Saint Helena Island StrathE2E Implementation files

Saint Helena Island Implementation STRATHE 2 E Júlia P Olher , Jack H Laverick , Michael Heath , Douglas C Speirs , Maria A Gasalla June , 2025 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 I ntroduction M odel D om a in F ixed P hysic a l Ba ckground M odel a re a proportions Da ily disturb a nce r a tes S ediment porosity H ydr a ulic conductivity S ediment org a nic nitrogen content F ixed biologic a l C onfigur a tion p a r a meters E vent timing p a r a meters ( not fitted ) E cologic a l drivers F ishing fleet Ba ckground S t a tic ge a r - definition p a r a meters in the fishing fleet model Da t a processing to derive fishing p a r a meters Ta rget d a t a La ndings a nd byc a tch P roduction a nd PB r a tios D iet proportions A mbient a l t a rget Ha rverting r a tio M odel fitting E nd A cknowledgements A ppendix R eferences 3 Introduction This document describes the configuration of StrathE2E for Saint Helena Island and its parameterisation to enable stationary state fitting for both the baseline period (2010-2019) and future projections (20202069). These represent contrasting periods of environmental conditions. Volumetric and seabed habitat data define the physical configuration of the system. We regard these as being fixed in time. Similarly, we regard the physiological parameters of the ecology model as being fixed in time. Some of these are set from external data. The remainder are fitted, as detailed here. Changes in the model performance between the different time periods therefore stem from the hydrodynamic, hydro-chemical and fishery driving data. These are detailed in the ecological drivers and fishing fleet sections. In the StrathE2E model all marine lifeforms are explicitly or implicitly accounted for, but aggregated into coarse groups or ‘guilds’ defined mainly by feeding characteristics and diet preferences (Figure 1). All state variables, except macrophytes, are expressed solely in terms of nitrogen mass. For more information about the StrathE2E model check the documentations available. 4| Introduction Figure 1: Ecological guilds of the StrathE2E model (Heath, 2021). Department of Mathematics and Statistics, University of Strathclyde, Glasgow, UK, E-mail: [email protected] Oceanographic Institute, University of Sao Paulo, Brazil, E-mail: juliapetr[email protected] The code written to support this parameterisation is available on github. 5 Model Domain Saint Helena (15.96°S, 5.70°W) is a remote volcanic island with its top rising approximately 800 meters above sea level (Carleton et al. 2010). It is located off the South Mid-Atlantic Ridge and forms part of the British Overseas Territory of Saint Helena, Ascension, and Tristan da Cunha. The island has a 200-nautical-mile Exclusive Economic Zone (EEZ), which is governed and utilized by the St Helena Government and the local population. In 2016, Saint Helena’s entire EEZ was designated as a IUCN Category VI sustainable-use Marine Protected Area (MPA) by the Environmental Protection Ordinance (SHG 2016). Saint Helena does not have reef-building corals. Instead, its narrow inshore habitats consist of boulder and bedrock reefs, as well as sandy beaches (SHG 2023). The majority of the EEZ comprises pelagic habitats and deep ocean waters. The model splits the domain into three zones, inshore/shallow, offshore/shallow, and offshore/deep (Figure 2). The inshore/shallow zone covers waters shallower than 60 m or within 4.5 km from shore. The offshore zone encompasses the remaining area of the model domain and is further divided into a shallow and a deep layer. The 6| Model Domain shallow layer represents water from the surface to 60 m depth, and shares a boundary with the inshore shallow zone. The offshore/deep zone covers the same area as the offshore/shallow zone, but represents water between 60 m and 600 m deep. There is a second internal boundary between the two offshore zones. Figure 2: The spatial structure of StrathE2E; Ocean volumes and seafloor habitats. StrathE2E is built around a simplified spatial structure which represents shelf seas. These spatial units are connected to each other and to boundaries as shown to the right. The volumes connected to each spatial component are highlighted in blue. The seafloor within the StrathE2E model can be classified into eight habitat types. These include three sediment classes: fine (muddy, 1), medium (sandy, 2), and coarse (gravel, 3), with the fourth class (rock, 0) indicating the absence of soft sediment. These sediment classes are defined in both the inshore/shallow and offshore/deep zones. As of V.4, StrathE2E2 can represent an offshore “overhang” where open ocean does not contact the seafloor (Figure 2). The perimeter of the offshore zone in the model domain is the edge of the MPA, with a “false bottom” (overhang) that exchanges with the deep sea. By using the MPA to define the model domain, we accurately represent the protected area status of Saint Helena Island, which also corresponds to its EEZ. The sea surface area of the model domain was estimated to be 450,906.82 km². 7 Figure 3: Map of the model domain. StrathE2E defines seabed sediment habitats as inshore (light gray) or offshore/overhang (dark gray). Within each zone, three sediment classes can be represented – fine (muddy, 1), medium (sandy, 2) and coarse (gravel, 3). A fourth class (rock, 0) represents an absence of soft sediment. In this model implementation the inshore and offshore were divided into 25% sand, 25% gravel and 50% rock sediment, where most of the model domain (>99.7%) is overhang with a false seabed. The sea surface area of the model was estimated to be 450,906.82 km². 8| Fixed Physical Fixed Physical Background Water column inshore / shallow and offshore / deep zone area proportions and layer thicknesses ; seabed habitat area proportions and sediment properties : The depth boundary between deep and shallow layers was determined using vertical diffusivity values from NEMO-MEDUSA (Yool, Popova, and Anderson 2013) and mixed layer depth from Globcolour (GlobColour 2023). The shallow-deep layer division was set at 60 m, with a bottom depth of 600 m for the deep layer. The offshore zone then extends to the edge of the EEZ, but with a “false bottom” that exchanges with the deep sea (without seafloor). In the coastal zone, the polygon is defined for areas shallower than 60 m or within 4.5 km of the coast. The sediment classes were roughly designated as sand, gravel, and rock based on the seabed habitats from the habitat map denoting seabed type in the Marine Management Plan (SHG 2023). 9 Parameters for relationship between median grain size , sediment porosity and permeability . Permeability is used as the basis for estimating hydraulic conductivity which is a parameter in the representation of sediment processes in the model : Porosity (proportion by volume of interstitial water) and permeability of each sediment habitat were derived from median grain sizes using empirically-based relationships. D50 = median grain size (mm); parameters p1 = -1.227, p2 = -0.270, p3 = -0.436, p4 = 0.366 (Heath et al. 2021) where D50* = 0.11 ≤ D50 ≤ 0.50 p5 = -9.213, p6 = 4.615 (Heath, Wilson, and Speirs 2015). These relationships are coded into the StrathE2E2 R-package with the parameters in the csv setup file for the North Sea model. The parameters are probably a reasonable starting point for any future model of a new region. Derivation of the parameters is described in the following text sub-sections. Parameters for in - built relationship between sediment mud content , and slowly degrading ( refractory ) organic nitrogen content of seabed sediments ( see description in this document ): Values for each sediment type derived from parameterised relationships between total organic nitrogen content of sediments (TON%, percent by weight), mud content (mud%, percent by weight) and median grain size (D50, mm). p7 = 0.657, p8 = -0.800 p9 = -1.965, p10 = 0.590 Proportion of TON estimated to be refractory = 0.9 log 10( porosity ) = p 3+ p 4 ⎛ ⎝ ⎞ ⎠ 1 1 + e ( ) − log 10( D 50)− p 1 p 2 permeability = 10 p 5∙ D ∗ p 6 50 mud % = 10 p 7∙ Dp 8 50 TON % = 10 p 9∙ mud % p 10 16 | Fixed biological Planktivorous We found no specific data on the reproductive period of planktivorous species on Saint Helena, except for the report that chub mackerel ( Scomber japonicus ) spawns there in March (Edwards 1990). Therefore, we inferred a spawning period during the austral spring and summer and recruitment in autumn and winter, also based on data from small pelagics at the Saint Peter and Saint Paul Rocks (SPSP, see more details in https://www.marineresourcemodelling.maths.strath.ac .uk/strathe2e/articles/Implementations.html). This was primarily derived from flyingfish ( Cypselurus cyanopterus ). In SPSP, spawning occurs from December to March (Lessa et al. 1999; Pinheiro et al. 2011), with recruitment occurring from April to June (Lessa, Nóbrega, and Junior 2004). Based on this, the planktivorous guild’s spawning period was set from December 1 to March 31st, and recruitment from April 1st to June 30th. Demersal The reproductive period for the demersal guild was determined based on the spawning patterns of the rock hind grouper ( Epinephelus