Southern Forests 2017: 1–15 Printed in South Africa — All rights reserved Copyright © NISC (Pty) Ltd SOUTHERN FORESTS ISSN 2070-2620 EISSN 2070-2639 http://dx.doi.org/10.2989/20702620.2016.1254914 This is the fnal version of the article that is published ahead of the print and online issue Survival and long-term growth of eucalypts on semi-arid sites in a Mediterranean climate, South Africa Ben du Toit*, Gideon F Malherbe, Anton Kunneke, Thomas Seifert and C Brand Wessels Department of Forest and Wood Science, Stellenbosch University, Stellenbosch, South Africa * Corresponding author, e-mail: ben@sun.ac.za Four experiments were established on the semi-arid west coast plain of South Africa during the 1990s. The trails tested the survival and growth of several eucalypt species and hybrids, some of which were established in a climate that is drier than their natural distribution range. The aridity indices (AI; defined as mean annual precipitation [MAP]/mean annual potential evapotranspiration) ranged from 0.21 to 0.36 and MAP from 228 to 423 mm. The driest trial site (AI = 0.21 and MAP = 228) had high levels of mortality. However, a number of species (in particular, Eucalyptus gomphocephala, E. camaldulensis and E. tereticornis, as well as individual hybrids of the latter two species with E. grandis) survived and grew well at the remaining sites. Eucalyptus cladocalyx survived well and attained competitive growth rates only on the wettest site in the group (AI = 0.36). The dominant height of the top-performing genotypes at age 5 ranged between 9 and 10 m on the two wetter sites. This corresponded to mean annual increment values in excess of 10 m 3 ha −1 a −1 , which is comparable to volume obtained at more favourable aridity indices in the summer rainfall zone of South Africa and exceeds the growth rates obtained in several other arid zone studies globally. The E. grandis × E. camaldulensis hybrid ranked among the top performers in two trials, but its susceptibility to recently introduced pests and relatively poor wood quality makes it a less attractive choice for planting. The high density and durability of timber, acceptable growth rate (given the low rainfall conditions), and low pest and disease incidence make E. gomphocephala and E. cladocalyx the species of choice for planting in the drier and relatively wetter sections of the semi-arid zone, respectively. Keywords: agroforestry, dryland forestry, Eucalyptus, Western Cape coastal plain Introduction Southern Forests is co-published by NISC (Pty) Ltd and Taylor & Francis South Africa is a country that relies on plantation forestry for the sustainable supply of timber to several downstream industries: construction and building, pulp and paper, furniture, pole treatment plants and composite board manufacturing. Planted forests also make substantial contributions to timber used on farms, for example utility timber, fencing, and biomass for firewood and energy. With only 0.5% of the country’s surface area being covered in closed-canopy indigenous forests, coupled with limited utilisation potential due to the moderately slow growth rates of even the faster-growing indigenous species in southern Africa (de Cauwer et al. 2017; Gush 2017), the government decided to start a plantation forestry industry at the end of the eighteenth century. The forest industry grew rapidly up to a point where some excess logs were exported by the end of the nineteenth century. The oversupply situation was, however, short lived and already by 2005 an increasing shortage of timber was predicted for the future (DWAF 2005). This shortage of timber became more of a concern in the Western Cape after the announcement by government that the forestry industry in the province will be reduced substantially by means of an exit strategy over a period extending up until 2020, whereby traditional plantation areas would be suspended from production in favour of conservation land uses (VECON 2006). There is currently a drive to find alternative, renewable energy sources to reduce South Africa’s reliance on fossil fuel burning for energy (DME 2003). In addition, the Working for Water Programme (Binns et al. 2001) is busy eradicating large areas of invasive trees (mainly Australian acacias), a resource that has been extensively used as firewood in peri-urban areas of the Western Cape province (du Toit et al. 2010a). There is thus also a newly developing need for an alternative (non-invasive) form of woody biomass in the province, for use in peri-urban areas (du Toit et al. 2010a). The reduction in suitable forestry land has resulted in the need to explore the potential to establish plantations or woodlots on non-traditional forestry sites, to meet future woody biomass needs (Seifert et al. 2016; Gush 2017). An initiative to investigate the feasibility of potential afforesta- tion of low rainfall areas (<450 mm) along the West Coast region of the Western Cape was started in the early 1990s (van Wyk et al. 2001). The project was dubbed DIRAP § This article is based on a paper presented at the Symposium on Silviculture and Management of Dryland Forests, Stellenbosch University, South Africa, 16–19 March 2015, jointly organised by IUFRO unit 1.02.05 and the Department of Forest and Wood Science, Stellenbosch University