of 61.40 ± 5.79 days, whereas studies shorter than 120 days reported a shorter mean half‐life of 50.34 ± 18.9 days.

Pre‐treatment of macroalgae

Almost half of all studies pre‐treated macroalgal biomaterial before experimentation began (42%), commonly by oven‐drying or freeze‐drying macroalgae (Josselyn & Mathieson, 1980; Luo, Dai, et al., ²⁰²²; Smith & Foreman, ¹⁹⁸⁴). This is often done with the justification of ensuring that the macroalgae is dead, as well as minimizing morphological differences between species for cross‐species comparisons of degradation rates (Kim et al., 2021; Lopes et al., ²⁰¹¹). It also enables researchers to record initial dry weight measurements of macroalgal tissue. Eight studies included in this review independently assessed the impact of pre‐treating macroalgae on subsequent macroalgal degradation dynamics, and more than half (57%) observed that degradation quickened after macroalgae were pre‐treated, compared to un‐treated macroalgae, whereas 28% documented a negligible effect (Table S6). Our analysis showed that macroalgal half‐life was not significantly impacted by pre‐treatment (p = 0.3549). Interestingly, our analysis did find that macroalgae that were pre‐treated had significantly higher proportions of recalcitrant biomaterials compared to macroalgae that were not pre‐treated (p = 0.00691; Figure 9). The mechanism behind this difference is unknown, although it is possible that different pre‐treatment methods (including oven drying, air drying, freezing, and others) impact the stability of recalcitrant biomaterials, ultimately altering degradation dynamics.

FIGURE 9.

Violin plot displaying how macroalgae which had been pre‐treated have higher percent recalcitrance (%) as compared to macroalgae which were not pre‐treated.

Comparison to other marine macrophytes

Many studies included in this review compared tissue degradation dynamics between vascular marine plants and macroalgae, and all of these studies observed that macroalgae degraded faster than seagrasses (Bourguès et al., 1996; Buchsbaum et al., ¹⁹⁹¹; Gladstone‐Gallagher et al., ²⁰¹⁶; Josselyn & Mathieson, ¹⁹⁸⁰; Kristensen, ¹⁹⁹⁴; Lanari, Claudino, et al., ²⁰¹⁸; Liu et al., ²⁰²⁰; Lopes et al., ²⁰¹¹; Rice, ¹⁹⁸²; Rice & Tenore, ¹⁹⁸¹; Thomson et al., ²⁰²⁰; Twilley et al., ¹⁹⁸⁶; Vichkovitten & Holmer, ²⁰⁰⁴) and faster than mangroves (Gladstone‐Gallagher et al., 2016; Kristensen, ¹⁹⁹⁴; Rice, ¹⁹⁸²; Rice & Tenore, ¹⁹⁸¹). Buchsbaum et al. (1991) observed that phenolic content decreased at a faster rate in the algae Fucus vesiculosis than in seagrasses throughout degradation time, which might explain this trend. In other words, macroalgal phenolic compounds may be more susceptible to degradation as compared to seagrass phenolic compounds. Also, seagrasses and mangroves contain particularly recalcitrant cell structural compounds such as lignocellulose, which are especially hard for microbes to degrade and are absent from macroalgae (Benner & Hodson, 1985; Klap et al., ²⁰⁰⁰).

Furthermore, after comparing the results of our study to those of similar review papers on marine macrophyte degradation, macroalgae decay faster than other forms of marine macrophyte detritus, on average, as evidenced both by a faster mean degradation rate and a larger mean decay constant (Lovelock et al., 2017; Ouyang et al., ^{2017, 2023}; Simpson et al., ²⁰²³; Table S7; Figure 10). Ouyang et al. (2023) compiled litter degradation data across different types of coastal vegetated ecosystems in their review and observed that macroalgae decay was, on average, the fastest, as evidenced by a higher mean decay constant (k). The mean k value for macroalgae was highest at 20 ± 2 × 10⁻² · day⁻¹, while the mean k value for mangroves, seagrasses, and tidal marsh plants were 1.5 ± 0.1 × 10⁻² · day⁻¹, 1.6 ± 0.1 × 10⁻² · day⁻¹, and 6.0 ± 0.5 × 10⁻³ · day⁻¹, respectively (Ouyang et al., 2023).

