Please cite this article in press as: Smith, A.C., et al., Modelling above- and below-ground mass loss and N dynamics in wooden dowels (LIDET) placed across North and Central America biomes at the decadal time scale. Ecol. Model. (2010), doi:10.1016/j.ecolmodel.2010.09.018 ARTICLE IN PRESS G Model ECOMOD-5977; No. of Pages 15 Ecological Modelling xxx (2010) xxx–xxx Contents lists available at ScienceDirect Ecological Modelling journal homepage: www.elsevier.com/locate/ecolmodel Modelling above- and below-ground mass loss and N dynamics in wooden dowels (LIDET) placed across North and Central America biomes at the decadal time scale Amanda C. Smith a , Jagtar S. Bhatti b , Chen Hua c , Mark E. Harmon d , Paul A. Arp a,∗ a Faculty of Forestry and Environmental Management, University of New Brunswick, 28 Dineen Drive, PO Box 44555, Fredericton, New Brunswick E3B 6C2, Canada b Northern Forestry Centre, 5320 - 122nd Street, Edmonton, Alberta, T6H 3S5 Canada c Biology Department,University of Illinois-Springfield, Springfield, IL 62703H, USA d Department of Forest Ecosystems and Society, Oregon State University, Corvallis, OR 97331-5752, USA article info Article history: Received 5 March 2010 Received in revised form 1 September 2010 Accepted 14 September 2010 Available online xxx Keywords: Mass Nitrogen Wooden dowels Tropical Temperate Boreal forests Grasslands Wetlands Tundra abstract This article focuses on modelling above and below-ground mass loss and nitrogen (N) dynamics based on the wooden dowels (Gonystylus bancanus [Miq.] Kurz) of the decadal Long-term Intersite Decomposition Experiment (LIDET) data. These dowels were placed at 27 locations across North and Central America, involving tropical, temperate and boreal forests, grasslands, wetlands and the tundra. The dowel, inserted vertically into the soil with one half remaining exposed to the air, revealed fast mass and N losses under warm to humid conditions, and slow losses under wet as well as cold to dry conditions. The model formu- lation, referred to as the Wood Decomposition Model, or WDM, related these losses to (i) mean annual precipitation, mean monthly January and July air temperatures, and (ii) mean annual actual evapotran- spiration (AET) at each location. The resulting calibrations conformed well to the time-in-field averages for mass remaining by location: R 2 = 0.83 and 0.90 for the lower and upper parts, respectively. These val- ues dropped, respectively, to 0.41 and 0.55 for the N concentrations, and to 0.28 and 0.43 for N remaining. These reductions likely refer to error propagation and to as yet unresolved variations in N transference into and out of the wood specific to each individual dowel location. Recalibrating the model parameters by ecosystem type reduced the R 2 values for actual versus best-fitted mass loss by about 0.15. Doing the same without location- or ecosystem-specific adjustments reduced the R 2 values further, by about 0.3. © 2010 Published by Elsevier B.V. 1. Introduction Predicting the rate at which wood decays and mineralizes is important for assessing past, current and future ecosystem-level carbon (C) and nitrogen (N) responses under varying and changing climate conditions (Laiho and Prescott, 2004). Quantifying these processes, however, is a complex task because of their depen- dence on wood type, size, shape, density, lignin content, presence of wood preservatives, configuration of placement, wood-consuming organisms at work, and antecedent conditions (Harmon et al., 1995; Stevens, 1997). For example, woody debris that remains dry min- eralizes fairly slowly. In contrast, wood that remains moist decays more quickly by providing optimal conditions for the entry and growth of decay-causing organisms such as fungi, bacteria, insects ∗ Corresponding author. Tel.: +1 506 453 4931; fax: +1 506 453 3538. E-mail addresses: Jagtar.Bhatti@NRCan-RNCan.gc.ca (J.S. Bhatti), chen40@uis.edu (C. Hua), Mark.Harmon@oregonstate.edu (M.E. Harmon), arp1@unb.ca, arp1@unb.ca (P.A. Arp). and wood dwellers. Wood placed into the ground may decay even more quickly than wood resting on the ground, depending on differ- ences in moisture content and the physical, chemical and biological conditions of the adjacent soil (Busse, 1994; van der Wal et al., 2007). With regard to N, decaying wood has low N concentrations prior to decay (Hungate, 1940). Hence, transference of exogenous N from adjacent soil and decaying litter is likely to occur on account of physico-chemical processes such as diffusion from N-enriched soil solution into wood and biological processes such as N 2 fixation, and transfer of exogenous N and other nutrients into the wood via invading organisms, especially fungal mycelia (Becker, 1971; Ausmus, 1977; Freya et al., 2003). Ecologically, decaying wood may therefore provide temporary storage for N and other nutrients for later use (Boddy and Watkinson, 1995; Pyle and Brown, 1999). To gain insight into the overall mass and N dynamics in decay- ing wood, recent forest litter studies dealing with forest litter decay across widely ranging site and climate conditions have also pro- duced data for wood decay. Among these studies are: the Long-term Intersite Decomposition Experiment in the United States (LIDET, 1995; Parton et al., 2007; Adair et al., 2008), the Decomposition 0304-3800/$ – see front matter © 2010 Published by Elsevier B.V. doi:10.1016/j.ecolmodel.2010.09.018