adscensionis ), a species of ecological and fisheries importance in Saint Helena Island. Most species within the grouper subfamily Epinephelinae form spawning aggregations and spawn over extended periods during the winter and spring (Edwards 1990). On Saint Helena and its closest 17 neighbor, Ascension Island, the spawning season occurs from June to November, peaking in August, coinciding with the austral winter (Edwards 1990; AIG 2016; Nolan et al. 2017; Riley, Ball, and Cowburn 2020). Fishers in Saint Helena also report that grouper are difficult to catch in winter months, another indicator of the spawning period (Riley, Ball, and Cowburn 2020). In the absence of specific recruitment data, we set the recruitment period to start 90 days after the end of spawning and continue for 150 days. Carnivorous and scavengers benthos Reports from the Saint Helena Government indicate that lobsters enter early reproductive stages from October 1st to December 31st, with a closed fishing season from January to the end of March due to the advanced reproductive stages of the brown spiny lobster ( Panulirus echinatus ) and stump lobster ( Scyllarides obtusus ) (SHG 2022). Based on this information, the spawning period for the carnivorous and scavenging benthos guild was set from January 1st to March 31st, with the recruitment period from April 1st to June 30th. In the absence of specific recruitment data, we assumed recruitment to begin 30 days after the spawning period ends, lasting for 120 days. Filter and deposit benthos Research on the reproductive activities of filter and deposit feeding benthos in the Saint Helena is limited. We adopted the same spawning and recruitment periods used for the FD benthos guild in the South Brazil Bight implementation (see more details at https://www .marineresourcemodelling.maths.strath.ac.uk/strathe2e/articles /Implementation_The_Brazilian_Shelf.html) - spawning period from September 1st to February 1st and recruitment from August 1st to January 1st. Table 4: Biological event timing parameters, constant across the time periods. The data are processed in the model setup to calculate the immigration flux parameters needed in the ecology model. Spawning and recruitment durations were established assuming a month with 30 days. 18 | Fixed biological P a r a m e t e r V a lu e P e l agic fi s h sp a wn i n g st a rt da y 330 P e l agic fi s h sp a wn i n g d ur a t i on ( da ys ) 120 P e l agic fi s h r ec ru i tm e nt st a rt da y 91 P e l agic fi s h r ec ru i tm e nt d ur a t i on ( da ys ) 90 D e m e rs a l fi s h sp a wn i n g st a rt da y 150 D e m e rs a l fi s h sp a wn i n g d ur a t i on ( da ys ) 180 D e m e rs a l fi s h r ec ru i tm e nt st a rt da y 240 D e m e rs a l fi s h r ec ru i tm e nt d ur a t i on ( da ys ) 150 F i lt / de p be nt h os sp a wn i n g st a rt da y 240 F i lt / de p be nt h os sp a wn i n g d ur a t i on ( da ys ) 120 F i lt / de p be nt h os r ec ru i tm e nt st a rt da y 210 F i lt / de p be nt h os r ec ru i tm e nt d ur a t i on ( da ys ) 150 C a rn / s ca v be nt h os sp a wn i n g st a rt da y 1 C a rn / s ca v be nt h os sp a wn i n g d ur a t i on ( da ys ) 90 C a rn / s ca v be nt h os r ec ru i tm e nt st a rt da y 120 C a rn / s ca v be nt h os r ec ru i tm e nt d ur a t i on ( da ys ) 120 Extra - domain stock biomass of migratory , and the proportion invading the domain each year . Start and end dates for the annual invasion , and start and end dates for the emigration . ( see description below ). The migratory fishes in the region are part of a broader Atlantic stock and perform transatlantic movements, utilizing multiple spawning grounds. Key migratory species making seasonal transits through Saint Helena waters include skipjack tuna ( Katsuwonus pelamis ), bigeye tuna ( Thunnus obesus ), and yellowfin tuna ( Thunnus albacares ), along with other tuna-like species and sharks. The data on the biomass of these migratory fish stocks, the proportion entering the Saint Helena EEZ, and the timing of their migrations is detailed below. 19 We assumed that there is no feedback between fishing and environmental conditions and the biomass and migration patterns. In this version of StrathE2E the timing of immigration and emigration, and the mass influx across the ocean boundary during the annual immigration phase are treated as period-specific external driving data. The model setup code calculates the parameters which are needed in the ecology model. These are the only fixed (i.e. non-fitted) ecology model parameters which are period-specific. Saint Helena likely serves as a feeding ground for sub-adult fish, which migrate to spawning grounds upon reaching maturity (Wright et al. 2020). Bigeye tuna ( Thunnus obesus ) and yellowfin tuna ( Thunnus albacares ) primarily spawn in tropical waters, such as the Gulf of Guinea and the Gulf of Mexico, during the summer (Carleton et al. 2010; FAO 2011). Recent tagging studies on yellowfin tuna show that they remain in Saint Helena waters for extended periods, ranging from six months to a year, during which they feed and grow (Wright et al. 2020). They are typically recruited into the fishery between February and April each year and leave the population upon reaching a length of 120 cm (Wright et al. 2020). Yellowfin tuna reach an average size of 96 to 120 cm at Cardno Seamount in May/June, suggesting migration out of the EEZ at this size (Wright et al. 2020, 2021). We utilized outputs from the SEAPODYM model to characterize the movement dynamics of tunas in the Atlantic Ocean, as part of research conducted within the Mission Atlantic project (Merillet, Titaud, and Conchon 2024). These outputs were based on the same four climate projections derived from NEMO-ERSEM physical outputs used in the StrathE2E driving data (see pages 26-29). The projections are driven by two SSPs (SSP3-7.0 and SSP1-2.6) and two CMIP6 models (CNRM-CM6-1- 20 | Fixed biological HR and GFDL-ESM4). Two species, skipjack tuna ( Katsuwonus pelamis ) and albacore tuna ( Thunnus alalunga ), were modeled using SEAPODYM-MASS, an age-structured dynamic population model that accounts for prey availability (Lehodey, Senina, and Murtugudde 2008; Lehodey, Murtugudde, and Senina 2010). Historical data were provided from 1990 to 2014, along with projections from 2015 to 2070, on a monthly basis and averaged across the Saint Helena model domain. Data included local biomass estimates for each SSP variant, forcing model variant, and total stock biomass in the Atlantic. To establish the baseline period from 2010-2019, historical data from 2010–2014 were combined with projections from 2015–2019. The remaining time series were divided into decadal intervals from 2020– 2029 to 2060–2069 for each of the four climate projections. To identify the start and end days for annual immigration and emigration periods, we interpolated data to obtain daily values averaged by decade for each projection. To synchronize with the StrathE2E annual cycle, each month was standardized to 30 days. Seasonal cycles for each decade and projections were extracted using the decompose function in R, excluding trends and random components - for more information check the function documentation. Immigration and emigration periods were defined as the days when biomass increased from the 20th to 80th percentile and decreased from the 80th to 20th percentile, respectively. Figure 4: Seasonal component of migratory fish biomass derived from SEAPODYM model outputs. Points A to D indicate the start and end dates of immigration and emigration for each decade (2010–2019 to 2060–2069), under two climate scenarios (SSP1-2.6 and SSP3-7.0) and two forcing models (CNRM and GFDL). 21 Discussions with the model’s authors suggested that parameterization favored skipjack biomass in tropical and equatorial regions. To address this, we scaled the local tuna biomass using a factor derived from regional stock assessment. The scaling equation below was applied in all Mission Atlantic case studies to adjusted SEAPODYM outputs to better reflect local stock assessments. Where no local assessments were available, as in the Saint Helena domain, α defaulted to 1. Additionally, another scaling factor ( ) was applied in cases where skipjack and albacore alone do not represent the entire migratory guild, as in the Saint Helena Island domain. The scaled biomass was calculated as: Where: - is the proportion of the Atlantic stock entering the model domain annually, from SEAPODYM; - is the scaling factor to adjust the local biomass relative to known stock assessments; - is the Atlantic stock from SEAPODYM; - is the Atlantic biomass from stock assessments; - is the local biomass from stock assessments; - is the scaling factor to the others Atlantic stocks that may enter the model domain In the case of Saint Helena, was derived from stock assessments from the 2010s. Part of the management of Atlantic migratory fish stocks falls under the International Commission for the Conservation of Atlantic Tunas (ICCAT), which relies on a consensus-based decisionmaking process among its member nations. To estimate the ‘global’ stock biomass of migratory fish assumed to enter the model domain, we compiled data for 18 species, including key species such as the aforementioned