A review of subtidal kelp forests in Ireland: From first descriptions to new habitat monitoring techniques

Abstract Aim Kelp forests worldwide are important marine ecosystems that foster high primary to secondary productivity and multiple ecosystem services. These ecosystems are increasingly under threat from extreme storms, changing ocean temperatures, harvesting, and greater herbivore pressure at regional and global scales, necessitating urgent documentation of their historical to present‐day distributions. Species range shifts to higher latitudes have already been documented in some species that dominate subtidal habitats within Europe. Very little is known about kelp forest ecosystems in Ireland, where rocky coastlines are dominated by Laminaria hyperborea. In order to rectify this substantial knowledge gap, we compiled historical records from an array of sources to present historical distribution, kelp and kelp forest recording effort over time, and present rational for the monitoring of kelp habitats to better understand ecosystem resilience. Location Ireland (Northern Ireland and Éire). Methods Herbaria, literature from the Linnaean society dating back to late 1700s, journal articles, government reports, and online databases were scoured for information on L. hyperborea. Information about kelp ecosystems was solicited from dive clubs and citizen science groups that are active along Ireland's coastlines. Results Data were used to create distribution maps and analyze methodology and technology used to record L. hyperborea presence and kelp ecosystems within Ireland. We discuss the recent surge in studies on Irish kelp ecosystems, fauna associated with kelp ecosystems that may be used as indicators of ecosystem health and suggest methodologies for continued monitoring. Main Conclusions While there has been a steady increase in recording effort of the dominant subtidal kelp forest species, L. hyperborea, only recently have studies begun to address other important eco‐evolutionary processes at work in kelp forests including connectivity among kelp populations in Ireland. Further monitoring, using suggested methodologies, is required to better understand the resilience of kelp ecosystems in Ireland.


| BACKG ROU N D
Suringar; Kraan, 2017). Along semi-exposed rocky coastlines, L. hyperborea forms dense forests, whereas in the calmer, shallow regions of tidal loughs and fjords, S. latissima can form small forests on hard substratum. The paucity of subtidal research makes it difficult to determine the current and historical scope of these habitats along Ireland's shorelines, but the synergistic threats of ocean warming and increased commercial harvesting present a critical need to better understand the historical and present-day distribution of these important species in order to protect current and future marine forests.
As ecosystem engineers (Jones, Lawton, & Shachak, 1994), kelp provides structure for shallow marine habitats as a resource and a habitat for many organisms (>300 macrofauna species in L. hyperborea forests in Ireland throughout the year (Schoenrock et al., 2020).
In Ireland, observations indicate kelp forests are seasonal homes to commercial species like the edible crab (Cancer pagurus Linnaeus), European lobster (Homarus gammarus Linnaeus), and multiple species of juvenile fish that inhabit the kelp canopy (Schoenrock et al., 2020). There are natural seasonal patterns in these marine communities, but certain species inhabit kelp forests throughout the year including echinoderms, such as the common star (<3 cm in diameter, Asterias rubens Linnaeus), spiny sea star (Marthasterias glacialis Jullien), and the common urchin (Echinus esculentus Linnaeus); a full summary of species is provided in Table 1, (Schoenrock et al., 2020). The constant presence of these taxa in kelp forests over two years of monitoring suggests that they are indicators of healthy ecosystems in the west of Ireland. Regions to the north, south, and east have similar communities (Schoenrock et al., 2020), but these regions have not been as thoroughly surveyed as the west of Ireland.
Kelp forest decline has been observed world-wide driven by warming oceans and heatwave events, anthropogenic inputs Note: A lack of these species within a L. hyperborea forest may indicate (a) the ecosystem is unhealthy or (b) the habitat is small ("kelp park" see Parr, 2020) or comprised of mixed kelps (see Table 5 where communities change with kelp species).
Harvesting is also a threat to kelp forests across Europe, particularly France and Norway where the commercial exploitation of L. hyperborea and L. digitata has been occurring for decades (Valero, Engel, Billot, Kloareg, & Destombe, 2001). The mechanical removal of L. hyperborea results in the removal of whole individuals. Recovery can take greater than 5 years (Lorentsen, Sjøtun, & Grémillet, 2010) in these systems. We show that regular and systematic monitoring is urgently needed in order to conserve and inform policy makers to foster resilience, which we define as the ability of this ecosystem to recover from a disturbance and maintain ecosystem function.