FIGURE 10.

Comparison of mean decay constant values reported by similar systematic reviews of marine macrophyte degradation. Mangrove litter mean k values are extracted from (Simpson et al., 2023), salt marsh and seagrass litter mean k values are derived from (Ouyang et al., 2023), and macroalgae mean k value is taken from this review. Error bars are standard error; sample sizes are included above each point.

Macroalgal tissue composition

Carbon and nitrogen content, as well as carbon:nitrogen (C:N) ratios, were often measured in macroalgal detritus degradation studies, as this gives insight into the quality of macroalgae as a food source, as well as the rate that these two key nutrients are released to the ecosystem throughout degradation (Norderhaug et al., 2006). Thirty‐seven percent of studies included in this review reported how carbon and nitrogen content changed during degradation. We had predicted that higher initial nitrogen content and lower initial carbon content would predict faster degradation, since high nitrogen content in primary producer tissues is thought to be more palatable to microbes and grazers who are often nitrogen limited (Enríquez et al., 1993; Sterner & Hessen, ¹⁹⁹⁴). Filbee‐Dexter et al. (2022) and Wright et al. (2022) both observed that higher initial carbon content in macroalgal tissues predicted slower degradation, although in both cases initial carbon content only partially explained variations in macroalgal degradation across sites and/or species. When we tested whether initial carbon content, initial nitrogen content, or initial carbon:nitrogen ratios significantly predicted macroalgal biomaterial half‐life using a LMM with study as a random effect, there were no significant effects (p = 0.66 for C:N; p = 0.82 for carbon; p = 0.43 for nitrogen; Table S8).

Across marine, aquatic, and terrestrial plants, C:N ratios typically decline with degradation time (Enríquez et al., 1993). Many studies on macroalgal degradation have reported that macroalgal detritus becomes enriched in nitrogen as decomposition progresses, thus reducing the C:N ratio (Birch et al., 1983; Brouwer, ¹⁹⁹⁶; Hwang et al., ²⁰²³; Krumhansl & Scheibling, ^2012a; Norderhaug et al., ²⁰⁰⁶; Smith & Foreman, ¹⁹⁸⁴; Twilley et al., ¹⁹⁸⁶). Nitrogen enrichment of detritus is usually attributed to the microbial transformation of detritus throughout degradation (Norderhaug et al., 2003; Tenore et al., ¹⁹⁷⁹), although some macroalgal degradation studies have reported an increase in the C:N ratio throughout degradation (Pedersen et al., 2021; Urban‐Malinga et al., ²⁰⁰⁸; Vichkovitten & Holmer, ²⁰⁰⁴). Across all studies included in this review that measured C:N, 64% reported a decrease of greater than 10% in C:N, whereas 16% reported an increase in C:N of more than 10%. The average percent change in C:N across the entire degradation experiment is −21.37% (N = 140), which may mean that macroalgae become less palatable to consumers throughout degradation time since grazers prefer C:N ratios that are not too high nor too low (Norderhaug et al., 2006).