skipjack, yellowfin, bigeye, and albacore tuna, along with wahoo ( Acanthocybium solandri ), frigate tuna ( Auxis thazard ), Galapagos shark ( Carcharhinus galapagensis ), common dolphinfish ( Coryphaena hippurus ), Atlantic sailfish ( Istiophorus albicans ), shortfin mako ( Isurus oxyrinchus ), white marlin ( Kajikia albida ), blue marlin ( Makaira nigricans ), blue shark ( Prionace glauca ), Atlantic bonito ( Sarda sarda ), smooth hammerhead ( Sphyrna zygaena ), southern λ α = , λ = ln ( ) x ( t ) X ( t ) ln( P ) Y ( t ) X ( t ) scaled  biomass ( t ) = [ P ] α × X ( t ) × λ P α X ( t ) Y ( t ) x ( t ) λ λ 22 | Fixed biological bluefin tuna ( Thunnus maccoyii ), Atlantic bluefin tuna ( Thunnus thynnus ), and swordfish ( Xiphias gladius ). Biomass data were sourced from Sea Around Us (SAU), which utilizes ICCAT data and the CMSY method - Pauly, Zeller, and Palomares (2023). SAU estimates, available for 2010–2018, were averaged for this analysis. The combined biomass of the 18 key stocks in the Atlantic, which may enter the Saint Helena EEZ, totaled 5,503,844.45 tons. The scaling factor was calculated by dividing the global migratory biomass by SEAPODYM’s skipjack and albacore stock estimates. This factor was then applied to SEAPODYM’s local skipjack and albacore biomass estimates to calculate total migratory biomass averaged by decade, model, and SSP. The annual proportion of Atlantic migratory biomass entering the Saint Helena domain was determined by dividing the scaled local biomass by the scaled Atlantic migratory biomass. An increase in the proportion of migratory fish migrating to Saint Helena suggests that the island will become more suitable for tunas by the 2060s under these climate projections, except in the SSP1-GFDL model variant (Table 5). This trend was integrated into StrathE2E to investigate its potential impacts on the entire food web and fisheries. Figure 5: Scaled annual mean migratory fish biomass from SEAPODYM model outputs for each decade (2010–2019 to 2060–2069), under two climate scenarios (SSP1-2.6 and SSP3-7.0) and two forcing models (CNRM and GFDL). To assess the proportion of peak biomass that remained within the model domain even in the emigration period we extracted the maximum and minimum values from the seasonal cycle. Compared to λ 23 the literature available, yellowfin tuna exhibit high fidelity to the area and remain in St.Helena waters for extended periods due to elevated productivity levels associated with the island and seamounts (Wright et al. 2018, 2021). The full set of parameters used for migratory fish event timing is detailed in lines 17–26 of the event_timing_SAINT_HELENA_*.csv files for each decade under the four climate projections. Table 5: Time-variant biological event timing parameters related to migratory fish for the baseline period (2010-2019) and last decade of analysis (2060-2069) across the four climate projections, used as parameters data. 2010-2019 2060-2069 SSP 1CNRM SSP 3CNRM SSP 1GFDL SSP 3GFDL SSP 1CNRM SSP 3CNRM SSP 1GFDL SSP 3GFDL O cea n bi om a ss 5214177 5278278 5767905 5808957 6503421 6745468 6407583 6766477 P ropn o f o cea n popn e nt e r i n g p e r y ea r 0.01273 0.01299 0.01319 0.01334 0.01206 0.01171 0.01146 0.01158 I mm ig r a t i on st a rt da y 203 207 260 239 71 48 71 66 I mm ig r a t i on e n d da y 346 352 343 332 162 149 187 170 P ropn o f p eak popn w hich r e m ai ns 0.9352 0.9197 0.8931 0.8943 0.8697 0.8988 0.8196 0.8540 E m ig r a t i on st a rt da y 59 65 56 45 233 220 258 241 E m ig r a t i on e n d da y 132 136 189 143 358 335 358 353 Table 6: Fixed biological event timing parameters related to migratory fish, constant across the time periods and climate projections P a r a m e t e r V a lu e M ig r a tory fi s h sw i t ch (0= o ff 1= on ) 1.000 M ig r a tory fi s h ca r b on to w e t w eigh t ( g / g ) 0.184 M o de l d om ai n s ea sur face a r ea ( k m 2) 450906.822 24 | Ecological drivers Ecological drivers Monthly resolution time-varying physical and chemical driving parameters for the model were derived from a variety of sources: Temperature, vertical mixing coefficients, volume fluxes, and boundary nutrient, detritus and phytoplankton concentrations from outputs of an ensemble of NEMO-ERSEM coupled hydrogeochemical model runs with a 2015 historical/future split from Artioli et al. (2023). Surface shortwave radiation from CMIP6 used to force the NEMO-ERSEM coupled hydro-geochemical runs mentioned above. Freshwater volume outflows from CMIP6 used to force the NEMO-ERSEM coupled hydro-geochemical runs mentioned above. River nitrate and ammonia concentrations taken from the IMAGE model to force the NEMO-ERSEM coupled hydrogeochemical runs mentioned above. Atmospheric deposition of nitrate and ammonia from ISIMIP3a (Inter-Sectoral Impact Model Intercomparison Project; Yang and Tian (2022)) with SSP projects from ISIMIP3b used to force the NEMO-ERSEM coupled hydro-geochemical runs mentioned above. Remote sensing data products on Suspended Particulate Matter (Globcolour L3b; ftp://ftp.hermes.acri.fr/GLOB/merged/month/). Wave height, period, and direction from the ERA-5 reanalysis monthly means accessed through CDS for 1980-2021 (Hersbach et al. 2020). Details of how these data were processed are given below, supported by the nemomedusR and MiMeMo.tools packages. NEMO - ERSEM ensemble : Four different NEMO-ERSEM runs were used to parameterise different versions of the Saint Helena Island implementation of StrathE2E. These four runs are a 2x2 factorial cross of two future projected scenarios (SSP1-2.6 and SSP3-7.0) with a historical hindcast from 2015, and forcing 25 by two atmospheric CMIP6 models (CNRM-CM6-1-HR and GFDL-ESM4) produced by Artioli et al. (2023). In the following sections model output was processed for a 2010-2019 baseline period, and then decadal projections from 2020-2029 until 2060-2069. Vertical mixing coefficients between the model layers : Vertical diffusivity from the NEMO-ERSEM coupled hydro-geochemical model output was interpolated for each grid cell at the 60m boundary depth between the shallow and deep layers of the offshore zone, and the 600m boundary at the deep sea overhang. These values were summarised as monthly averages into period-specific climatological annual cycles of data for decadal periods for all combinations of SSPs and forcings. Monthly averaged temperatures for each water column layer : Derived by monthly averaging values at grid points within the inshore and vertical layers of the offshore zones from the NEMO-ERSEM coupled hydro-geochemical model output, weighted by grid point volumes. Values were summarised into period-specific climatological annual cycles of data. Monthly averaged suspended particulate matter ( SPM ) concentrations ( mg . m -3) in the shallow zone and the deep zone upper layer : Monthly averaged values of inorganic suspended particulate matter in sea water are available from the Globcolour project, starting from September 1997. These data are derived from satellite observations using the algorithm of Gohin (Gohin 2011). Data were downloaded from the ftp server (ftp://ftp.hermes.acri.fr/GLOB/merged/month/). We summarised these values as zonal statistics for the model domain to acquire a climatological annual cycle of data for the 2010-2019 simulation period only. Monthly average light attenuation coefficients for the inshore and offshore surface layers : Light attenuation in open water was parameterised from a linear relationship between the light attenuation coefficient and suspended particulate matter concentration (SPM) (Devlin et al. 2008). 32 | Fishing fleet Data processing to derive fishing parameters Landings Data from the Environment and Natural Resources Directorate (ENRD) includes annual landings per species. It is compiled by the Statistics Office of the St.Helena Government from various sources, including the Fisheries Corporation and the Agriculture and Natural Resources Department (https://www.sainthelena.gov.sh/st‐helena/statistics/). This dataset contains information on 20 species distributed across migratory, demersal, planktivorous, and CS benthos guilds. The fleets were allocated to each species based on information from official government reports. When available, landings were also divided into inshore and seamount categories using the proportions provided in the CR087 report. We also utilized data from the ICCAT, which records and stores pelagic fishing data (ICCAT 2023). We filtered this data using the UK-Saint Helena flag in the T2CE (Task 2 – Catch and Effort) database. We verified that the landing data for the period 2010-2019 for pelagic species matched those available in the ENRD database but were divided into two fleets: rod and reel, and longline. We considered the rod and reel fleet equivalent to the Pole and Line fleet in our model and used this division accordingly. The landing data were consistent with the official fisheries statistics of St. Helena, and we applied the same fleet divisions as those present in the ICCAT data, except for swordfish ( Xiphias gladius ), which was not present in the official St. Helena database for this period. For swordfish, we used the landing values from the ICCAT data. 33 Figure 7: Catches per year and per fleet compiled for the Saint Helena EEZ StrathE2E model. Pole and line The average annual landings between 2010 and 2019 were 329.5 tonnes, with a maximum of 877.88 tonnes in 2011 and a minimum of 99.6 tonnes in 2013. All landings targeted the migratory fish guild. Yellowfin tuna contributed nearly 50% of the total landings, followed by skipjack tuna with 26%, and bigeye tuna with 16%. The remaining catch consisted of wahoo, albacore/longfin tuna, marlins, and sharks. Longlines Longline fishing occurred only between 2011 and 2013, resulting in a total catch of 23.71 tonnes. Only species from the migratory fish guild were landed, primarily swordfish and bigeye tuna. Handlines A total of 132.28 tonnes were landed between 2010 and 2019, with a minimum of 2.57 tonnes in 2015 and a maximum of 25.13 tonnes in 2010. The most fished guild was demersal fish, with 81.219 tonnes of grouper and 21.46 tonnes of conger eel. Other species caught included cavalley, dorado, and yellowtail. Handlines and dip - nets A total of 27.58 tonnes of planktivorous fish were caught, with the only species identified in the database being mackerels. 