| L AMINARIA HYPERBORE A RECORDS IN IREL AND
In November 2019, the written records of phycologists and old texts were accessed in the Linnaean Society of London to investigate the study of L. hyperborea from 1700 to present day. National herbaria were visited to examine kelp voucher specimens from Ireland over the same time period, though we must note that there are, in general, very few records of large brown algae in these herbaria: National University of Ireland Galway (M. Guiry, 1 record, Finavarra County Clare), The National Botanic Gardens of Ireland (2 records, Clifden County Galway), Trinity College (0 records), and Natural History Museum of London (3 records, Clare Island). The lack of voucher specimens is likely due to the difficulty in preserving thick thalli on paper, and many specimens were likely placed in formalin rather than pressed for preservation (Tsuda & Abbott, 1985). Precise collection details were not noted on many herbarium sheets making it difficult to ascertain the location of collection; therefore, herbarium data were not included when mapping kelp records along the coastline through time. Site records (coordinates) for L. hyperborea were downloaded or donated from the Global Biodiversity Information F I G U R E 1 Laminaria hyperborea "period of first record" for Ireland from pre-1950 (1913 was the only record), from 1950-1969, 1970-1989, 1990-1999, 2000-2009, and 2010 to 2018 (most recent year of record on data platforms) Facility (10 January 2020, www.gbif.org), Ocean Biogeographic Information System (10 January 2020, www.obis.org), National Biodiversity Data Centre (19 September 2019, www.biodi versi tyire land.ie), and the Environmental Protection Agency (17 September 2019). All data were concatenated and quality-filtered for duplicate records (i.e., same coordinates, same date) and correct geographical location (i.e., points on land were removed). Data were then sorted by year of recording and number of years recorded to highlight sites accessed earliest within the country ("period of first record"; Figure 1) and of great interest ("number of years recorded"; Figure 2).
Overall, recording effort increased as we approach the present day, with a boom in the 1990s, however, few sites were recorded multiple times (Table 2).
Records for kelp forest sites were provided through recreational Comhairle Fo-Thuinn (CFT) dive clubs throughout Ireland, the BIOMAR data set (Picton & Morrow, 2006), and recent Irish Research Council and Environmental Protection Agency projects (Schoenrock et al., 2020). These were analyzed separately to highlight the distribution of kelp ecosystems (not just individual kelp  Figure 1); however, there is some overlap. We hope that future observations of L. hyperborea will include more metadata, such as whether the kelp was found in situ, and whether it was in a kelp forest, park, other habitat, or on a strandline (as beach wrack).
The difference between a kelp forest or park is generally described as a reduced density of large kelp individuals, generally adjacent to a kelp forest, although there is no ecological distinction to date (Parr, 2020).

