Spatio-temporal and cumulative effects of land use-land cover and climate change on two ecosystem services in the Colombian Andes

Highlights

• Cumulative climate change and LULC dynamics can have marked effects on ecosystem services.

• We studied such effects on water yield and carbon sequestration in two watersheds in Colombia.

• Climate scenarios had greater effect on water provision than land-use land-cover scenarios.

• Carbon sequestration was greater in forest and shrubland areas farther from Bogota.

• It is key to incorporate climate conditions in land planning for sustainable provision of services.

Abstract

Climate change can have marked effects on ecosystem service (ES) provision in the Andes, particularly in peri-urban areas. In addition to global-change related processes, cumulative effects such as changing socio-political dynamics, environmental policies, and conflicts are also changing type and magnitude of land use–land cover (LULC) dynamics in the Colombian Andes. Studies in the region have investigated the effects of LULC change, deforestation and extreme climatic events on the hydrology of watersheds and carbon sequestration. Yet, less is known on how the cumulative effects of climate and LULC changes will drive water yield and carbon sequestration. To investigate these cumulative effects, we study two different watersheds near Bogota, Colombia and their ES for the period 2016–2046. We use IPCC-LULC scenarios, expert elicitation, hydro-meteorological data, and integrated modelling using temporal LULC change and ESs valuation models to parse out effects of LULC versus climate change on two representative ESs. Our results show forest and shrublands remain stable during the analysis period. However, urban conversion of agricultural pastures is substantial. We found that climate change scenarios had greater effect on water yield and supply than LULC scenarios in both watersheds. However, carbon sequestration was greater in rural forest and shrubland areas farther from Bogota. In contrast to current land use zoning being promoted by local elected officials, our findings indicate that land-use development and policies in near-urban basins need to minimize urbanization in agriculture and pasture LULCs, as these can have substantial effects on water yield. Similarly, land use polices in ex-urban areas need to conserve forested and shrubland areas to maximize their carbon offset potential. Collectively, our results highlight the need to incorporate climate change conditions in decision making and land use planning processes, in order to maintain the capacity of ecosystems, both urban and rural, to provide services to society.

Introduction

Several anthropogenic threats affect ecosystems at multiple scales. Environmental change, the combination of multi-scale climate alteration and disturbances on natural and socio-ecosystems, cumulatively act to alter and degrade ecosystem properties. Among these affected properties are ecosystem composition, structure, and function, which support the capacity of ecosystems to provide ecosystems services that benefit humans (Grimm et al., 2013; Sharp et al., 2018). These changes occur at multiple spatial-temporal scales, including large-scale human induced changes in climate that can force large-scale changes that affect ecosystem function and distribution (Fu et al., 2017). Similarly, short-term, local-scale ecosystem disturbances related to land use and their interactions with climate change, can have the potential to affect ecosystem structure and function. For example, agricultural practices in the central Andes of Colombia affect hydrological (García-Leoz et al., 2018) and biogeochemical (Suescún et al., 2017) functions; and the interactions between land use and climate variability exacerbate these effects (Suescún et al., 2019). When considered at larger scales, forest ecosystems are often related to the maintenance of hydrological regulation (Mercado-Bettin et al., 2019) and atmospheric circulation in South America (Molina et al., 2019).

Other socio-economic and ecological factors and disturbances such as changing socio-political dynamics, globalization, land use and environmental policies, and the recent post-conflict process, are also changing both the type and magnitude of land use – land cover (LULC) dynamics in the Colombian Andes, particularly in peri-urban areas (Rubiano et al., 2017; Salazar et al., 2018). For example, in the past few decades agricultural expansion and changing socio-ecological dynamics have historically resulted in deforestation in several regions of the Colombian Andes (Armenteras et al., 2017). This growing agricultural front has resulted in vast deforestation and is mostly related to increased demand for crop and livestock grazing lands (Armenteras et al., 2013). Additionally, recent urbanization trends are characterized by a sizable rural migration to urban areas, such as Bogota and Medellin, the two largest cities in Colombia, and have resulted in significant land use – land cover changes (LUCC), particularly land abandonment with subsequent forest regeneration in these former rural and peri-urban agricultural areas (Mendoza and Etter, 2002). Land speculation, poorly defined land tenure regimes, illicit crops and mining are also major drivers of LUCC in Colombia, especially in regards to loss of forest cover and mercury related water pollution (Dávalos et al., 2011). The dynamics of these cumulative LULC changes also lead to biodiversity and habitat loss, as well as altered biogeochemical cycles (Nelson et al., 2010). Specifically, changes in forest biomass and soil quality can potentially alter carbon dynamics, evapotranspiration, as well as infiltration and water regulation functions (Sharp et al., 2018). As such, cumulative effects will directly affect the capacity of ecosystems to provide key services such as carbon storage and water regulation, which are fundamental in the tropical Andes, where a large portion of the population lives (Bradley et al., 2006; García-Leoz et al., 2018).

There is a large body of literature studying the effects of LULC and climate change on ecosystem structure and subsequent functions and services (Grimm et al., 2013; Fu et al., 2017). Most of these spatially-explicit studies have used a combination of functional models and scenarios to study the effects on forest ecosystem services in association with forest management (Bottalico et al., 2016), urbanization (Delphin et al., 2016) and combined climate-LULC effects (Fu et al., 2017). Scenarios are regularly used in these models and their output can inform management, policy and planning decisions, and climate change adaptation and mitigation strategies (Nelson et al., 2010; Liang et al., 2018a, Liang et al., 2018b). Geospatial methods can also be used to spatialize information related to such scenarios of forest change (Lagrosa et al., 2018; Liang et al., 2017a, Liang et al., 2017b), urbanization (Jiang et al., 2017) and climate change (Jorda-Capdevila et al., 2018).

Previous studies have investigated the effects of LUCC, deforestation, and extreme climatic events on the hydrology of Andean watersheds (Ochoa-Tocachi et al., 2016; Bonnesoeur et al., 2019). However, less known is how the combined effects of tropical forest loss, climate change and urbanization are driving water regulation and biomass in peri-urban Andean watersheds that are fundamental for the provision of key-ecosystem services. Understanding the effects of rapid socio-ecological changes on regulation ecosystem services in the Andes will provide key information for management and conservation goals, as well as the development of climate change related policies, strategies, and incentives. Nevertheless, in the Colombian context, up to date and accurate geospatial data are often unavailable, access to sites is difficult, and land use dynamics are rapidly changing (Rubiano et al., 2017; Etter et al., 2006).

In this study, we explore the cumulative effects of socio-ecological disturbances on two case study watersheds and their ecosystem services in the Colombian Andes. The first larger watershed includes a dynamic urban area with over 8 million inhabitants, while the second is a smaller more rural and peri-urban watershed. Both watershed as such typify land use dynamics in upper Andean areas. Our specific objectives are three-fold. First, we model LUCC in both watersheds during the period of 2016–2046. Second, we estimate changes in water yield and carbon storage, in both watersheds during this same period. Finally, we assess and discuss the overall cumulative effects that forest cover dynamics, climate change, and urbanization will have on these watersheds and the provision of these two key ecosystem services.