Brown algae synthesize polyphenolics (including phlorotannins), which are thought to protect their tissues from UV damage, microbial attack, and grazing (Duggins & Eckman, 1997; Geiselman & McConnell, ¹⁹⁸¹; Paul et al., ²⁰⁰⁶). Thus, many researchers have been interested in whether polyphenolic content has a predictive or explanatory role in brown algal degradation. Wright et al. (2022) observed that higher polyphenolic content predicted slower decomposition rates both across various Laminarian species as well as between specific individuals of the same species. Similarly, Pedersen et al. (2021) observed that kelp stipes contained higher portions of polyphenolics and decayed slower than kelp blades, which had lower polyphenolic concentrations. However, Gilson et al. (2021) observed that polyphenolic content did not predict differences in degradation rates between kelp species. Therefore, while measuring polyphenolic content can give some insight into brown algal degradation dynamics, polyphenolic concentrations do not fully explain brown algal degradation dynamics. This may be because there are many forms of polyphenolics, and some forms are more labile than others (Paul et al., 2006). However, polyphenolics may leach from the algae during degradation, removing their protective capacity during the later stages of degradation (Norderhaug et al., 2006; Pedersen et al., ²⁰²¹). In summary, although C:N ratio typically declines as macroalgal degradation progresses, reducing the palatability of macroalgal detritus, the decline in polyphenolics during brown algal degradation may make detritus less chemically defended and therefore more accessible to consumers as degradation progresses.

Impact of microbial community on degradation

The microbial community is very likely an important driver of macroalgal degradation, as many studies have shown that degradation progresses faster when microbes are present or present in higher concentrations (Li, Feng, Xiong, He, et al., 2023; Li, Feng, Xiong, Shao, et al., ²⁰²³; Lucas et al., ¹⁹⁸¹; Perkins et al., ²⁰²³; Stuart et al., ¹⁹⁸¹). Moreover, the primary potential mechanism behind the strong impact of temperature on organic matter degradation is that temperature accelerates microbial metabolisms (Burke et al., 2004; Cornwell et al., ²⁰⁰⁸). Also, the high variability in macroalgal degradation dynamics across studies may be partially attributable to variations in microbial communities caused by certain experimental choices that have been taken in laboratory studies (including whether natural or artificial seawater was used, whether microbes were artificially enriched/spiked, or whether natural sediment was included in experimental tanks; Huang et al., ²⁰²³; Li et al., ²⁰²²; Rieper‐Kirchner, ¹⁹⁸⁹).

Previous studies have investigated how the microbial community changes throughout macroalgal degradation. Bacterial abundance usually increases during the early stages of degradation, and then begins to gradually decline, presumably since the amount and accessibility of biomaterial decrease with time (Chen et al., 2020; Lucas et al., ¹⁹⁸¹; Manikandan et al., ²⁰²¹; Rieper‐Kirchner, ¹⁹⁸⁹). Bacteria from the phylum Proteobacteria typically dominate the bacterial community in the preliminary stages of degradation, then the bacterial community generally diversifies as degradation progresses (Brunet et al., 2021; Chen et al., ²⁰²⁰; Feng et al., ²⁰²²; Hu et al., ²⁰²³; Ihua et al., ²⁰¹⁹; Liang et al., ²⁰²¹; Lozada et al., ²⁰²³; Xie et al., ²⁰²⁴; Zhang et al., ²⁰²⁴; Zhang, Qin, et al., ²⁰²²). Across these studies, algal‐degrading bacterial communities vary substantially with environment, location, oxygen availability, and algal species. Common algal‐degrading bacteria that increase in prominence in later stages of degradation include bacteria from the genera Flavobacterium (known for their fucoidan‐degrading capacity; Brunet et al., ²⁰²¹; Liang et al., ²⁰²¹; Lozada et al., ²⁰²³; Sakai et al., ²⁰⁰²) as well as bacteria from the phylum Planctomycetes (known for their ability to degrade sulphated polysaccharides; Hu et al., ²⁰²³; Lage & Bondoso, ²⁰¹⁴; Lozada et al., ²⁰²³; Wegner et al., ²⁰¹³; Xie et al., ²⁰²⁴). Also, Perkins et al. (2023) noted that fungi play an important role in kelp degradation, since after the addition of fungicide, kelp degradation progressed significantly slower.