34 | Fishing fleet Pots and Hand - Gathering A total of 0.46 tonnes of the lobsters (Carnivorous and scavengers benthos guild) Panulirus echinatus and Scyllarides obtusus were landed from 2010 to 2019. Discards Guild-specific discard rates (the proportion of the catch discarded) were derived by calculating discard ratios and applying these to the landing values described earlier. We considered that fishery discards in the St Helena fleets are minimal, primarily due to the use of “one-by-one” techniques, which allow fishermen to avoid unwanted species. According to (Carleton et al. 2010), the pelagic fleets in Saint Helena occasionally incidentally catch cavally ( Pseudocaranx dentex ), dolphinfish ( Coryphaena spp. ), blue marlin ( Makaira nigricans ), and pelagic sharks. However, they report that, aside from some sharks, nearly all incidental catches are either sold or retained by fishermen for home consumption. Gilman, Suuronen, and Chaloupka (2017) evaluated discard rates in global tuna, tuna-like, and billfish fisheries by RFMO area and gear type. In the ICCAT area of the South Atlantic, no discard rate records were found for handline and pole and line tuna fisheries, which are also considered to have low or negligible discard rates by Kelleher (2005). Based on this, we considered the discard rates to be zero for both fleets. 35 Regarding the baitfish fishery, small-scale fisheries for small pelagics are considered to have low or negligible discard rates (Kelleher 2005), and there are no non-target species in this fleet in St Helena according to Carleton et al. (2010). We also assumed there were no discards in the pot and hand-gathering fleets for lobster. In contrast, longline fishing trials within Saint Helena’s EEZ have reported significant bycatch levels, including blue sharks and other shark species, in addition to the target swordfish (Collins 2017). The FAO global database (Kelleher 2005) indicates that smaller longliners, which typically have shorter trips, tend to retain more sharks and other nontarget species, resulting in a discard rate of 15% for these vessels. Based on this, we adopted the discard rate of 15% for longlines for the migratory guild. Bycatch We did not find records of bycatch occurrence involving marine mammals or seabirds in the St Helena fleets. Seabirds such as boobies, noddies, and diving petrels do associate with fishing vessels and sometimes take live or cut bait, but no seabird bycatch was observed or reported in any UK Overseas Territories recently (ICCAT 2022a). Rough-tooth dolphins ( Steno bredanensis ) often take bait off hooks and can hinder fishing activities, but they are not adversely affected by fishing operations (Carleton et al. 2010). While the occasional seasonal harvesting of dolphins occurred until the 1970s, there is no evidence of any fishery-related cetaceans deaths in recent decades (Carleton et al. 2010). However, it is anticipated that some seabird bycatch may occur in the longline fleets. We extrapolated seabird bycatch rates from similar fleets from another island in the South Mid-Atlantic Ridge, Saint Peter and Saint Paul Rocks (see more details in https://www .marineresourcemodelling.maths.strath.ac.uk/strathe2e/articles /Implementations.html). We applied a mortality rate of 0.05 birds per day for longline fleets and 0.143 birds per day for handlines (Bugoni et al. 2008). These daily bird mortality rates were multiplied by the estimated fishing days for each fleet (see subsection below for fishing effort). To estimate bycatch in terms of biomass, we used the average weight of these species as reported in the literature. The total bycatch from both fleets was estimated at approximately 9 individual birds, equivalent to around 0.047 tonnes per year (Table 7). This data was converted to mMN.m⁻².y⁻¹ using the appropriate conversion factor with molar nitrogen mass (see detailed nitrogen mass sources per unit wet weight at the North Sea Implementation). 36 | Fishing fleet It is important to note that the methodology developed here is a simplification, designed only to parameterize the Saint Helena EEZ StrathE2E model. The annual bycatch data presented should not be regarded as a comprehensive assessment of bycatch in the region. Table 7: Seabird guild bycatch rates per fleet applied in the baseline period for the Saint Helena Island implementation. Average weight from compilation in target data of booby, noddy and petrels. Bycatch individuals per fishing day estimated using Bugoni et al. (2008) for tropical longlines F l ee t F i s hi n g da ys B y ca t ch ( i n di v id u a ls ) A v e r age s eabi r d w eigh t ( t ) B y ca t ch ( t ) B y ca t ch ( m MN / m ² / y ea r ) L on g l i n e s 9.12 ± 19.84 0.46 ± 0.99 0.0051 0.0023 ± 0.0051 1.30 E -08 ± 2.83 E -08 H a n d l i n e s 60.5 ± 50.02 8.65 ± 7.15 0.0051 0.0443 ± 0.0366 2.47 E -07 ± 2.04 E -07 Fishing effort We utilized data from ICCAT, which records and stores pelagic fishing data, filtering for the UK-Saint Helena flag in the T2CE database, Task 2 – Catch and Effort (ICCAT 2023). We verified that the landing data for pelagic species from 2010-2019 matched those available in the ENRD. We used fishing effort in terms of fishing days for longline and line days (considered equivalent to fishing days) for rod and reel. The “Rod and Reel” gear category in the ICCAT database was renamed to “Pole and Line” to be consistent with the catch data. This resulted in an average of 1885.6 fishing days from 2010-2019 for the Pole and Line fleet targeting tuna and tuna-like species. We assumed 8 hours of fishing per day for conversion to fishing hours. For the longline fleet, we obtained an average of 30.67 fishing days from ICCAT records from 2013-2015 for longline targeting swordfish and tunas (Table 8). We assumed 12 hours of fishing per day for conversion to fishing hours. Table 8: Pole and line effort compiled for the baseline period implementation of Saint Helena EEZ model from ICCAT data, assuming 8 hours of fishing per day. 37 Y ea r F i s hi n g da ys E ff ort ( s . m 2 . da y -1 ) 2010 2162 3.84 E -07 2011 3974 7.05 E -07 2012 2498 4.43 E -07 2013 1196 2.12 E -07 2014 1194 2.12 E -07 2015 1222 2.17 E -07 2016 1076 1.91 E -07 2017 1638 2.91 E -07 2018 1710 3.03 E -07 2019 2176 3.86 E -07 For the handline fleet, we considered the effort targeting primality the grouper present in the Blue Belt Programme report for the Saint Helena EEZ (Riley, Ball, and Cowburn 2020) from 2010 to 2019, obtaining an average of 60.5 fishing days per year. For the baitfish fleet, there are no official records of effort. We made a rough estimate based on information from Collins (2017) that each vessel would normally collect 20-30 kg of baitfish per fishing day. We used 25 kg per fishing day and the annual mackerel landing value from 2010-2019 (2.76 kg) to estimate approximately 110 fishing days per year. There were no available indicators of the effort directed at lobster fishing on Saint Helena. Therefore, we used estimates of approximately 10 kg of spiny lobsters per fishing day using traps in another SouthAtlantic island (Freire et al. 2015). From 2010 to 2019, an average of 46 kg of lobsters were landed per year, roughly suggesting that five fishing days would be necessary annually. The final unit for input into the model for each fleet was sec.m-2.d-1 (Table 9). Table 9: Effort compiled per fleet for the baseline period implementation of Saint Helena EEZ model 38 | Fishing fleet F l ee ts S tr a t h E 2 E S ai nt H e l e n a E ff ort ( s . y ea r -1 ) E ff ort ( s . m 2 . da y -1 ) P ol e a n d l i n e 5.43 E +07 3.34 E -07 L on g l i n e s 3.97 E +05 2.45 E -09 H a n d l i n e s 2.61 E +06 1.61 E -08 H a n d l i n e s a n d di p - n e ts ( bai t ) 7.92 E +05 4.88 E -09 P ots a n d ha n d ga t he r i n g 2.16 E +05 1.33 E -09 Fleet Distribution per habitat The distribution of each fleet across habitats was assessed by integrating literature data with sediment habitat information. The majority of fleets operate offshore within the overhang zone of the model. Exceptions include the artisanal handlines and hand-gathering and pot fishing fleets, which are predominantly concentrated in the inshore zone. Handline fishing for grouper and pot fishing for lobsters are confined to the inshore rock and gravel habitats area and are not present around the Bonaparte and Cardno seamounts (Collins 2017; Rees et al. 2016; Riley, Ball, and Cowburn 2020). Baitfish ( Decapterus spp. ) are caught with nets at the surface in waters 10 to 20 meters deep (Carleton et al. 2010), usually near sandy and rocky inshore beaches (Edwards 1990). We allocated the fishing effort per habitat as outlined in Table 10. Table 10: Spatial distribution of fishing activity across seabed habitat classes, inferred for Saint Helena based on the characteristics of local fishing fleets. The ‘deep ocean’ habitat refers to the portion of the offshore zone that extends into deep ocean waters, without a seafloor. 