14Landscape structure and fragmentation have important effects on ecosystem 15 services, with a common assumption that fragmentation reduces service 16 provision. This is based on fragmentation's expected effects on ecosystem 17 service supply, but ignores how fragmentation influences the flow of services to 18 people. Here, we develop a new conceptual framework that explicitly considers 19 the links between landscape fragmentation, the supply of services, and the flow 20 of services to people. We argue that fragmentation's effects on ecosystem service 21 flow can actually be positive or negative and use our framework to construct 22 testable hypotheses about the effects of fragmentation on final ecosystem service 23 provision. Empirical efforts to apply and test this framework are critical to 24 improve landscape management for multiple ecosystem services. Humans continue to heavily modify natural ecosystems around the world, 31 with negative consequences for biodiversity (see Glossary) and natural capital 32 [1,2]. At the same time, demand for ecosystems to provide benefits, or services, 33 to society is growing rapidly [3]. This has significantly increased the need to 34 understand and manage landscapes simultaneously for ecosystem services and 35 biodiversity. Recently, the potential of managing landscape structure [4][5][6], and 36 in particular landscape fragmentation [7,8], for these multiple goals has been 37 highlighted. Interest in landscape fragmentation -the breaking apart of areas of 38 natural land cover into smaller pieces independent of a change in the amount of 39 natural land cover -has a long history in ecology [9]. Consequently, a well-40 developed understanding exists of its effects on biodiversity and ecosystem 41 functioning [10]. However, the shift in research interest from biodiversity 42 towards the concept of ecosystem services has recast what before were solely 43 ecological questions into social-ecological ones [11][12][13]. This recasting means 44 that predictions about the ecological effects of landscape fragmentation on 45 biodiversity and ecosystem functioning are unlikely to translate directly into 46 ecosystem service provision. This will be especially true if fragmentation has 47 contrasting effects on people and how they interact with ecosystems to produce 48 ecosystem services compared to biodiversity and ecosystem functioning. It is 49 therefore critical to rethink how fragmentation alters all of the components of 50 ecosystem service provision in order to improve landscape management for 51 multiple services. 52 4 Ecosystem service provision depends on three elements: supply, demand, 53 and flow (Figure 1), each of which can respond differently to landscape 54 fragmentation. Ecosystem service supply is the potential for natural capital to 55 generate a benefit for people, irrespective of it being realized or used [14]. In 56 turn, ecosystem service demand is the level of service provision desired or 57 required by people, and is influenced by human needs, values, ...