Key knowledge gaps

It remains unclear how persistent macroalgal‐derived carbon is within the ocean, preventing macroalgae from being included in current blue carbon budgets (Pessarrodona et al., 2023). One key area of uncertainty that remains is the understanding of macroalgal‐derived exudate (DOC, POC, and POM) degradation dynamics. There is reason to suspect that macroalgal‐derived DOC persisting in the ocean may be a key pathway for macroalgal carbon sequestration (Pessarrodona et al., 2023); there is a vast oceanic DOC pool that has remained relatively stable throughout time (Baltar et al., 2021; Cai & Jiao, ²⁰²³), and macroalgae are one of the main sources of DOC that contribute to coastal DOC pools (Lønborg et al., 2020; Wada & Hama, ²⁰¹³). Additionally, although the effect of temperature on macroalgal degradation is quite well established, the effects of other abiotic factors like depth, oxygen concentration, substratum type, salinity, and pH on degradation have been variable across contexts and need further study (Table 3 and Table S6). This is particularly important considering that we may expect these abiotic factors to change as climate change progresses, necessitating more study on how these factors impact macroalgal degradation dynamics.

Macroalgae represent a highly diverse, polyphyletic group, in contrast with other vegetated blue carbon ecosystems that are relatively more homogeneous (mangroves, seagrasses, and tidal marshes; Krause‐Jensen et al., ²⁰¹⁸). This diversity likely causes variations in degradation trajectories, since different macroalgal taxonomic and functional groups have distinct cell wall compounds (Lee & Ho, 2022; Popper et al., ²⁰¹¹; Synytsya et al., ²⁰¹⁵). This diversity may explain why macroalgae k‐values are more variable across the literature when compared with other more uniform types of marine macrophytes (Figure 10). Our analysis showed that brown algae are more recalcitrant when compared to red algae (with no significant difference when compared to green algae). This finding should be interpreted with caution, however, due to the over‐representation of brown algae in the data set (and corresponding under‐representation of other algal groups), highlighting the need for more study on other algal groups. Indeed, only 99 species of macroalgae have been assessed for their degradation dynamics (for a full list of species see Supplemental Datafile S6), which is only a tiny fraction of the thousands of species of macroalgae currently recognized (Guiry, 2024). Thus, many species of macroalgae remain unstudied, limiting our ability to make well‐founded conclusions about trends in degradation across all macroalgae. Particularly, due to the loss of kelps and fucoids in many ecosystems, species of turfing algae and crustose coralline algae are becoming increasingly dominant (Filbee‐Dexter & Wernberg, 2018; Matsunaga et al., ¹⁹⁹⁹; Rogers‐Bennett & Catton, ²⁰¹⁹; Smith et al., ²⁰²⁴), so understanding how these groups of algae contribute to coastal carbon cycling is paramount. Additionally, coralline algae are very thermally stable (Trevathan‐Tackett et al., 2015), meaning that they may be important contributors to carbon sequestration; however, whether the carbon dioxide produced by coralline algae during CaCO₃ production exceeds their contributions to carbon sequestration is contested (Smeaton et al., 2017; van der Heijden & Kamenos, ²⁰¹⁵).

Recommended methods

In future studies, striving to fit models that accurately explain macroalgal degradation data is imperative in order to understand the longevity of macroalgal biomaterials in the ocean (Lian et al., 2023; Pessarrodona et al., ²⁰²³). This review has shown that macroalgal degradation is not always best explained by exponential decay models, despite the common assumption that all organic matter degrades following an exponential decay curve (Berg & McClaugherty, 2003; Wider & Lang, ¹⁹⁸²). Crucially, carefully fit models will enable us to accurately assess if there is recalcitrant matter, which will lead to a better understanding of the carbon sequestration potential of macroalgae (Hurd et al., 2022; Pessarrodona et al., ²⁰²³). In this vein, increasing the experimental duration of degradation experiments to at least 1 month, preferably longer, would enable more precise model fitting and estimation of recalcitrance while also reducing the risk of overfitting models. It is also important to sample decaying macroalgal biomaterials frequently throughout their degradation times, as this improves our ability to fit models to degradation trajectories; at least four to five datapoints is ideal for subsequent model fitting. Currently, 25% of studies included in this review made only one final assessment of degradation (plus an initial assessment), which severely limited our ability to understand degradation trajectory and estimate percent recalcitrance (Figure 4f).