39 F l ee ts S tr a t h E 2 E S ai nt H e l e n a S ha llow ro ck S ha llow s a n d S ha llow g r a v e l D ee p o cea n P ol e a n d l i n e 0.00 0.0 0.00 1 L on g l i n e s 0.00 0.0 0.00 1 H a n d l i n e s 0.50 0.0 0.50 0 H a n d l i n e s a n d di p - n e ts ( bai t ) 0.25 0.5 0.25 0 P ots a n d ha n d ga t he r i n g 0.50 0.0 0.50 0 Seabed abrasion We assumed no seabed abrasion rate (plough rate) per unit of time, as the fleets are mainly pelagic. The only fleet with contact with the seafloor is pots and traps, however the plough rate is minimum or no existent. Other processing 1. Total catch by fleet and guild was calculated by combining landings with discard/bycatch values. 2. When catch was 0 discard rates were set to 1 3. Fishing power was calculated as catch/activity per fleet. 4. Demersal non quota and quota limited were combined into a single Demersal guild for catch, landings and discards. 40 | Target data Target data The observed (target) data and its standard deviation (SD) are used in the fitting process to train StrathE2E, ensuring it returns reasonable values during the simulated annealing scheme. For more information about the the methodology implemented in StrathE2E check the documentations available. The ecosystem state indices included in the optimization process comprise data on the gross and net production of commercial guilds, production-to-biomass (PB) ratios, dietary proportions of top predators, fishery landings, bycatch, primary production, chlorophyll, and other relevant properties, all located in the file “annual_observed_SAINT_HELENA_MA.csv.” The landings and bycatch data compiled in the Fishing fleet section (pages 34-37) were used as target data, expressed in units of mMN/m²/year. The SD for landings was calculated based on the 2010– 2019 time series, while for bycatch, the SD was assumed to be 25% of the corresponding values. To optimize fishing gear activity rates we used target data concerning the harvest ratios (the proportion of exploitable biomass captured per day within each guild), fitting a scaling parameter for the effort-toharvest ratio. The resultant values can be found in the files “region_harvest_SAINT_HELENA_MA.csv” and “zonal_harvest_r_SAINT_HELENA_MA.csv”. Landings and bycatch The landings and bycatch data compiled in the Fishing Fleet Section were also used as target data. Annual landings from all fleets were summed for each guild and averaged over the 2010-2019 period. Data in tonnes were converted to units of mMN/m²/year, with molar nitrogen mass (see detailed nitrogen mass sources per unit wet weight at the North Sea Implementation). The standard deviation for landings was calculated based on the 2010–2019 time series, while for bycatch, the standard deviation was roughly assumed as 25% of the corresponding values. 41 Table 11: Landings and bycatch data from the Saint Helena relevant to the period 2010-2019, or a general value when specific period data was unavailable. The standard deviation for fishery landings was determined through a detailed analysis of multi-year data. For bycatch, the standard deviation was a rough estimate in order to assign a weight to a specific measure in the likelihood calculation. A nnu a l m ea sur e SD o f m ea sur e U n i ts D e s c r i pt i on 1.2500 e -05 1.0300 e -05 m MN / m 2/ y A nnu a l pl a n k t i vorous fi s h l a n di n g s ( l i v e w eigh t ) 3.0600 e -05 2.0100 e -05 m MN / m 2/ y A nnu a l de m e rs a l fi s h l a n di n g s ( l i v e w eigh t ) 0.00172 0.00107 m MN / m 2/ y A nnu a l m ig r a tory fi s h l a n di n g s ( l i v e w eigh t ) NA NA m MN / m 2/ y A nnu a l susp / de p be nt h os l a n di n g s ( l i v e w eigh t ) 1.0300 e -07 1.3100 e -07 m MN / m 2/ y A nnu a l ca rn / s ca v be nt h os l a n di n g s ( l i v e w eigh t ) NA NA m MN / m 2/ y A nnu a l ca rn zoopl a n k ton l a n di n g s ( l i v e w eigh t ) 2.6000 e -07 2.3300 e -07 m MN / m 2/ y A nnu a l b y ca t ch o f bi r d s Production and PB ratios Calculating the production involved compiling and adjusting biomass values for the model area and converting them to molar nitrogen mass (see detailed nitrogen mass sources per unit wet weight at the North Sea Implementation). Subsequently, convert biomass into gross production using PB ratios. For seabirds, cetaceans, and pinnipeds, net production was calculated using the relationship net production = 0.6 * gross production. We compiled biomass data and PB ratios for the main species of each guild, including both commercial species present in our fisheries database and non-commercial species of ecological importance in the region. In our literature search, we encountered a limited number of recent studies quantifying biomass in the Saint Helena EEZ. Details regarding the sources of biomass and PB ratios, along with the calculated values, are available in Table 12. The compilation conducted for each guild is described in the subsequent sections. Table 12: PB ratios utilized to estimate gross (demersal fish) and net (seabirds and cetaceans) production as target data for each commercial guild in the model. Production values are presented in 48 | Target data oceanic shrimps play a crucial role in the diets of tunas (Laptikhovsky et al. 2021) and are likely abundant in the region. However, we found no biomass data for significant populations of carnivorous, scavenging, filter-feeding, or deposit-feeding benthic species. As a result, production was not used as a target parameter for either benthos guild in the model. Zooplankton Regarding the zooplankton guilds, we found limited local information, including details on their proportion in the tuna diet (Laptikhovsky et al. 2021) and descriptions of cephalopod species (SHG 2023). However, to the best of our knowledge, biomass estimates are not available for the St. Helena EEZ. As a result, zooplankton production was not used as target data for the omnivorous and carnivorous zooplankton guilds. Diet proportions We reviewed the dietary preferences of the main seabirds species occurring in Saint Helena Island. The dietary proportions identified were subsequently utilized as target data for the fitting process. We then calibrated the resource-consumer matrix of the model using a simulated annealing scheme to align the 2010-2019 model with observed ecosystem state data. Although we did not include the dietary proportions of cetaceans due to the absence of specific local studies, the species occurring around Saint Helena are known to consume prey in other regions that are already represented in the base model preference matrix. 49 Seabirds Saint Helena is home to an internationally significant population of redbilled tropicbirds ( Phaethon aethereus ) (Beard et al. 2023). A detailed study by (Beard et al. 2024) conducted from 2013 to 2018 revealed that squid comprised 73.3% of the diet, while fish constituted 26.6%. Juvenile squid from the Ommastrephidae family, especially neon flying squid ( Ommastrephes cylindraceus ), were found in 51.1% of the samples. The study highlighted a high proportion of squid in the diet of red-billed tropicbirds in Saint Helena, a pattern that contrasts with other regions where these birds primarily feed on fish. The authors note that it remains unclear whether this preference for squid is inherent or reflects recent changes in prey availability due to fisheries or climate change. The planktivorous fish guild accounting for 26.6% of the seabirds’ diet and other guilds (i.g. squids in CS zooplankton) for the remaining proportions were integrated into the target data in the Saint Helena model (Table 13). Table 13: Proportions of the fish guilds in the diet of seabirds considered as target data in the Saint Helena Island model. P roport i on P l a n k t i vorous D e m e rs a l M ig r a tory D i s ca r d s S our ce s S eabi r d s 0.266 0.000 0.000 0.000 B ea r d e t a l . (2024) 50 | Target data Ambiental target The annual total primary production for Saint Helena averaged over the 2010–2019 period was obtained from satellite observations in the Global Ocean Colour (GlobColour 2023) with spatial resolution of 4 × 4 km. The inshore-to-offshore ratio of annual mean phytoplankton concentration in the surface layer was also derived from these observations. Chlorophyll concentrations from remote sensing satellite data were converted to nitrogen units using a carbon-to-chlorophyll weight ratio of 20 and the Redfield molar ratio of carbon to nitrogen. Surface chlorophyll data from GlobColour (2023) were also used to generate monthly concentration values for the “monthly_observed_SAINT_HELENA_2010-2019.csv” file. Nitrate data were obtained from the World Ocean Atlas 2023 (WOA23; Reagan et al. (2023)) and categorized into shallow (≤60 m) and deep (up to 600 m) layers, as well as two seasonal periods: May–August and November–February. A depth-resolved, 1° × 1° gridded monthly climatology of nitrate was used for this analysis. Table 14: Annual ambiental data from the Saint Helena relevant to the period 2010-2019, or a general value when specific period data was unavailable. Sources: World Ocean Atlas (Reagan et al, 2024), NEMOERSEM and GlobColour, Copernicus. 