Vegetation census data from montane meadow plots are used to test a predict
ed connection between the species-area relationship, S = cA(z), and the dep
endence of interpatch species turnover on patch area A, interpatch distance
D, and the species-area exponent z. At small spatial scales, from D approx
imate to 1-10 m, where species-area parameters can be independently estimat
ed, the prediction is confirmed; at larger scales, from D approximate to 1-
10(4) m, the scale-dependence of z is deduced. A predicted dependence of sp
ecies richness on the shape of censused patches is also confirmed. Our resu
lts indicate that readily obtainable species-turnover data between distant
small patches can be used to estimate species-area exponents at landscape s
cales where census data for nested areas are generally not available, there
by improving our ability to estimate landscape-scale species richness and r
arity.