Table 1 Hypotheses proposed to explain latitudinal biodiversity gradients.

Associated predictor metrics tested in this study and transformations applied in linear models are also listed. SST, sea surface temperature; Chl, chlorophyll; NO3, nitrate; PAR, photosynthetically active radiation; sqrt, square root.

HypothesisMetricMeasurement
Temperature
Species richness increases with temperature (76), perhaps due
to increased metabolic rates leading to faster generation times
and hence increased rates of mutation, evolution, and accumulation
of species (30). Alternatively, niche conservatism may constrain
thermal tolerances of most species to warm conditions (13, 31, 77),
or temperature may indirectly link to diversity through its influence
on productivity.
SST (log)Mean SST of sites, transformed to units
of energy (−1/kT, where k is the Boltzmann
constant and T is in Kelvin) (30).
Primary productivity
Light irradiance and nutrients drive primary productivity,
which, at low to moderate levels at least, increases abundance
and species richness (32, 33).
Chl (log)
NO3 (log)
PAR (sqrt)
Satellite derived mean Chl, NO3, and PAR
measurements per 5–arc min (9.2 km) grid
cells compiled in Bio-ORACLE (ocean rasters
for analysis of climate and environment) (71).
Area
Species richness increases with geographic area, independent
of habitat diversity, in part through increased persistence of
animals with large range requirements (34).
Shelf (sqrt)Total continental shelf area (water depths,
<200 m), from NOAA (National Oceanic
and Atmospheric Administration) NGDC
(National Geophysical Data Center) (78).
Coral (log x + 1)Total area of coral reef within ecoregion,
assessed through “Reefs at Risk” (79).
Natural disturbance regime
Species richness decreases with, or peaks at intermediate
levels of, environmental disturbance.
CyclonePercentage of sites surveyed in an ecoregion
with a cyclone of at least category 3 passing
within 50 km in the past 10 years, from IBTrACS
(International Best Track Archive for Climate
Stewardship) (80).
SST range (log)Difference between annual mean monthly
maximum and minimum SST, averaged
across years, from Bio-ORACLE (71).
Climatic stability
Species richness is highest at locations with landscapes
little affected by recent cataclysmic geological and
climatic events and thus have avoided major extinctions.
Stability (log)Inverse of SD of mean water temperature
since the last glacial maximum, estimated
for three time periods: 0 ky (thousand years),
8 ky, and 21 ky before present, from
Paleo-MARSPEC (Ocean Climate Layers
for Marine Spatial Ecology) (81).
Fragmentation and connectivity
Species richness varies with rate of speciation, which is
dependent on frequency of vicariant events, including
breaks in population connectivity through isolation of
island populations, coastal fragmentation, and changing
dispersal pathways into adjoining regions.
Coast (log)Total length of coastline (km) within
the ecoregion.
ConnectNumber of abutting coastal ecoregions.
Islands (log)Number of separated coastline features on
map of the ecoregion.
Human disturbance
Species richness is negatively affected by human activities.
Pop (log x + 10)Estimated population density, calculated
by fitting a smoothly tapered surface to
each settlement point on a year 2000
world population density grid (82), using
the quadratic kernel function described
in the study of Silverman (83). Populations
were screened for a density greater than
1000 people per 0.04° cell, and the
search radius was set at 3.959° (28).