51 A nnu a l m ea sur e SD o f m ea sur e U n i ts D e s c r i pt i on 862.81158 50.00000 m MN / m 2/ y A nnu a l tot a l pr i m a ry pro d u c t i on 0.07402 0.03000 m g / m 3 A v e r age w i nt e r ( N ov - F eb ) ch lorop h yll c on c s ha llow l a y e r 0.09887 0.05000 m g / m 3 A v e r age summ e r ( M a y - A u g ) ch lorop h yll c on c s ha llow l a y e r 0.66133 0.58195 m MN / m 3 A v e r age w i nt e r ( N ov - F eb ) n i tr a t e c on c s ha llow l a y e r 0.31693 0.09501 m MN / m 3 A v e r age summ e r ( M a y - A u g ) n i tr a t e c on c s ha llow l a y e r 18.41458 0.67838 m MN / m 3 A v e r age w i nt e r ( N ov - F eb ) n i tr a t e c on c dee p l a y e r 18.09236 0.27485 m MN / m 3 A v e r age summ e r ( M a y - A u g ) n i tr a t e c on c dee p l a y e r NA NA m MN / m 3 A v e r age w i nt e r ( N ov - F eb ) a mmon ia c on c s ha llow l a y e r NA NA m MN / m 3 A v e r age summ e r ( M a y - A u g ) a mmon ia c on c s ha llow l a y e r NA NA m MN / m 3 A v e r age w i nt e r ( N ov - F eb ) a mmon ia c on c dee p l a y e r NA NA m MN / m 3 A v e r age summ e r ( M a y - A u g ) a mmon ia c on c dee p l a y e r 1.04900 0.40000 di m e ns i onl e ss I ns h or e o ff s h or e r a t i o o f a nnu a l m ea n p h yt sur face l a y e r c on c Table 15: Monthly ambiental data from the Saint Helena relevant to the period 2010-2019, or a general value when specific period data was unavailable. Sources: GlobColour, Copernicus. 52 | Target data M ont h V a r iab l e m edia n low e r ce nt i l e upp e r ce nt i l e U n i ts low ce nt v a lu e upp ce nt v a lu e J a n sur face ch lorop h yll 0.061 0.058 0.066 m g m 3 0.200 0.800 F eb sur face ch lorop h yll 0.054 0.052 0.057 m g m 3 0.200 0.800 M a r sur face ch lorop h yll 0.053 0.049 0.055 m g m 3 0.200 0.800 A pr sur face ch lorop h yll 0.058 0.055 0.060 m g m 3 0.200 0.800 M a y sur face ch lorop h yll 0.068 0.064 0.073 m g m 3 0.200 0.800 J un sur face ch lorop h yll 0.087 0.081 0.091 m g m 3 0.200 0.800 J ul sur face ch lorop h yll 0.109 0.106 0.126 m g m 3 0.200 0.800 A u g sur face ch lorop h yll 0.132 0.121 0.139 m g m 3 0.200 0.800 S e p sur face ch lorop h yll 0.151 0.140 0.159 m g m 3 0.200 0.800 O c t sur face ch lorop h yll 0.135 0.129 0.146 m g m 3 0.200 0.800 N ov sur face ch lorop h yll 0.109 0.098 0.112 m g m 3 0.200 0.800 D ec sur face ch lorop h yll 0.072 0.066 0.083 m g m 3 0.200 0.800 Harverting ratio To derive the regional harvest ratio, we calculated the harvest ratio for each commercial guild by dividing the annual catches (landings + discards) of each guild by the biomass of the compiled commercial species (refer to the catch methodology in the Fishing Fleet Section and biomass methodology in the Target Data Section). For predator guilds, the annual bycatch value was used instead of the catch value. The input unit for StrathE2E is the harvesting ratio per day. 53 This provided sufficient information to calculate initial values for the scaling parameters linking effort to harvest ratios using the function e2e_calculate_hrscale(). We estimated the effort-harvest ratio scaling parameters required to produce the best fit to the 2010-2019 target data using the function e2e_optimize_hr(). This optimization was applied exclusively to the Migratory, Demersal, and CS benthos guilds, as the fishing model does not include DF benthos and CS zooplankton, and for the planktivorous guild we lacked the necessary biomass data to perform these calculations. Migratory fish Atlantic-wide stock assessments conducted by ICCAT indicate that some migratory species, such as bigeye tuna ( Thunnus obesus ) and Atlantic white marlin ( Kajikia albida ), were overexploited or undergoing overfishing during the baseline period (ICCAT 2018, 2019a). In contrast, the yellowfin tuna ( Thunnus albacares ) and the eastern stock of skipjack tuna ( Katsuwonus pelamis ) was likely neither overfished nor subject to overfishing during the baseline period or the most recent assessment ICCAT (2022b). However, it is worth noting that Saint Helena’s contribution to Atlantic fisheries catches is minimal. To calculate the harvesting ratio, we divided the estimated catch (discussed in the Fishing Fleet Section) by the estimated biomass (outlined in Event Timing Section) averaged for the four climate projections. Based on rough estimates derived from catch and biomass data used for model parameterization, we estimated a harvesting ratio of 2.14E-05. 54 | Target data Demersal The rock hind grouper ( Epinephelus adscensionis ) and the deep-water bullseye ( Cookeolus japonicus ) are highly vulnerable to overfishing, and no maximum sustainable yield estimates are currently available for Saint Helena (Choat and Robertson 2008). According to Riley, Ball, and Cowburn (2020), the grouper is considered to be sustainably fished in Saint Helena; however, there are signs of fishing pressure on the population, and more data are needed to conduct a comprehensive stock assessment (Choat and Robertson 2008). To calculate the harvesting ratio, we divided the estimated catch (discussed in the Fishing Fleet Section) by the estimated biomass of the guild (outlined in the Target Data Section). Note that some species are part of the harvesting ratio of the guild although it is not harvested. Based on rough estimates derived from catch and biomass data used for model parameterization, we estimated a demersal fish harvesting ratio of 3.72E-07. Seabirds For seabirds, we performed a rough estimate of the harvesting ratio by dividing the daily bycatch data (as detailed in the Fishing Fleet Section) by the annual biomass estimate (detailed in Target Data Section). This calculation resulted in a harvesting ratio of 1.32E-05. 55 Model fitting Model parameter optimization is crucial for aligning the model with observational data. The stationary fitting method adjusts parameters so that the model’s annual averages, integrated fluxes, or seasonal cycles closely match as closely as possible long-term ecosystem observations. The StrathE2E2 applies a stationary state fitting method that relies on likelihood estimation and a simulated annealing process using the Metropolis-Hastings algorithm. This algorithm automates the acceptance or rejection of new randomly generated parameter sets to maximize the likelihood of observed target indices given the model’s parameter values. All the details regardings the StrathE2E computational scheme is detailed at: StrathE2E2 version 4.0.1: Model parameter optimization, sensitivity analysis and Monte Carlo computation of credible intervals of outputs. In this implementation, model fitting was performed using a common set of parameters for both GFDL models SSPs and another for both CNRM SSPs, utilizing the corresponding NEMO-ERSEM environmental driving data from the 2010–2015 period. Additionally, the 2010–2019 guild-level harvest ratios for fish and invertebrates were incorporated, along with a comprehensive dataset on the ecological state of the system for the same period, shared across CNRM and GFDL variations, as detailed at the Target data Section. The results are shown graphically in Figures 8-11. The optimization process returned an overall likelihood of 0.764 for CNRM SSP1-2.6, 0.767 for CNRM SSP3-7.0, 0.754 for GFDL SSP1-2.6, 0.750 for GFDL SSP3-7.0. 