| HIS TORI C AL RECORDS , 170 0 -190 0
During the 1700s, many natural historians began extensive descriptions of what were termed "the Algae" from the initial growth to fructification (formation of reproductive structures) in intertidal species (Greville, 1830 and authors listed within). Some of these researchers focused on seaweeds of the British Isles (present-day Éire and the United Kingdom), and many were women and clergymen (e.g., Reverend David Landsborough of Glasgow, 1847) who had a keen interest in the natural world (see Table 3). This work may have been driven by the need to understand marine harvest because "kelp," potash of all seaweeds, was an economic resource used in agriculture, as packaging material, and an iodine source (Harvey, 1849), and this is certainly a driver for renewed interest in seaweeds from the 1940s (Ara Mara; South & Titley, 1986) to today (Monagail & Morrison, 2020). Cultural uses of kelp were also observed: One herbarium record from The National Botanic Gardens of Ireland is a rosary made from the stipes of L. digitata collected from Glencolumbkille, County Donegal ( Figure 4). Historical records of Irish or Irish-based phycologists is very thoroughly outlined in Guiry (2012); but unfortunately, the information provided at this time does not define the distribution of any recorded seaweeds, but instead refers to ecological aspects like zonation on the shoreline (e.g., the taxon now termed S. latissima was thought to only live between high and low tides). Laminaria spp. and other seaweeds were often described as tangles (Harvey, 1849), and herbarium records  (Guiry & Guiry, 2020).
Interesting notes on the ecology of seaweeds include seaweed support of food webs and community structure: "The Algae, therefore by supporting the base, support the structure" (Harvey, 1849), which is potentially the first description of seaweeds as ecosystem engineers and/or providing ecosystem services. Observations were also noted on ecological interactions between coralline algae and fleshy algae (Lamouroux, 1826), the annual phenology (annual growth patterns) of marine algae, variation in zonation from subtidal to intertidal ("land flora") across regions, and the distribution of dominant brown algae (Cocks, 1859;Harvey, 1849). Most of these studies were restricted to coastlines where seaweed could be easily observed. Cocks (1859) even notes his thought that there would be little space for seaweed below the tidelines (i.e., subtidal). There could be a good deal of data on subtidal marine algae in sounding records of the British Admiralty dating back to 1,580 which would have refuted this idea; however, these records were and are not easily accessible. Up to the twentieth century, much of phycology in Ireland was focused on species descriptions and distributions to the extent of providing "presence" data on certain portions of the coastlines (broadly noted as "Northern Ireland" or "County Cork" for example) which is still an issue with some reports today (see Scally, Pfieffer, & Hewitt, 2020).

| HIS TORI C AL RECORDS , 190 0 -2 018
Some of the first comprehensive surveys of natural environments occurred on Lambay and Clare Island in the early 1900s, and these multidisciplinary reports provide a comparison of intertidal algal communities between 1910 to the 1990s (Cotton, 1909(Cotton, , 1912, but more recent surveys have collected subtidally, expanding the flora record for these locations (Rindi & Guiry, 2004 (Kain, 1963), reproduction (Kain & Jones, 1964), competition and growth (Creed, Kain, & Norton, 1998;Kain, 1962Kain, , 1969Kain, , 1976aKain, , 1977, and description of succession and subcanopy/understory seaweeds (Kain, 1976b(Kain, , 1982(Kain, , 1989 The first distribution record with multiple georeferenced data points of large seaweeds in the UK and Ireland was published by Crisp and Southward in 1958, as a side note to their record of intertidal invertebrates (Crisp & Southward, 1958). From 1950 to 1990, multiple studies referenced seaweeds in specific regions (see Table 3); for instance, Morton (1994)  BIOMAR data are unique in the fact that they can be analyzed to highlight the impact that kelp species and region have on faunal assemblages within kelp ecosystems (Tables 4 and 5). SACFOR scales were given a numerical value (0 = absent, 1 = rare, 2 = occasional, 4 = frequent, 5 = common, 6 = abundant, and 7 = super abundant) for each site record, and a Bray-Curtis similarity matrix was created with species data across sites, and finally, similarity of species compositions within kelp forests, (a) within the same geographical region (Table 4) and (b) within forests dominated by different kelp species (Table 5), were evaluated using an analysis of similarity (ANOSIM; Clarke & Gorley, 2006).
Regional differences were apparent in kelp communities; for example, more species contribute to community similarity in kelp forests in west Ireland than in other regions ( Note: People are listed with known names or initials, dates of activity specifying life span, date of only phycology publication or first year of publication if still active (noted as "From…"), and specific information about the phycologist or naturalist (publication information or occupation).