Pre‐treating macroalgae prior to experimentation may limit the confidence with which studies' findings can be used to understand how algae decay in situ, particularly since we determined that pre‐treating algae significantly influences macroalgal percent recalcitrance. Thus, we do not recommend pre‐treating macroalgae prior to degradation if one's goal is to understand the carbon sequestration potential of macroalgae better. Likewise, since litterbag mesh size impacts macroalgal degradation dynamics, size should be considered carefully to enhance repeatability and ecological relevance. The median mesh diameter used by studies included in this review was 1 mm, so moving forward, using 1‐mm diameter mesh in macroalgal degradation litterbag studies is a reasonable choice.

Lastly, in the endeavor to estimate the carbon sequestration potential of macroalgae, one key aspect future researchers should focus on is the relative recalcitrance of macroalgal organic matter. There are many alternate approaches researchers can use to assess the relative recalcitrance of organic matter, which include thermogravimetric analyses of tissue stability (Trevathan‐Tackett et al., 2015), quantifying the amount of already known recalcitrant compounds in tissues (which can be done through a vast variety of methods, reviewed in Cai & Jiao, ²⁰²³), or taking marine sediment core samples to categorize the age and source of carbon contained within them (Hansen et al., 2022; Ørberg et al., ²⁰²³; Queirós et al., ²⁰²³). The method used in this review, fitting models to degradation data and determining at what point the degradation rate approaches zero, is beneficial, as it can give us an idea of recalcitrance by making use of already collected data. However, it is important to note that an inherent issue with the common in situ litterbag/mesh bag approach is that these studies cannot track the fate of the biomaterial (whether it be tissue, POM, POC, or DOC) that escapes through the holes in the mesh. Conducting more studies that measure the longevity of macroalgal POM, POC, and DOC will enable a better understanding of the recalcitrance of all types of macroalgal biomaterials. However, recalcitrance is not a straightforward concept, as it can depend on a myriad of factors (Jiao et al., 2014; Schmidt et al., ²⁰¹¹; Shen & Benner, ²⁰¹⁸). Therefore, more work is necessary to determine the mechanisms that confer recalcitrance in macroalgae, including the relative importance of molecular stability, abiotic factors, and microbial community.

CONCLUSIONS

Macroalgal degradation is a complex process with a variety of important implications for ecosystem functioning and marine carbon cycling (Lastra et al., 2008; Ouyang et al., ²⁰²³; Renaud et al., ²⁰¹⁵; Watanabe et al., ²⁰²⁰). By summarizing the current literature on the topic of macroalgal degradation, we aimed to shed light on this important topic and highlight future directions of research. One of our main findings was that there is substantial variation in macroalgal decay trajectories, meaning that authors should select and compare appropriate macroalgal decay curves in the future to garner a more precise understanding of macroalgal degradation and relative recalcitrance (i.e., longevity). Next, we identified a variety of factors that impact degradation dynamics in macroalgae: Increasing temperature accelerates macroalgal degradation; larger litterbag mesh sizes slow macroalgal degradation; brown algae are more recalcitrant than reds; and pre‐treating macroalgae prior to experimentation impacts estimates of percent recalcitrance. The impacts of other drivers on macroalgal degradation (including biomaterial type, functional group, degradation environment, and experiment duration) vary across contexts and, therefore, ought to be investigated further. In addition, we highlighted that detrital photosynthesis is an important aspect of macroalgal degradation, that macroalgal tissue composition (including carbon content and polyphenolic content) does not fully explain variations in macroalgal degradation dynamics, and that the microbial community is a major driver of macroalgal degradation. Lastly, we emphasized that many knowledge gaps persist, such as the lack of geographic and taxonomic diversity in the literature and the lack of study on macroalgal DOC and POM/POC degradation.

