Patterns of biological nitrogen fixation during 60,000 years of forest development on volcanic soils from south-central Chile

Publicado en New Zealand Journal of Ecology, v. 38(2): 189- 200
Autores

Pérez, C.A., Thomas, F., Silva, W.A., Segura, B., Gallardo, B. and Armesto, J.J.

Año de publicación 2014
DOI N/A
Afiliaciones
  • Instituto de Ecología y Biodiversidad, Las Palmeras 3425, Santiago, Chile
  • Geobotany, Faculty of Regional and Environmental Sciences, Behringstra&betae 21, Universität Trier, 54296 Trier, Germany
  • Facultad de Ciencias Forestales y de Conservación de la Naturaleza, Universidad de Chile, Santa Rosa 11315, Santiago, Chile
  • Departamento de Ecología, Pontificia Universidad Católica de Chile, Alameda 340, Santiago, Chile

 

Programa

CRN3

Proyecto CRN3035
Keywords

Abstract

Biological nitrogen fixation (BNF) is a key process for ecosystem development on new substrates. On young volcanic substrates, the near absence of nitrogen (N) and the presence of available phosphorus (P) in the soil should stimulate the activity of diazotrophic, N-fixing, bacteria. Our main hypothesis is that ecosystem N gain through BNF is tightly coupled to the development of progressive and maximum phases of ecosystem succession, as element contents build up. We investigated ecosystem development in a well-established 60 000-year-old chronosequence in Llaima volcano, south-central Chile. Tree basal areas and total carbon (C), N and P contents in soils showed declining trends over the millennial chronosequence following maximum values between 326 and 3470 years of succession. Maximum rates of BNF (3&ndash6 kg N ha&ndash1 year&ndash1) were recorded in the interval from 776 to 3470 years of substrate age, partly associated with high soil C and P contents, and with the lowest denitrification rates in the chronosequence (0.17&ndash0.42 kg N ha&ndash1 year&ndash1). Accordingly, over this time interval, there was a positive balance of gaseous N fluxes (5.7&ndash2.8 kg N ha&ndash1 year&ndash1), which produced the lowest 15N signal in the surface soil (&delta15N = &minus4.6 &permil). In turn, the first stage of the chronosequence was characterised by low symbiotic N fixation associated with Racomitrium moss carpets (0.07 kg N ha&ndash1 year&ndash1) that did not compensate for denitrification losses (2.6 kg N ha&ndash1 year&ndash1), yielding a negative balance of N gas exchanges. At 3470 years in the chronosequence, BNF rates declined (1.04&ndash2.3 kg N ha&ndash1 year&ndash1) and denitrification increased (1.2&ndash2.5 kg N ha&ndash1 year&ndash1), leading to N losses and enriching the &delta15N signal in the surface soil (&minus2.4 &permil). Finally, the oldest sites of the chronosequence presented BNF rates that balanced denitrification rates (1.3 kg N ha&ndash1 year&ndash1), thus supporting the hypothesis that a net nitrogen budget of zero for N gaseous fluxes characterises the retrogressive phase in this chronosequence.