Part 1: how warm is a nest and who hatches? Part 2: what a seagrass reef can feed over the decades.
The science
Sand temperature. A buried clutch feels the damped, delayed annual temperature wave of the sand: T(z,t) = T̄ + A e−z/δ cos(ω(t − t0) − z/δ), δ = √(2α/ω) with α = 3×10−7 m²/s (about 1.7 m for the annual wave). Shade lowers the mean by 2.5 °C and the amplitude by 25 %, dark sand adds 1.5 °C, climate warming adds ΔT. Late in development the embryos warm the nest (metabolic heating, up to +1.5 °C).
Development and sex. Development speed dθ/dt = 2(T−29)/10/55 d (55 d at 29 °C, Q10 = 2). Sex is set by the mean temperature in the middle third of development (θ from 1/3 to 2/3): P(female) = 1/(1 + e−(T−Tpiv)/s), s = TRT/(2 ln 19), so 5 % to 95 % female spans the transitional range. Hatch success falls with the peak nest temperature: 0.92/(1 + e(Tmax−35)/0.7).
Population. Seven stages (first-year, juvenile, subadult, adult, split by sex). One year of food-limited reef: seagrass dS/dt = rS(1 − S/K) − cN S/(h + S), K = reef health; food satisfaction φ scales survival (φ0.4). Females breed every 3 years, 4 clutches of 110 eggs; mating success 1 − e−3·males/females (so extreme female bias lowers fertile eggs); first-year survival is Beverton-Holt crowded. The matrix eigenvalue λ is the growth rate in a healthy, uncrowded reef.
Honest simplifications. Sand heat is a one-dimensional analytic wave (no rain, no moisture, no daily cycle, no beach slope); a single temperature for sex and for hatching; all demographic rates are illustrative round numbers in the right order of magnitude, not a fit to a particular stock; migration, bycatch, predators and nest loss are not modelled. Results show how the pieces interact, not a forecast for any real beach.
Models built with Blender (turtle, corals, fish, seagrass, reef floor) and rendered with Three.js.