56 | Model fitting Table 16: Observational indices of the 2010-2019 state of the Saint Helena MPA Implementation to which the CNRM SSP1-2.6 model variant was fitted, the corresponding SD, model data and Chi (partial likelihood of the observed value given the parameters). M ea sur e SD M o de l da t a C hi U n i ts D e s c r i pt i on 862.81 50.00 938.30 1.14 m MN / m 2/ y A nnu a l tot a l pr i m a ry pro d u c t i on 0.17 0.02 0.16 0.06 m MN / m 2/ y A nnu a l de m e rs a l fi s h g ross pro d u c t i on 1.8 e -04 1.3 e -04 1.7 e -04 4.2 e -03 m MN / m 2/ y A nnu a l n e t pro d u c t i on o f bi r d s 1.7 e -03 1.3 e -03 8.3 e -04 0.22 m MN / m 2/ y A nnu a l n e t pro d u c t i on o f ce t acea ns 0.27 0.20 0.35 0.09 di m e ns i onl e ss P roport i on pl a n k t i vorous fi s h i n die t o f bi r d s 1.3 e -05 1.0 e -05 1.5 e -05 0.02 m MN / m 2/ y A nnu a l pl a n k t i vorous fi s h l a n di n g s ( l i v e w eigh t ) 3.1 e -05 2.0 e -05 2.6 e -05 0.03 m MN / m 2/ y A nnu a l de m e rs a l fi s h l a n di n g s ( l i v e w eigh t ) 1.7 e -03 1.1 e -03 1.7 e -03 1.2 e -03 m MN / m 2/ y A nnu a l m ig r a tory fi s h l a n di n g s ( l i v e w eigh t ) 1.0 e -07 1.3 e -07 1.1 e -07 2.0 e -03 m MN / m 2/ y A nnu a l ca rn / s ca v be nt h os l a n di n g s ( l i v e w eigh t ) 10.00 7.50 13.54 0.11 / y A nnu a l g ross PB r a t i o l a rv ae o f susp / de p be nt h os 10.00 7.50 10.20 3.7 e -04 / y A nnu a l g ross PB r a t i o l a rv ae o f ca rn / s ca v be nt h os 10.00 7.50 3.92 0.33 / y A nnu a l g ross PB r a t i o susp / de p feedi n g be nt h os 1.12 0.84 0.69 0.13 / y A nnu a l g ross PB r a t i o ca rn / s ca v feedi n g be nt h os 17.30 12.97 2.39 0.66 / y A nnu a l g ross PB r a t i o omn i v zoopl a n k ton 6.65 4.99 2.08 0.42 / y A nnu a l g ross PB r a t i o ca rn i v zoopl a n k ton 4.00 3.00 0.85 0.55 / y A nnu a l g ross PB r a t i o l a rv ae o f pl a n k t i vorous fi s h 4.00 3.00 2.27 0.17 / y A nnu a l g ross PB r a t i o l a rv ae o f de m e rs a l fi s h 2.25 1.69 0.65 0.45 / y A nnu a l g ross PB r a t i o pl a n k t i vorous fi s h 0.73 0.55 0.02 0.84 / y A nnu a l g ross PB r a t i o de m e rs a l fi s h 1.30 0.98 0.10 0.75 / y A nnu a l g ross PB r a t i o m ig r a tory fi s h 5.40 4.05 0.02 0.88 / y A nnu a l n e t PB r a t i o bi r d s 0.05 0.04 0.06 5.3 e -04 / y A nnu a l n e t PB r a t i o ce t acea ns 0.07 0.03 NA NA m g / m 3 A v e r age w i nt e r ( N ov - F eb ) ch lorop h yll c on c s ha llow l a y e r 0.10 0.05 NA NA m g / m 3 A v e r age summ e r ( M a y - A u g ) ch lorop h yll c on c s ha llow l a y e r 0.66 0.58 0.27 0.23 m MN / m 3 A v e r age w i nt e r ( N ov - F eb ) n i tr a t e c on c s ha llow l a y e r 0.32 0.10 0.27 0.11 m MN / m 3 A v e r age summ e r ( M a y - A u g ) n i tr a t e c on c s ha llow l a y e r 1.05 0.40 0.63 0.54 di m e ns i onl e ss I ns h or e o ff s h or e r a t i o o f a nnu a l m ea n p h yt sur face l a y e r c on c 2.6 e -07 2.3 e -07 3.3 e -07 0.04 m MN / m 2/ y A nnu a l b y ca t ch o f bi r d s 0.15 0.05 0.16 0.02 di m e ns i onl e ss P roport i on o f m ac rop h yt e a nnu a l n i tro ge n upt ake e xport ed a s beachca st 57 Figure 8: Annual integrated or averaged results from the best-fit CNRM SSP1-2.6 2010-2019 stationary model variant, compared with observed data from Saint Helena MPA. Black boxes and whiskers represent the corresponding variability in measurements aggregated over the 2010– 2019 period. The visualization was generated using the function e2e_compare_obs(selection=‘ANNUAL’). 64 | End End Acknowledgements Financial support for the development of the Saint Helena EEZ implementation of StrathE2E came from the European Union Horizon 2020 research and innovation programme (Mission Atlantic - No.862428). We are grateful to Merillet Laurène and CLS team for the tunas data from SEAPODYM, Silvia Malagoli from disturbance rates approximations, and Yuri Artioli and PLM team for the NEMO-ERSEM model outputs. Appendix The species mentioned in this documentation and/or in the data used in the construction of the model are in Table 20 by guild of the StrathE2E model. Table 20: List of the main species considered per StrathE2E guild in the Saint Helena Island EEZ model. 65 C ommon n a me E nglish Ta xon C et a ce a ns Southern right whale Eubalaena australis Pygmy sperm whale Kogia breviceps Dwarf sperm whale Kogia sima Humpback whale Megaptera novaeangliae Killer whale Orcinus orca Sperm whale Physeter macrocephalus False killer whale Pseudorca crassidens Pan-tropical spotted dolphin Stenella attenuata Rough toothed dolphins Steno bredanensis Bottlenose dolphin Tursiops truncatus S e a birds Black noddy Anous minutus Brown noddy Anous stolidus Bulwer's petrel Bulweria bulwerii Lesser frigatebird Fregata ariel Great frigatebird Fregata minor White tern/Fairy tern Gygis alba Madeiran storm petrels Oceanodroma castro Sooty tern Onchyprion fuscata White-faced storm petrel Pelagodroma marina Red-billed tropicbird Phaethon aethereus Sooty shearwater Puffinus griseus Sargasso shearwater Puffinus lherminieri Masked booby Sula dactylatra Brown booby Sula leucogaster Red-footed booby Sula sula 66 | End C ommon n a me E nglish Ta xon S e a birds Arctic Skua Stercorarius parasiticus Pomarine Skua Stercorarius pomarinus M igr a tory Wahoo Acanthocybium solandri Thresher shark Alopias vulpinus Silky shark Carcharhinus falciformis Galapagos shark Carcharhinus galapagensis Oceanic whitetip shark Carcharhinus longimanus Dolphinfish Coryphaena spp. Pompano dolphinfish Coryphaena equiselis Other sharks n.i. Elasmobranchii Black marlin Istiompax indica Marlins n.i. Istiophoridae Shortfin mako Isurus oxyrinchus Atlantic white marlin Kajikia albida Skipjack tuna Katsuwonus pelamis Blue marlin Makaira nigricans Blue shark Prionace glauca Whale shark Rhincodon typus Mackerels, tunas, bonitos Scombridae n.i. Yellowtail Seriola lalandi Almaco jack Seriola rivoliana Hammerhead shark Sphyrna spp. Albacore Thunnus alalunga Yellowfin tuna Thunnus albacares Bigeye tuna Thunnus obesus 67 C ommon n a me E nglish Ta xon M igr a tory fish Tuna n.i. Thunnus spp. Swordfish Xiphias gladius D emers a l fish Island Hogfish/Parrotfish Bodianus insularis Ocean triggerfish Canthidermis sufflamen St Helena butterflyfish Chromis sanctaehelenae Brown chromis Chromis multilineata Deep-water bullseye Cookeolus japonicus Rock hind grouper Epinephelus adscensionis Spotted moray Gymnothorax moringa Glasseye snapper Heteropriacanthus cruentatus Squirrelfish Holocentrus adscensionis Other moray eels Muraenidae Blackbar soldierfish Myripristis jacobus Bastard cavalley pilot Stegastes sanctaehelenae Greenfish Thalassoma sanctaehelenae P l a nktivorous fish Mackerel scad/Kingston Decapterus macarellus Stonebrass scad Decapterus muroadsi Round scad/Summer stonebrass Decapterus punctatus Rainbow runner Elagatis bipinnulata Pufferfish Lagocephalus lagocephalus Mackerel Scomber japonicus/colias Bigeye scad/Steenbrass Selar crumenophthalmus Ca rnivorous a nd sc a vengers benthos Common octopus Octopus occidentalis Brown spiny lobster Panulirus echinatus 68 | End References AIG. 2016. “Ascension Island Government: Epinephelus Adscensionis Species Action Plan.” The Ascension Island Biodiversity Action Plan, Ascension Island Government Conservation Department, Georgetown, Ascension Island . https://www.ascension.gov.ac/wp‐content/uploads /2019/09/ROCKHIND‐GROUPER‐SAP‐edited.pdf. Artioli, Y., A. Katavouta, J. Holt, J. Harle, G. Galli, L. Mora, L. Merillet, and A. Conchon. 2023. “Deliverable 6.2 - 2020-2070 Future Projections (NEMO-ERSEM & SEAPODYM).” Mission Atlantic Project Report, Work Package 6 - Dynamics of ecosystem state; resources. www.missionatlantic.eu. Beard, A., R. M. Mirra, E. Clingham, L. Henry, R. J. Thomas, and F. Hailer. 2023. “Breeding Ecology, Population Size and Nest Site Preferences of Red-Billed Tropicbirds at St Helena, South Atlantic Ocean.” Emu - Austral Ornithology 123 (3): 185–94. https://doi.org/10.1080/01584197.2023 .2205595. Beard, A., R. J. Thomas, J. C. Xavier, E. Clingham, L. Henry, R. M. Mirra, and F. Hailer. 2024. “A Taste for Squid: Red-Billed Tropicbird (Phaethon Aethereus) Diet in the South Atlantic.” Waterbirds 47 (1): 1–16. https://doi .org/10.1675/063.047.0106. Bester, M. N. 2021. “Status of Pinnipeds on Mid-Atlantic Ridge Islands, South Atlantic Ocean.” Polar Biology 44 (4): 865–71. https://doi.org/10 .1007/s00300‐021‐02838‐z. Bridges, A. E. H., D. K. A. Barnes, J. B. Bell, R. E. Ross, and K. L. Howell. 2021. “Benthic Assemblage Composition of South Atlantic Seamounts.” Frontiers in Marine Science 8. https://doi.org/10.3389 /fmars.2021.660648. Brown, J. 2014. Marine Life of St Helena . St Helena: St Helena Government; Pisces Publications. Brown, J., A. Beard, E. Clingham, R. Fricke, L. Henry, and P. Wirtz. 2019. “The Fishes of St Helena Island, Central Atlantic Ocean - New Records and an Annotated Check-List.” Zootaxa 4543: 151–94. https://doi.org/10 .11646/zootaxa.4543.2.1. 69 Bugoni, L., T. S. Neves, N. O. Leite, D. Carvalho, G. Sales, R. W. Furness, C. E. Stein, F. V. Peppes, B. B. Giffoni, and D. S. S. Monteiro. 2008. “Potential Bycatch of Seabirds and Turtles in Hook-and-Line Fisheries of the Itaipava Fleet, Brazil.” Fisheries Research 90 (1): 217–24. https://doi.org /10.1016/j.fishres.2007.10.013. Capitani, L., J. N. de Araujo, E. A. Vieira, R. Angelini, and G. O. Longo. 2021. “Ocean Warming Will Reduce Standing Biomass in a Tropical Western Atlantic Reef Ecosystem.” Ecosystems 25 (4): 843–57. https:// doi.org/10.1007/s10021‐021‐00691‐z. Carleton, C., P. Medley, T. Southall, and M. Gill. 2010. “MSC Sustainable Fisheries Certification: St Helena Pole and Line and Rod and Line Tuna Fisheries for Albacore, Bigeye, Yellowfin and Skipjack Tuna.” Public Certification Report - Prepared for the St Helena Development Agency. Food Certification International Ltd. Choat, J. H., and D. R. Robertson. 2008. “An Ecological Survey of the St Helena and Ascension Island Populations of the Jack (Epinephelus Adscensionis) with a Review of Management Options.”https://doi.org /10.13140/RG.2.1.2333.9368. Christensen, L. B. 2006. “Reconstructing Historical Abundances of Exploited Marine Mammals at the Global Scale.” Retrospective Theses and Dissertations, 1919-2007. PhD thesis, University of British Columbia. https://doi.org/10.14288/1.0074892. Clingham, E., L. Henry, and A. Beard. 2013. “Monitoring Population Size of St. Helena Cetaceans.” Environmental Management Division, Saint Helena Government. Collins, M. A. 2017. “St Helena Fisheries Sector: Review & Strategy (20162025).” Www.sainthelena.gov.sh/wp-content/uploads/2012/08/StHelena-Fisheries-Strategy-Summary.pdf. A report for the St Helena Government. Cowburn, B., J. Graham, M. Schratzberger, J. Brown, L. Henry, E. Clingham, A. Beard, and P. Nelson. 