Current distributions of plant species have been shaped by past periods of climate change (Hewitt 2000, Kreyling et al. 2012). However, present rapid climate and land use changes threat biodiversity and impact natural and anthropogenic systems as well, in all continents (IPCC. PCC, 2019), leading to large-scale shifts in species’ ranges (Bardou et al. 2020). This is particularly relevant in drylands that cover c. 40% of the land global surface which deliver significant ecosystem services. Drylands are considered one of the most sensitive areas prone to suffer the effects of climate and land cover changes due to increasing hydro-climatic variance that will characterize forecasted 21st century warming (Huang et al. 2015). Global expansion of drylands has occurred over the past 60 years (Feng and Fu 2013, Huang et al. 2017a) and it is projected to accelerate (Huang et al. 2016a). In addition, approximately one third of global population, more than 2 billion people, inhabit this environment (Safriel et al. 2005). Thus, the expanded drylands together with extensive land use may aggravate the risk of land degradation and ecosystem vulnerability (Huang et al. 2017b). Therefore, estimates on projected range shifts of drylands are crucial to avoid global desertification intensification (Reynolds 2011).
Drylands include hyper-arid, arid, semiarid, and dry sub humid areas (Feng and Fu 2013), where temperature and precipitation forecasts vary with latitude and geographic region. It has been shown that temperate drylands might be reduced by a third and converted to subtropical drylands (Schlaepfer et al. 2017). Also, major expansions of semiarid regions would occur over the north side of the Mediterranean, southern Africa, and North and South America (Feng and Fu 2013). Nonetheless, early temperature and precipitation projections based on 20th century data indicate that while warming characterized most dry regions, the increase rate is less than the average for global drylands in the Sahel at the Horn of Africa and Southern South America (Hulme 1996). In contrast to the majority of dryland regions of the world, where no significant wetting or drying was measured, the estimations predict an increase in precipitation of approximately 18% for southern South American drylands (Hulme 1996, CONAF- CONAMA 2006). Similarly, hydroclimatological conditions, for the region, using local climate models for two-time series 1961–1990 and 2071–2100 indicated that dry zones of Argentina yielded wetter trends associated to greater water availability and an aridity index decrease, particularly towards temperate regions (Zaninelli et al 2019). Thus, different responses might be expected, particularly for widespread dryland species spanning through tropical, subtropical, and temperate environments (Vidal R. et al 2022).
Under the rapidly increasing demand for food due to population growth, urbanization, and increasing incomes in developing countries, efforts are placed on production of livestock in arid rangelands not naturally suited for extensive agricultural practices (Delgado 2005, Tadey and Souto 2016). These detrimental effects are deepened in regions with high endemism rates beholding species highly adapted to specific climatic conditions (Olson et al. 2001, Sánchez-Azofeifa et al. 2005, Miles et al. 2006, Souto et al. 2015). Developing evidence-based management strategies that support the adaptive capacity of species is crucial under predicted climate change, which is leading to increased warming and summer droughts (Inouye, 2000; Hartmann, 2011), with potentially large repercussions for ecosystems productivity (Ciais et al., 2005; Gazol et al., 2018). Here we focus on improving the understanding of how dryland species responded to past and will do so under projected future climatic conditions to have a hint of potential responses of dominant Larrea species inhabiting Monte Desert, the largest continuous arid region of South America.
Ecological niche models (ENM) consist of the projection of environmental niches in the landscape. They are used to describe environmental tolerances of species and allow projecting potential responses of their ranges to past and future climate scenarios, estimating the favorability of the habitat (Peterson and Soberon, 2012). Environmental niche models can also predict the suitability of habitats across a landscape, and are increasingly been used to design conservation and restoration plans (Soberon et al 2017). A growing number of studies compare predicted habitat suitability generated by ENMs from different species to test hypotheses on historical processes that might be responsible for current distributions and also to forecast evolutionary responses of species under predicted future climates (Soberon et al 2017). Species distribution models (SDMs) and ecological niche models are constructed by correlating observations of occurrence with data on environmental conditions at occupied sites (Bradie and Leung 2017). Increasingly, conservation biologists are using species distribution models (SDMs) to assess the environmental parameters that shape species’ distributions today and project these species–environment relationships into past and future conditions (Phillips et al. 2006; Elith and Leathwick 2009). In this sense temperate deserts in the Global South are being highly degraded and information is needed on future responses of key taxa. Particularly, temperate and semi-arid regions of Southern South America are undergoing rapid habitat conversion as a result of several human activities (i.e. grazing, logging, agriculture). These arid ecosystems contain many endemic species and have played an important role in the evolution of South American biota (Ojeda et al 1998).
