Tropical Forest Ecology: Sterile or Virgin for Theoreticians? 129
per year. Finally, letKb eth erat eat which rainfall,
dust from the air, and weathered bedrock supply
nitrogen per unit area of soil. Then gain balances
loss of soil nitrogen when
k(U 0 −U)+ 2 mU 0 =K
Since the forest takes up nutrient at the rate
U=U 0 /( 1 +Rv/R)=U 0 R/(R+Rv),
U 0 −U=[U 0 (R+Rv)−U 0 R]/(Rv+R)
=U 0 Rv/(Rv+R);
K= 2 mU 0 +k(U 0 −U)
= 2 mU 0 +kU 0 Rv/(Rv+R);
U 0 =K(Rv+R)/[kRv+ 2 m(Rv+R)]
U=U 0 R/(R+Rv)
=KR/[kRv+ 2 m(Rv+R)]
Nutrient uptake is increased by increasing fine
root biomass, especially if 2mR(k+m)Rv,
and by decreasing tree mortality, especially when
2 m(Rv+R)kRv.
Poor soils favor reducing nutrient losses. There-
fore, King’s theory predicts that on poorer soil,
trees, branches, and leaves should be longer lived,
wood production lower, and nutrients more effi-
ciently translocated from dying plant parts and
old wood to their new counterparts.
A fertility gradient in Amazonia supports some
of th es epr edictions. Soils ar emor ef ertil eat
Amazonia’s western edge, near the Andes, than in
central and eastern Amazonia (Malhietal. 2004).
As predicted, wood production is lower on poorer
soil. Bakeret al. (2004a, p. 360) found that wood
production, th eannual incr eas ein abov e-ground
woody biomassAGB, in tons ha−^1 year−^1 ,ofa
plot’s trees≥10 cm dbh surviving from on ec en-
sus to the next, was related to their annual basal
area increaseBA,inm^2 ha−^1 year−^1 , by the
regression
AGB=9.57(BA)+0.12(r^2 =0.89)
Between 1985 and 1992, annual basal area
increase averaged 0.64 m^2 ha−^1 in western Ama-
zonia and 0.4 m^2 ha−^1 in less fertile central and
eastern Amazonia (Lewisetal. 2004); accordingly
wood production averaged 6.3 tons ha−^1 year−^1
in western Amazonia and 4 tons ha−^1 year−^1
further east.
Also as predicted, tree mortality is lower on
poorer soils. Between 1985 and 1992, annual
mortality rate for trees≥10 cm dbh averaged
2% in western Amazonia and only 1% further
east, whether measured as the proportion of
trees dying per year, or the proportion of basal
area comprised by these dying stems (Table 3
in Lewiset al. 2004). Trees in central and east-
ern Amazonia are indeed built to last: their
wood density (g dry weight cm−^3 fresh volume),
a good measure of wood strength (Putzet al.
1983), averages 0.684, compared with 0.571
in western Amazonia (Bakeret al. 2004b). Poor
soil so promotes longevity over reproduction
that, in nutrient-starved heath forest of Borneo,
Table 8.3 Number of free-standing woody plants≥10 cm dbh (N), number of species among them (S), and
Fisher’sα, in relation to annual rainfall (P) and rainfall during the driest quarter (P 3 ) (both in mm) in 25 ha subplots
of selected continental forest dynamics plots of the Center for Tropical Forest Science.
Site NSα PP 3
Yasuni, Ecuadorian Amazonia 17,546 820 178 3081 594
Lambir, Sarawak, Malaysia 15,916 851 193 2664 498
Pasoh Reserve, Malaysia 13,276 604 130 1788 318
Korup, Cameroon 12,296 261 47 5272 172
Barro Colorado Island, Panama 10,728 206 36 2551 131
Huai Kha Khaeng, Thailan d10,938 185 32 1476 46
Note: Fisher’sαis defined by the relationS=αln( 1 +N/α)(Conditetal. 2004, p. 79).
Sources: Climate data from table 4.3 of Leigh (2004);N,S, andLfrom table 7.1 of Conditetal.(2004).