TA B L E 3 (Continued)
F I G U R E 4 Image of a kelp rosary within the herbarium at The National Botanic Gardens of Ireland. The "beads" of the rosary are likely made from made from the stipes of Laminaria digitata collected from Glencolumbkille, County Donegal also affected community assemblages, but too few replicates exist in mixed and A. esculenta forests to define species driving differences (Table 5). When compared with a recent study in the west of Ireland (Table 1), species associated with L. hyperborea forests are notably different (Table 5), potentially due to the quantitative versus qualitative data collection methodology, and survey focus. For instance, kelp blades where many hydroids reside (e.g., Electra pilosa, Tables 4 and 5) were not included in the swath surveys used for community analysis in Schoenrock et al. (in review). Moving forward, creating a standard monitoring methodology would benefit analysis of data and highlight [changing] patterns in species distribution and habitat usage over time. U. pinnatifida, are found in areas with high average temperatures (but also restricted to man-made or modified structures, e.g., harbors), while the native A. esculenta is found in regions with colder average temperatures (Yesson et al., 2015a) and is thought to be more susceptible to temperature than the Laminaria spp. of the region (Müller, Laepple, Bartsch, & Wiencke, 2009 major freshwater sources (see figure 5 in Yesson et al., 2015a), though this potentially overestimates its distribution along the east coast because the coastline has more sand/mud/marsh habitats than rocky coastline (Neilson & Costello, 1999). More interestingly, the study also indicates regions suitable for species range expansions includ-   (Estes & Duggins, 1995;Hagen, 1995;Ling et al., 2015). The blades of L. hyperborea annually regenerate, starting growth in winter and reaching maximum length mid-summer, and producing sori from October to March (Kain & Jones, 1964). The zoospores produced within sori disperse ~200 m and settle to develop into gametophytes and following fertilization, juvenile sporophytes (Fredriksen, Sjøtun, Lein, & Rueness, 1995). This life cycle may facilitate resilience of kelp populations, allowing for refuge from environmental and biological stressors as either (a) a large sporophyte is too large for grazers or (b) a microscopic stage is safe from storms or otherwise that would uproot large sporophytes (i.e., bet-hedging: Lubchenco & Cubit, 1980). However, our understanding of the role of kelp gametophytes as a spore bank is limited to only a handful of studies (e.g., Robuchon, Couceiro, et al., 2014).

Recovery of kelp ecosystems after large disturbances is an im
Resilience may also be conferred through genetic diversity as genetic variation is the essential evolutionary mechanism with which species can respond to environmental stochasticity. Larger, outcrossed populations tend to be more genetically diverse, than smaller, often inbred, populations. Studying these patterns in the sea can be challenging as not all predictions from terrestrial environments necessarily apply (i.e., chaotic genetic patchiness: Galindo,  and Portugal), Australia, Chile, and California. Interestingly, species that had population genetic data (Macrocystis pyrifera, Lessonia nigrescens, and L. digitata) harbored the highest levels of diversity in areas with strong harvesting pressure. Population connectivity (with kelp species) is largely affected by habitat discontinuity (e.g., Billot, Engel, Rousvoal, Kloareg, & Valero, 2003), and patterns of isolation by distance are common (e.g., Robuchon, Le Gall, Mauger, & Valero, 2014). Understanding how genetic diversity is partitioned and how populations are connected to one another is a necessity in order to determine how populations could recover from harvesting (Robuchon, Le Gall, et al., 2014) or from disturbances, such as heatwaves seen within the Pacific Ocean (Wernberg et al., 2019).

| CON CLUS IONS
Studies of kelp forests in Ireland are historically rare and contain mostly qualitative information. Kelp records with georeferenced data points date back to 1913 and continued over the decades, with a pulse in records from the 1990s onward. Most records are single sightings of L. hyperborea, indicating that either people do not record multiple sightings of the same kelp forest, or many regions are not revisited. Recording effort should move toward documenting kelp ecosystems (presence of a forest) as well as abundance of "indicator species" within using standardized methodology. This would boost evidence that kelp forests are indicators of good environmental status and could be used operationalize MSFD legislation. Maintaining resilience of kelp forests and their associated species is important not only for the ecosystems, but the services they provide to civilization, which can be achieved through monitoring habitats and management of stressors (Krumhansl et al., 2016). Development of a remote sensing mapping tool (via satellite or otherwise) would aid in monitoring the distribution kelp forest distributions.

ACK N OWLED G M ENTS
We appreciate the help of the staff at National herbaria, especially Schoenrock-Rossiter.

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data used within this manuscript are freely available in National