Table 1. Network criteria, conservation targets, and metrics.

CBD network criteria (bold) including definitions quoted from CBD (29), metrics (italics), conservation targets, and metric equations used in this study, with relevant comments.

Network criteria Metrics	Definitions and metric equations (normalized to 0 to 5 range)	Conservation targets and comments
Important areas	“[Important Areas are] geographically or oceanographically discrete areas that provide important services to one or more species/populations of an ecosystem or to the ecosystem as a whole, compared to other surrounding areas…”
Major transform faults	APEI percent coverage/100% × 5.	The objective is to protect 100% of important areas. Scores are based on percent area conserved (for transition zones), percent by number of features conserved (for hybrid zones), and percent of length conserved (for transform faults).
Biogeographic transition zones
Genetic hybrid zones
Representativity	“Representativity is captured in a network when it consists of areas representing the different biogeographical subdivisions of the global oceans and regional seas that reasonably reflect the full range of ecosystems, including the biotic and habitat diversity of those marine ecosystems.”
Discrete habitat variables: Spreading ridge Active vents Inactive vents Fracture zones Seamounts	APEI percent coverage/50% × 5, where any score greater than 5 was set to 5.	The objective is to protect a representative amount (30 to 50%) of key habitat within the study region. Scores are based on percent area conserved (for spreading ridges), percent by number of features conserved (for active and inactive vents, and seamounts), and by percent of length conserved (for transform faults).
		Note: Active hydrothermal vents and other vulnerable marine ecosystems are at risk of serious harm from SMS mining activities. We expect 100% of active hydrothermal vent ecosystems and other habitats at risk of serious harm to be protected through conservation measures, including, but not limited, to APEIs.
Continuous variables that describe the regional seascape: Slopes Depth Seafloor POC flux	5 − (RMSE × 5)	The objective is to mimic the distribution of variables determined to be key drivers of biodiversity in proportion to their occurrence in the management subunit. Root mean square error (RMSE) was calculated as the difference between cumulative frequency distributions within the APEI scenario and the study region. All variables were classified into 10 to 15 bins to remove the effect of the number of bins on RMSE.
Connectivity	“Connectivity in the design of a network allows for linkages whereby protected sites benefit from larval and/or species exchanges, and functional linkages from other network sites. In a connected network individual sites benefit one another.”
Regional connectivity	6 − (max distance between cores/75th percentile median dispersal distance), where any score greater than 5 was set to 5.	The objective is to ensure that there is no major disruption to dispersal across the network of APEIs. The maximum distance between APEIs compared to median faunal dispersal distances is an indicator of the potential for disrupting dispersal within the entire management subunit.
Network population persistence	6 − mean gap ratio (that is, the mean distance between cores/mean core length), where any score greater than 5 was set to 5.	The objective is to promote the viability of populations by self-seeding within APEIs and/or dispersal between APEIs. By minimizing the difference in length of APEI core areas versus distance between core areas, species that on average disperse beyond the APEI have a good chance of being able to disperse to adjacent APEIs. Minimizing this “gap ratio” should enhance persistence of species across the network, as well as within individual APEIs, and increase resilience across the network to localized disturbances.
Replication	“Replication of ecological features means that more than one site shall contain examples of a given feature in the given biogeographic area. The term “features” means “species, habitats and ecological processes” that naturally occur in the given biogeographic area.”
Replication	Number of APEIs where any score greater than 5 was set to 5.	The objective is to have three to five replicate APEIs within a management unit, to decrease the likelihood of local catastrophes causing systemic biodiversity loss.
Viability and adequacy	“Adequate and viable sites indicate that all sites within a network should have size and protection sufficient to ensure the ecological viability and integrity of the feature(s) for which they were selected.”
Total area	(APEI percent coverage/50%) × 5, where any score greater than 5 was set to 5.	The objective is to conserve an adequate portion (30 to 50%) of the management unit to ensure the viability of populations within it. Total area conserved is a proxy for overall adequacy of a network. The total area metric was calculated similarly to the habitat representativity metrics above.
Within APEI persistence	5 × (APEI core length/200 km), where any score greater than 5 was set to 5.	The objective is to ensure that APEIs are large enough to maintain minimum viable populations, and metapopulations, within a single APEI. The larger the APEI, the greater the probability self-recruitment within the APEI, and the lower the percentage of larval export from the APEI, which should enhance the persistence of populations, metapopulations, and communities within an APEI. 200 km was used as the minimum scale required to encompass two times the median dispersal distance of 75% of deep-sea fauna with known dispersal scales (53).
Climate Change: Absolute similarity	5 − (RMSE × 5)	The objective is to conserve areas where climate impacts would be minimized. The more close distributions of key climate variables (pH, temperature, dissolved O2 concentrations, and seafloor POC flux) in the future (that is, 2100) APEI cores mimic the current (that is, 2013) distribution in the management unit, the less impact is expected. RMSE was calculated as the difference between cumulative frequency distributions within the APEI scenario and the study region. All variables were classified into 10 to 15 bins to remove the effect of the number of bins on RMSE.
Climate change: Relative local change	(APEI percent coverage/50%) × 5, where any score greater than 5 was set to 5.	The objective is to conserve 30 to 50% of the areas projected to be least affected by climate change. Least affected cells were defined as the 10% of cells with the lowest percent change between current (2013) and predicted (2100) values of the four key climate variables (pH, temperature, dissolved O2 concentrations, and POC flux to the seafloor). The percent of those cells falling in APEI cores for each scenario was calculated following the approach used for representativity metrics (continuously distributed variables).