In arid systems, the majority of studies have focused on predicting the effects of climate change on native species of the northern hemisphere rather than such impacts or landscape change on South American ones (Vidal R et al 2021). Many of these species lack comprehensive occurrence data, and the abiotic and biotic limits to species distributions are poorly understood (Lomolino and Heaney 2004). This information gap on the geographical distribution of many taxa, known as Wallacean shortfall (Whittaker et al 2005), can be addressed using SDMs to develop range and habitat suitability maps, which can be generated with as few as 5–25 occurrence points without knowledge of absence points (Hernandez et al. 2006; Pearson et al. 2006). Thus, it is of interest to fill the knowledge of understudied regions, which might contain key taxa vulnerable to changes in climate. Given the forecasted climatic change it is useful to compare the current distributions of species to assess their ability to cope with changes in climate (Sheller and Leon 2016) such as under the Last Glacial Maximum (LGM) and the Holocene, when agriculture and pastoralism emerged and spread (Boivin et al., 2016; Hofman and Rick, 2018). Although manipulative field- and laboratory‐based studies can identify factors that shape niches and ranges (Hargreaves et al. 2014, Lee‐Yaw et al. 2016), for most species logistical difficulties preclude examining large numbers of variables and studying range limits across broad geographic scales. So, alternatively, environmental limits can be inferred using models of species’ geographic ranges and niches.
Monte Desert, is a subtropical and temperate arid region that covers approximately 467,000 km2 (Oyarzabal, 2018). As other temperate drylands, it is an important biodiversity hotspot characterized by intermediate to high levels of species richness and endemism (Olson et al. 2001, Sánchez-Azofeifa et al. 2005, Miles et al. 2006), probably due to its particular climatic setting (Turchetto-Zolet et al. 2013, Camps et al. 2018). In late Miocene and Pliocene, the geology and climate of the region was modeled by earlier volcanic and tectonic events related with the elevation of the Andean mountain range (Ramos and Ghiglione 2008). Also, paleoclimatic records and model simulations of past climates suggest significant variations in the atmospheric circulation, temperature and precipitation patterns since the Last Glacial Maximum (LGM) (Labraga and Villalba 2009). Linked to the LGM, this region was affected by dry and cold climates that facilitated the advance of xerophytic vegetation of northwestern Argentina even northwards (Ab’Saber 2000, Iriondo, 2010). Nonetheless, the late Holocene was characterized by a less severe dry period (Argollo Bautista and Iriondo, 2008, Iriondo, 2010). Models of climate change for 2100 forecast contrasting seasons, with increasing summer rainfall, but almost no changes or small reductions in winter precipitations across temperate drylands and temperature increases, larger in summer than in winter (CONAF-CONAMA 2006, Labraga and Villalba 2008). Thus, ongoing climate change highlights the need to understand vegetation responses to climate oscillations particularly in regions with projected changes different from the increasing temperature-decreasing precipitation standard global pattern (IPCC. PCC, 2019).
Unfortunately, agricultural expansion and overgrazing exploitation of the region affects not only the dominant shrubs species highly consumed by livestock, such as Larrea sp. (Souto and Tadey, 2018). Also, have severe and accelerated consequences that might scale globally, as deepening South America`s most dramatic biodiversity decline, reducing global green surface and natural cover (Hansen et al. 2013, WWF 2014, Vallejos et al. 2015). Studying multiple species' populations, within and beyond current ranges, can help identify whether demographic responses to climate change exhibit directionality, indicative of range shifts, and whether responses are uniform across a suite of species.
Hereby, we modeled the suitability of three species of the genus Larrea, a closely related group of species dominant in arid South American native rangelands, threatened not only by climate but also by land use change. We deployed models for the present (1975–2016), and then projected them into two scenarios in the past (LGM and Mid Holocene) and eight future climatic scenarios (2050 and 2070, under two scenarios of greenhouse gas emissions and two different global circulation models); to track changes in distribution through time. We aim to estimate current habitat suitability using spatial distribution modelling to retrodict the effects of historical climatic changes that occurred during the Quaternary glaciations and Mid Holocene, when anthropogenic disturbance started, on the spatial distribution of Larrea sp., and to predict species' range suitability under future climates. We hypothesize increasing habitat suitability during glacial periods for more cold-tolerant Larrea species. Milder Mid Holocene temperatures favored suitability of all studied species, whereas increasing precipitation and temperature seasonality would differentially affect Larrea species according to their climatic tolerances. Drylands are highly threatened, in unstable equilibrium ecosystems, which have turned into a matter of great concern due to growing land degradation and productivity-loss, lately combined with climate change (Huang et al 2017). Despite their known importance as alternative forage and prevalence in the largest desert of South America, Larrea species are not fully understood, particularly in their ecological niches and how their sustainability would change with projected climates. We fill this gap assessing the spatial distribution and niche overlap of three native Larrea species under present climate conditions to retrodict their potential distribution in the past and anticipate them under future climate scenarios, in the end assisting biodiversity conservation and sustainability.