The extent of the contribution of macroalgae to marine carbon sequestration remains uncertain due to the complexities involved in tracking macroalgal‐derived carbon from its source to a sink where carbon sequestration is possible (i.e., soft sediments or the deep sea; Dolliver & O'Connor, ^2022a; Fujita et al., ²⁰²³; Krause‐Jensen et al., ²⁰¹⁸). Garnering a better understanding of macroalgal degradation will advance our understanding of macroalgal carbon sequestration because the longevity of macroalgal biomaterials influences the likelihood that those biomaterials reach habitats where carbon sequestration is probable (Filbee‐Dexter et al., 2024; Pessarrodona et al., ²⁰²³; Queirós et al., ²⁰²³). In the case of macroalgal‐derived DOC, we lack evidence that a substantial amount of macroalgal DOC can persist in the ocean for 100 years or more (Paine et al., 2021). Therefore, understanding the long‐term persistence of macroalgal DOC will greatly expand our capacity to quantify the contribution of macroalgae to marine carbon sequestration (Cai & Jiao, 2023; Lønborg et al., ²⁰²⁰). Thus, further targeted study of the degradation dynamics and recalcitrance of a broad range of macroalgal biomaterials under a broad range of conditions will improve our understanding of the carbon sequestration capacity of macroalgae in addition to furthering our knowledge of the role of macroalgal biomaterials as trophic subsidies and as crucial habitat and food sources for meiofaunal communities.

GLOSSARY

Blue carbon: carbon which is sequestered by marine ecosystems, examples of blue carbon ecosystems include mangroves, seagrass meadows and salt marshes.

Carbon sequestration: the capture and storage of carbon in an inert form, usually for at least 100 years.

Detritus: here we use this term to refer to macroscopic tissue debris released from macroalgae which could range from small tissue fragments to an entire, dislodged algae (alternate terms: wrack or litter).

Degradation: the transformation of organic molecules contained within organic matter into inorganic molecules, typically done by microorganisms. This process results in the release of nutrients to the system, as well as the release of carbon dioxide (alternate terms: decomposition or decay).

Dissolved organic carbon (DOC): organic carbon that is dissolved in a body of water, typically referring to carbon that can pass through either a 0.2‐ or 0.7‐μm filter.

Dissolved organic matter (DOM): refers to all fractions of dissolved organic matter in a body of water that can pass through a either a 0.2‐ or 0.7‐μm filter. Note that all studies included in this review chose to analyze change in macroalgal‐derived DOC, rather than DOM.

Exudates: defined here as macroalgal‐derived DOC, DOM, POM, and POC inclusively.

Foliose algae: defined here as macroalgae larger than turf and not within the kelp or fucoid functional groups.

Fucoid: brown algae from the order Fucales.

Kelp: brown algae from the order Laminariales.

Labile: organic matter which is susceptible to degradation.

Macroalgal biomaterial: we use this term throughout to refer to any organic matter derived from macroalgae that is released into the ocean. This includes the macroalgal exudates—DOM (including DOC), POM (including POC)—as well as detritus (macroalgal tissue fragments and whole detached plants).

Meiofauna: small invertebrates which live in benthic marine and/or freshwater environments that are smaller than macrofauna but larger than microfauna.

Particulate organic carbon (POC): particulate organic carbon from a freshwater or marine source. The operational definition of particulate organic matter/carbon varies substantially across the body of literature included in this review. Some authors define particulate matter/carbon as suspended particles that are caught on a filter (either 0.2 or 0.7 μm) or they define particulate matter/carbon as being smaller than a certain threshold size, ranging between 10 and 750 μm.

Particulate organic matter (POM): refers to all fractions of particulate organic matter from a freshwater or marine source. The operational definition of particulate matter varies across the literature included in this review, see above.

Recalcitrant: organic material which persists in the environment, avoiding degradation or ingestion (alternate term: refractory). There is no standard length of time for which organic matter must resist degradation for it to be deemed recalcitrant. In this paper, we designated that there is a portion of recalcitrant biomaterial if the degradation trajectory had a positive horizontal asymptote and the experiment lasted longer than 1 month.