2021. “Rocky Reefs of St Helena and the Tropical Atlantic: How the Lack of Coral and an Isolated Oceanic Location Drive Unique Inshore Marine Ecology.” Marine Ecology Progress Series 663: 31–49. https://www.int‐res.com/abstracts/meps /v663/p31‐49/. Devlin, M., J. Barry, D. Mills, R. Gowen, J. Foden, D. Sivyer, and P. Tett. 2008. “Relationships Between Suspended Particulate Material, Light 70 | End Attenuation and Secchi Depth in UK Marine Waters.” Estuarine, Coastal and Shelf Science 79 (September): 429–39. https://doi.org/10.1016/j.ecss .2008.04.024. Edwards, A. J. 1990. Fish and Fisheries of Saint Helena Island . Edited by England. Centre for Tropical Coastal Management Studies University of Newcastle upon Tyne. FAO. 2011. “Yellowfin Tuna - Atlantic.” In: Fisheries and Aquaculture [online]. https://firms.fao.org/firms/en/resource/iccat0019__2011__en ?status=published. Ferrari, D. S., S. R. Floeter, F. Leprieur, and J. P. Quimbayo. 2023. “A TraitBased Approach to Marine Island Biogeography.” Journal of Biogeography 50 (3): 528–38. https://doi.org/10.1111/jbi.14549. Ferrari, D. S., L. T. Nunes, K. L. Jones, C. E. L. Ferreira, and S. R. Floeter. 2024. “Thermal Tolerance as a Driver of Reef Fish Community Structure at the Isolated Tropical Mid-Atlantic Ridge Islands.” Marine Environmental Research 199: 106611. https://doi.org/10.1016/j.marenvres .2024.106611. Freire, K., J. A. Aragão, A. Araújo, A. Silva, M. Bispo, G. Velasco, M. Carneiro, et al. 2015. “Reconstruction of Catch Statistics for Brazilian Marine Waters (1950-2010).” Fisheries Centre Research Reports 23 (June): 3–30. Freire, K., V. Christensen, and D. Pauly. 2008. “Description of the East Brazil Large Marine Ecosystem Using a Trophic Model.” Scientia Marina 72 (September). Froese, R., and D. Pauly. 2024. “FishBase. World Wide Web Electronic Publication.” 2024. www.fishbase.org. Gilman, E., P. Suuronen, and M Chaloupka. 2017. “Discards by Global Tuna Fisheries.” Marine Ecology Progress Series 582 (November). https:/ /doi.org/10.3354/meps12340. GlobColour, Copernicus. 2023. “Global Ocean Colour, Bio-Geo-Chemical, L4 (Monthly and Interpolated) from Satellite Observations (1997Ongoing). ID OCEANCOLOUR_GLO_BGC_L4_MY_009_104 e.u. Copernicus Marine Service Information (CMEMS). Marine Data Store (MDS).”10.48670/moi-00281. Gohin, F. 2011. “Annual Cycles of Chlorophyll-a, Non-Algal Suspended Particulate Matter, and Turbidity Observed from Space and in-Situ in 71 Coastal Waters.” Ocean Science 7 (5): 705–32. https://doi.org/10.5194 /os‐7‐705‐2011. Greenstreet, S. P. R. 1996. “Scottish Fisheries Research Report: Estimation of the Daily Consumption of Food by Fish in the North Sea in Each Quarter of the Year.” Vol. 55. The Scottish Office Agriculture, Environment; Fisheries Department. Heath, M. R. 2012. “Ecosystem Limits to Food Web Fluxes and Fisheries Yields in the North Sea Simulated with an End-to-End Food Web Model.” Progress in Oceanography Special issue: End-to-end modelling: Towards Comparative Analysis of Marine Ecosystem Organisation (102): 42–66. Heath, M. R., D. C. Speirs, I. Thurlbeck, and R. J. Wilson. 2021. “StrathE2E2: An r Package for Modelling the Dynamics of Marine Food Webs and Fisheries.” Methods in Ecology and Evolution 12 (2): 280–87. https://doi .org/10.1111/2041‐210X.13510. Heath, M. R., R. Wilson, and D. C. Speirs. 2015. “Modelling the WholeEcosystem Impacts of Trawling. A Study Commissioned by Fisheries Innovation Scotland (FIS).” FIS. Hersbach, H., B. Bell, P. Berrisford, S. Hirahara, A. Horányi, J. MuñozSabater, J. Nicolas, et al. 2020. “The ERA5 Global Reanalysis.” Quarterly Journal of the Royal Meteorological Society 146 (730): 1999–2049. https: //doi.org/10.1002/qj.3803. ICCAT. 2018. “Report of the 2018 ICCAT Bigeye Tuna Stock Assessment Meeting.” Pasaia, Spain: ICCAT - International Commission for the Conservation of Atlantic Tunas. https://www.iccat.int/Documents /Meetings/Docs/2018/REPORTS/2018_BET_SA_ENG.pdf. ———. 2019a. “Report of the 2018 ICCAT White Marlin Stock Assessment Meeting.” Miami, USA: ICCAT - International Commission for the Conservation of Atlantic Tunas. https://www.iccat.int/Documents/SCRS /DetRep/WHM_SA_ENG.pdf. ———. 2019b. “Report of the 2019 ICCAT Yellowfin Tuna Stock Assessment Meeting.” Grand-Bassam, Cote d’Ivoire: ICCAT - International Commission for the Conservation of Atlantic Tunas. https:/ /www.iccat.int/Documents/Meetings/Docs/2019/REPORTS/2019_YFT _SA_ENG.pdf. ———. 2022a. “Report for Biennial Period, 2020-21, Part II (2021) - Vol. 3: Annual Reports.” Madrid, Spain: International Commission for the 72 | End Conservation of Atlantic Tunas (ICCAT). ———. 2022b. “Report of the 2022 Skipjack Tuna Data Preparatory Meeting (Online, 21-25 February 2022).” ICCAT - International Commission for the Conservation of Atlantic Tunas. https://www.iccat .int/Documents/CVSP/CV079_2022/n_1/CV079010001.pdf. ———. 2023. “International Commission for the Conservation of Atlantic Tuna (ICCAT) Statistical Databases.” Jones, C. D., J. K. Hughes, N. Bellouin, S. C. Hardiman, G. S. Jones, J. Knight, S. Liddicoat, et al. 2011. “The HadGEM2-ES Implementation of CMIP5 Centennial Simulations.” Geoscientific Model Development 4 (3): 543–70. https://doi.org/10.5194/gmd‐4‐543‐2011. Kelleher, K. 2005. “Discards in the World’s Marine Fisheries. An Update.” FAO Fisheries Technical Paper 470: 131. Laptikhovsky, Vl., J. Naulaerts, E. Clingham, M. A. Collins, M. Cranfield, L. Henry, A. Small, T. Stamford, J. Xavier, and S. R. Wright. 2021. “Comparative Feeding Strategies of Yellowfin Tuna Around St Helena and Adjacent Seamounts of the South Atlantic Ocean.” Journal of Applied Ichthyology 37 (1): 38–52. https://doi.org/10.1111/jai.14122. Lehodey, P., R. Murtugudde, and I. Senina. 2010. “Bridging the Gap from Ocean Models to Population Dynamics of Large Marine Predators: A Model of Mid-Trophic Functional Groups.” Progress in Oceanography 84 (1): 69–84. https://doi.org/10.1016/j.pocean.2009.09.008. Lehodey, P., I. Senina, and R. Murtugudde. 2008. “A Spatial Ecosystem and Populations Dynamics Model (SEAPODYM) – Modeling of Tuna and Tuna-Like Populations.” Progress in Oceanography 78 (4): 304–18. https:/ /doi.org/10.1016/j.pocean.2008.06.004. Lessa, R., P. Mafalda, R. Lucchesi, J. Júnior, T. Vaske Júnior, and D. Hellebrandt. 1999. “Distribution and Abundance of Ichthyoneuston at Seamounts and Islands Off North-Eastern Brazil.” Archives of Fishery and Marine Research 47: 133–46. Lessa, R., M. Nóbrega, and J. L. B. Junior. 2004. “Dinâmica de Populações e Avaliação de Estoques Dos Recursos Pesqueiros Da Região Nordeste.” Recife, Brazil: Program REVIZEE, Score Nordeste. https://doi.org/10.13140/RG.2.2.27882.88003. MacLeod, C. D., and E. Bennett. 2007. “Pan-Tropical Spotted Dolphins (Stenella Attenuata) and Other Cetaceans Around St Helena in the 73 Tropical South-Eastern Atlantic.” Journal of the Marine Biological Association of the United Kingdom 87 (1): 339–44. https://doi.org/10.1017 /S0025315407052502. Martin, S. M., M. Soeffker, A. Schofield, R. Hobbs, T. Glass, and S. A. Morley. 2021. “Cetaceans Sightings During Research Cruises in Three Remote Atlantic British Overseas Territories.” Frontiers in Marine Science 8. https://doi.org/10.3389/fmars.2021.660152. Merillet, L., O. Titaud, and A. Conchon. 2024. “Micronecton and Tuna (Albacore and Skipjack) Mean Biomass over 1985-2014 and over 20152070 for CNRM and GFDL Models (with SSP3-7.0 and SSP1-2.6) from SEAPODYM Model Outputs for Atlantic Ocean.” Collecte Localisation Satellites (CLS), France. 2024. https://marineinfo.org/id/dataset/8535. Morissette, L., J. L. Melgo, K. Kaschner, and L. Gerber. 2009. “Modelling the Trophic Role of Marine Mammals in Tropical Areas: Data Requirements, Uncertainty, and Validation.” Text. 2009. https://open .library.ubc.ca/collections/52383/items/1.0074740. Nolan, E. T., D. K. A. Barnes, J. Brown, K. Downes, P. Enderlein, E. Gowland, and P. Brickle. 2017. “Biological and Physical Characterization of the Seabed Surrounding Ascension Island from 100–1000 m.” Journal of the Marine Biological Association of the United Kingdom , 647–59. Oppel, S., A. Beard, D. Fox, E. Mackley, E. Leat, L. Henry, E. Clingham, et al. 2015. “Foraging Distribution of a Tropical Seabird Supports Ashmole’s Hypothesis of Population Regulation.” Behavioral Ecology and Sociobiology 69 (6): 915–26. https://doi.org/10.1007/s00265‐015‐1903‐3. Pace, M. C., D. M. Bailey, D. W. Donnan, B. E. Narayanaswamy, H. J. Smith, D. C. Speirs, W. R. Turrell, and M. R. Heath. 2021. “Modelling Seabed Sediment Physical Properties and Organic Matter Content in the Firth of Clyde.” Earth System Science Data Discussions 2021: 1–35. https://doi.org/10.5194/essd‐2021‐23. Pauly, D., D. Zeller, and M. L. D. Palomares. 2023. “Sea Around Us Concepts, Design and Data.” Pinheiro, P., F. Hazin, P. Travassos, P. Oliveira, F. Carvalho, and M. Rêgo. 2011. “The Reproductive Biology of the Rainbow Runner, Elagatis Bipinnulata (Quoy & Gaimard, 1825) Caught in the são Pedro and são Paulo Archipelago.” Brazilian Journal of Biology 71 (February): 99–106. https://doi.org/10.1590/S1519‐69842011000100015.