197 Abstract Wide-spaced trees on pastoral land (pasture-tree (PT) systems) are a widespread feature of many farmed landscapes. They offer the potential to increase carbon (C) storage, with implications for reducing atmospheric CO 2 -C. The effect of PT systems on soil C stocks to 1 m depth was determined for trees aged 14-16 years at densities of 73-111 stems per hectare at four North Island sites (two with poplar, two with alder). Across sites, mean soil C concentration was 1.9-8.5% and mean total soil C mass was 120-455 tonnes C/ha. For alder systems, total C mass of PT was 37% less than adjacent pasture (Open) at Poukawa (120 versus 189 tonnes C/ha), whereas at Ruakura, there was no significant difference between systems. Total C mass of PT systems involving poplar did not vary significantly from adjacent Open systems at Tikokino (328 versus 352 tonnes C/ha) and Woodville (154 versus 202 tonnes C/ha). Soil at 0.3-1.0 m depth comprised up to half of total C mass. Results suggested that poplar and alder had different effects on soil C. Keywords: pastoral hill country, wide-spaced trees, carbon sequestration, greenhouse gas (GHG) mitigation Key messages  Soil C mass (0-1 m depth) of pasture-tree (PT) systems involving poplar and alder was not significantly different from that of adjacent open pasture, except at one of two sites with alder where the PT system had 37% less soil C than open pasture.  Up to 50% of the C mass at 0-1 m soil depth resides deeper than 0.3 m. Introduction The carbon (C) in soil organic matter is a major component within the global C cycle (Eswaran et al. 1995; Lal 2004). Comparative estimates of organic C contained in living biomass (560 Pg) and atmospheric CO 2 -C (760 Pg) indicate that a small shift in the soil organic C pool has the potential to have a large impact on atmospheric concentrations (Lal 2004). This represents a potential greenhouse gas (GHG) mitigation option for key primary industries in New Zealand. On pastoral hill country, wide-spaced trees (pasture- tree (PT) systems) have been a prominent feature for 50+ years (Wilkinson 1999; McIvor et al. 2011). The primary purpose of the trees is to reduce the occurrence of mass movement erosion processes such as shallow landslides (Douglas et al. 2013) through altering soil strength and hydrological patterns. The tree species used are predominantly poplar (Populus spp.) and willow (Salix spp.) (McIvor et al. 2011), with other species used including wattle (Acacia spp.), gum (Eucalyptus spp.) and alder (Alnus spp.) (Van Kraayenoord & Hathaway 1986). Mature tree density in PT systems is usually less than 50 stems per hectare (sph), but can vary considerably with risk and type of erosion, tree survival, and the extent of thinning if high densities were planted. There has been increasing interest in the potential of trees at varying densities to sequester C in above- and below-ground biomass to counter increasing atmospheric concentrations of GHGs (Dube et al. 2012; PØrez- Cruzado et al. 2012; Baah-Acheamfour et al. 2014). In the few studies that have determined soil C mass in PT systems, sampling has been to soil depths less than 300 mm. In a Manawatu poplar-pasture system, maximum soil organic C content (0-75 mm depth) was <35 tonnes C/ha and C content was similar or less in PT than in paired Open pasture plots (Guevara-Escobar et al. 2002). In contrast, the organic C of soil (0-100 mm depth) averaged over 12 sites in central Alberta, Canada, was higher in a poplar silvopastoral system (81.3 g/kg) than in adjacent open pasture (53.1 g/kg) (Baah-Acheamfour et al. 2014). Differences in soil C in PT compared with Open systems on similar topography could be because of factors such as tree species/clone, stand age, tree density, grazing management, cultivation history and soil type. Trees modify their local environment (Douglas et al. 2006a; Douglas et al. 2006b), impacting on inputs into the C cycle of pasture litter and root biomass, and contribute leaf litter and root biomass of their own (Young 1997). There is little information for PT systems on how deep in the soil profile trees influence soil C stocks, if they alter total C mass, or simply shift the distribution of soil C down the profile. Trees also modify the physical, chemical and biological properties of soils (Guevara- Escobar et al. 2002; Power et al. 2003; Douglas et al. 2006b), which in turn influences C cycling. This study aimed to determine if PT systems comprising trees at densities similar to those found on a range of pastoral landscapes accumulate more soil C than Open pasture systems to 1 m depth. Furthermore, it Impacts of poplar and alder on soil carbon in pasture-tree systems G.B. DOUGLAS 1 , R.E. VIBART 1 , A.D. MACKAY 1 , M.B. DODD 1 and I.R. McIVOR 2 1 AgResearch Grasslands, Private Bag 11008, Palmerston North, New Zealand 2 Plant and Food Research, Private Bag 11030, Palmerston North, New Zealand grant.douglas@agresearch.co.nz ISSN 0110-8581 (Print) ISSN 2463-4751 (Online)