Turf algae: low lying, dense aggregations of foliose or filamentous algae of various taxonomic groups.

AUTHOR CONTRIBUTIONS

Jessica R. Kennedy: Data curation (equal); formal analysis (lead); investigation (equal); methodology (equal); visualization (lead); writing – original draft (lead); writing – review and editing (equal). Caitlin O. Blain: Conceptualization (lead); data curation (equal); funding acquisition (lead); investigation (equal); methodology (equal); supervision (lead); writing – review and editing (equal).

Supporting information

Figure S1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) flow diagram for this systematic review.

Figure S2. Example degradation trajectory that exhibits a three‐parameter exponential decay relationship (black circles). If the experimental duration is confined to the initial decay period (e.g., red triangles), the degradation trajectory may appear linear, and recalcitrance is challenging to estimate. Longer experiment durations are beneficial as it enables us to understand the longevity of macroalgal biomaterial.

Figure S3. Violin plots demonstrating how macroalgal half‐life (log₁₀ transformed) varies across macroalgal functional groups (a) and macroalgal classification (b). Neither class nor functional group was a significant predictor of half‐life.

Table S1. Details of the systematic search conducted by the authors including details on who conducted searches and when and the exact search terms used for each database.

Table S2. Global mean values for all macroalgal degradation studies included in this review. Decay constant values were only extracted from models that were best described by exponential decay curves and degradation rate values were only estimated from data described by linear models. Data are presented as mean ± standard error, sample size refers to the number of individual observations.

Table S3. To analyze the impacts of various factors on macroalgal half‐life we used a multifactor linear mixed effects model, with study included as a random effect. The model formula was: half‐life ~ functional group + class + experimental temperature + degradation environment + litterbag mesh size + light availability + pretreatment + (1|reference). Both half‐life and litterbag mesh size data were log‐transformed. The final row refers to the significance of the random effect of study.

Table S4. To analyze the impacts of various factors on macroalgal recalcitrance (%) we used a multifactor generalized linear mixed effects model with a zero‐inflated negative binomial distribution, and study included as a random effect. The model formula was: percent recalcitrance (%) ~ functional group + class + experimental temperature + experiment duration + degradation environment + litterbag mesh size + light availability + pretreatment + (1|reference). Litterbag mesh size data were log‐transformed. The final row refers to the significance of the random effect of study.

Table S5. Findings of studies which compared how algae from different functional groups degrade. In the half‐life column, species are listed in order of shortest to longest half‐life, in days. In the recalcitrance column, species are listed in order of smallest to biggest portion of percent recalcitrant material. Comparisons are based on this review's independent estimates of half‐life and percent recalcitrance. Turf algal species are in green text, kelps are brown, foliose are black, and fucoids are blue. Asterisks denote studies which did not test if the difference in degradation dynamics were statistically significant.

Table S6. A summary of the findings of prior studies which tested how a variety of discrete methodological and environmental factors impact macroalgal degradation dynamics. Asterisks denote studies which did not verify statistically whether the factor effected degradation.

Table S7. Comparison of the results of our review and the findings other comprehensive reviews on marine biomaterial decomposition. Data are presented as mean ± standard error or as mean (5th percentile‐95th percentile) or as just the mean value. N.S. stands for not specified.

Table S8. Linear mixed effects model analyzing the effect of carbon and nitrogen content on macroalgal half‐life. Formula: Half‐life ~ Initial C:N ratio + Initial percent carbon + Initial percent nitrogen + (1|Study).

Appendix S1. Systematic review reference list.

Kennedy, J. R. , & Blain, C. O. (2025). A systematic review of marine macroalgal degradation: Toward a better understanding of macroalgal carbon sequestration potential. Journal of Phycology, 61, 399–432. 10.1111/jpy.70031

DATA AVAILABILITY STATEMENT

All datas (Supplemental Datafiles 1–6) have been uploaded to the following link and are publicly accessible for viewing and downloading: https://osf.io/vmtfd/files/osfstorage; DOI: 10.17605/OSF.IO/VMTFD