Two different measurements are doing work here, and they do not carry the same weight. The 18 million figure comes from joining household-level poverty data to estimates of populations living close to mapped seagrass [5]. That is a spatial overlay with survey data attached, and it tells you how many people in poverty fish where meadows are, but it stops short of telling you how much of any given meal comes out of one. The 6.5 million figure is the load-bearing one, because it is a claim about exclusivity: those households have the habitat as their entire basis for subsistence and fishing-related income [3].
The arithmetic between the two is where the food-security argument actually lives. Divide 6.5 by 18 and you get about 36 percent [15]; subtract, and roughly 11.5 million people fish these meadows but have somewhere else to go [16]. A household with a second fishing ground takes a meadow's decline as a cut in earnings. A household without one takes it as a subtraction from what it eats. Those are different exposures, and lumping them into one headline total loses the distinction that matters for triage.
The study also reports that households in extreme poverty were significantly more likely to be wholly seagrass-dependent, and that seagrass-dependent households earned 23 percent less than other fishers [6][7]. The authors read this as a coastal poverty trap, where households with few alternatives stay tied to low-return livelihoods in a vulnerable habitat [8]. This design leaves the direction of the arrow unsettled. Poverty plausibly pushes people toward the cheapest fishery to reach, one you can walk into with limited gear [4], and thin returns then keep them there. A cross-sectional income comparison cannot separate that from the reverse reading, and the paper frames its result as a suggestion rather than a mechanism [8].
The thing this does not tell you is what happens after a meadow goes. No one here has tracked an actual seagrass loss event and measured the protein and cash that left with it. Dr. Benjamin Jones of Project Seagrass, one of the lead researchers, puts it as loss of food, income and security following loss of habitat [9], which is a reasonable inference from measured dependence, though it sits in a weaker evidentiary class than the count itself.
The count is the contribution, and it is explicitly a floor: the paper says at least 18 million [1]. Project Seagrass describes it as the first large-scale empirical evidence that these meadows function as a safety net for people in extreme poverty [12], and the same release notes that the habitat is largely absent from poverty, development and conservation agendas [10]. That absence is the practical finding. Coastal development and land-based pollution are named as the primary threats, with the most dependent communities holding little influence over the decisions driving decline [11]. Appraisals of those decisions have not had a number for the people with no substitute. Now there is one, with an error bar pointing upward.
