Spatial variations of litterfall silicon flux and plant-available silicon in highly weathered soil in a lowland mixed dipterocarp forest of Lambir Hills National Park in Borneo

Abstract

Tropical forest trees take up silicon (Si) and return it to the forest floor via leaf litterfall. Our objective was to explore to what extent litter Si flux and Si availability from the soil are spatially coupled. We examined these relationships within a 4-ha area of lowland mixed dipterocarp forest of Lambir Hills National Park in Borneo. Using leaf litter samples collected with litter traps, we found that Si concentration and flux of leaf litter ranged 2-23 mg Si g 1 and 0.8-13.1 g Si m 2 yr 1 , respectively, whereas water-extractable Si from 0-10 cm deep soil ranged from 5.9 to 24.5 mg kg 1 (0.7 to 3.0 g Si m 2 ) at 80 litter trap locations. There was no significant correlation among these three aspects of Si cycling via trees. Water-extractable soil Si from three 95 cm deep cores showed no significant change with depth, whereas in-situ measurements with six tension lysimeters showed higher soil-water Si concentration in the upper soil layer (0-5 cm depth). These results suggest that spatial variations of Si concentration and flux in leaf litter do not reflect those of soil Si availability, but are modulated by distribution of tree species that differ in Si uptake. Si returned to the soil via leaf litter did not show strong spatial signals probably because solubility of Si from dead leaves differs among species. At the stand level, our results are consistent with the perspective that litter Si input enriches plant-available Si pool in the upper soil horizons in tropical forests.

Litterfall silicon flux in relation to vegetation differences in old‐growth and logged lowland forests in Borneo

“…Hence, litterfall Si flux increases strongly when species that accumulate Si at high concentrations, such as bamboos, become highly abundant. The Si fluxes via tree leaf-litter excluding bamboo litters (53-94 kg ha À1 year À1 ) were similar to those previously reported from other tropical forests (33-85 kg ha À1 year À1 , Alexandre et al, 1997;Ishizawa et al, 2019;Lucas et al, 1993;Nakamura et al, 2019Nakamura et al, , 2020b. Not as strongly as bamboos, Macaranga with high dead-leaf Si also modulated Si flux in post-logging stands, compensating for the loss of Si flux due to the decreased abundance of Si-accumulating Dipterocarpaceae trees (Figure 2b).…”

“…Finally, the results supported our last hypothesis that water-extractable Si concentration in the soil should be related to these community-level changes in litter Si flux. The values of water extractable Si at our site were within the range reported from other tropical forests in SE Asia (7.4-21.5 mg kg À1 , Ishizawa et al, 2019;Nakamura et al, 2019Nakamura et al, , 2020b. Water-extractable soil Si was the highest in the RIL stand followed by the CL and oldgrowth stands (Figure 4), showing an among-site trend similar to that of the stand-level litter Si flux including bamboo litter (Figure 3).…”

“…Such species differences in physiological regulation of Si uptake in roots explain why tropical plants differ greatly in Si accumulation in shoots (0 < Si ≤126 mg g-leaf-dry-mass À1 ) as revealed by recent studies (Ishizawa et al, 2019 for Peninsular Malaysia;Nakamura et al, 2019 for Borneo;Schaller et al, 2018 in Panama). The presence of high Si accumulating species (>10 mg gdry-mass À1 defined by Ma & Takahashi, 2002) among dominant tropical trees also explains the reason why lowland tropical forests on deeply weathered soils cycle much larger quantities of Si via litterfall than temperate forests (Alexandre et al, 1997;Bartoli, 1983;Lucas et al, 1993;Nakamura, Nakagawa,, & Kitajima, 2020b). All these ecosystem-level studies in the tropics are from old growth forests, and little is known about Si flux via litterfall in post-logging secondary forests which represent the majority of tropical forests in recent decades (Chazdon, 2014).…”