On the importance of getting fine‐scale temperature records near any surface

Abstract

The SoilTemp database will identify the microhabitats that best buffer the amplitude of warming. The temperature heterogeneity at spatial scales below the meter also requires attention. A worldwide database of temperatures near any surface is still lacking. This article is a Commentary on Lembrechts et al., 26, 6616-6629. Our climate is changing globally. The pressing need to identify the ecological consequences of these changes boosted the development of macroecological approaches as well as climatological tools (e.g., Worldclim climate data provider) that can be used by predictive models to anticipate future biological changes. These approaches revealed some complex trends in how biodiversity responds to global changes (Scheffers et al., 2016) and some of the mechanisms underlying these trends have been identified (Sunday et al., 2014). Nevertheless, the major pitfall remains that the actual climatic conditions experienced by organisms in their microhabitat and across their home range are largely ignored (Pincebourde, Murdock, Vickers, & Sears, 2016). The notion of microclimate is part of our daily lives and this concept has long been applied to natural systems to detail the abiotic conditions experienced by organisms in their microhabitat. For instance, an aphid lives at the temperature of the leaf it attacks (Pincebourde & Casas, 2019) and a mussel experiences a body temperature resulting from the complex interaction of multiple environmental variables in the rocky intertidal zone (Helmuth, 1998). In both cases, the temperature experienced in fine can deviate from atmospheric temperature, sometimes strongly, by >15°C (Pincebourde et al., 2016). Therefore, it is crucial to identify these deviations in temperature for any kind of microhabitat anywhere on Earth. Such a huge task inevitably calls for collaborative work to develop global databases on microclimates. In this context, Lembrechts et al. (2020) are initiating a new global database of near-surface temperatures, especially focusing on aboveground (up to 2 m) temperatures and belowground soil temperatures. Mounting such a database of soil temperatures across biomes, continents and microhabitats is the first step toward (a) the empirical quantification of the buffering property of surfaces (i.e., how much of atmospheric temperature variation is buffered at the different soil layers); and (b) an assessment of the amplitude of microclimatic change actually experienced by organisms (plants, animals, fungi, and microorganisms) living in or near the ground (i.e., how much of the global/regional warming is transferred into the soil microhabitat over long temporal scales). This initiative of Lembrechts et al. (2020) is extremely welcome for climate change biologists and even more so for ecologists. Lembrechts et al. (2020) propose several environmental factors driving the surface temperature deviation to be reported in the database, including horizontal (topography, land cover type, distance to water bodies) and vertical (vegetation characteristics, snow cover, soil properties) features that are operating at the local or landscape scale. This approach is reasonable, especially because most microclimatic loggers deployed around the world are operating at these spatial scales. Nevertheless, the SoilTemp database would gain in strength and pertinence by including the micro-scale as well. Local heterogeneity in surface temperatures is important for species maintenance via various processes including behavioral thermoregulation (Woods, Dillon, & Pincebourde, 2015). We exemplify this point based on unpublished data (from the CANOPEE project funded by Region Centre-Val de Loire, France) illustrating (near) surface temperature heterogeneity at or below the meter scale along both the vertical and horizontal gradients of a broadleaf temperate forest (Figure 1). Along the vertical gradient of forests, different surfaces from belowground to the canopy top show quite distinct temperature signatures, with deviations of up to 20°C between 15 cm deep in the soil and the air at the canopy surface (Figure 1a). This gradient illustrates the buffering effect of the forest understory (Zellweger et al., 2020). Canopy gaps that allow solar radiation to reach the ground can rapidly and markedly influence soil temperature while they have little effect on air temperature at 2 m (Figure 1a). Therefore, over short timeframes, the temperature experienced by organisms living in the soil versus at different heights in the canopy can differ as much as macroclimate temperature varies across 20° of latitude. In addition, soil surface temperature can be quite heterogeneous at the micro-scale over the horizontal layer. The temperature at the leaf litter surface in early spring, as measured with an infrared camera, can be quite variable, with a temperature range of about 16°C across a short distance of less than 1 m (Figure 1b). This variability results from the interaction of sun angle, leaf litter microtopography and shading effects by vegetative elements above the ground. This effect is extremely dynamic and depends on the sun's course in the sky. Soil surfaces can quickly reach high temperatures when receiving direct solar radiation, which influences the temperature within the soil at 5 cm belowground (Figure 1a,b). Two dataloggers only 40 cm apart can measure soil temperatures differing by as much as 4°C due to this effect. Obviously, deploying a single temperature logger within this thermal mosaic will sample only a fraction of the local temperature range, and therefore temporal series from logger networks are better candidates to estimate the degree of thermal heterogeneity at the micro-scale. Determining the amplitude of the buffering ability of near-ground surfaces requires high temporal resolution when measuring temperatures. Lembrechts et al. (2020) are judiciously calling for “time series spanning one month or more, with temperature measurements a maximum of 4 hours apart.” In the context of meta-analysis, getting series with a temporal resolution of a few hours will certainly complement the temperature data obtained at a finer temporal resolution, thereby enriching the analysis. Nevertheless, soil temperature recorded with a time step of >1 hr may not provide an accurate estimate of maximal temperatures experienced by some soil organisms (Figure 1b: the duration of the temporal series is 4 hr). Acute thermal events of about 1 hr matter for ectotherms as these events can increase an organism's thermal tolerance (heat hardening) or induce irreversible physiological damage (Ma, Ma, & Pincebourde, 2021). A high temporal resolution may be especially important for organisms with a weak dispersal ability such as Isotoma Collembola which disperse by only 5 cm/week (Ojala & Huhta, 2001) and which may not be able to escape from acute heat exposure. The SoilTemp database focuses on temperatures near soil surfaces, including air temperature 2 m aboveground and belowground temperatures (Lembrechts et al., 2020). We see this important call for data as a first step toward a more global near-surface temperature database. In the future, the call could be expanded to other types of interfaces with the atmosphere including plant organ temperature (e.g., leaf surfaces and plant inner tissues such as stems, woody tissues or galls) and the temperature of different layers in water bodies (ponds, rivers, wetlands). Indeed, the forest canopy, along with soil, is one of the key forest ‘compartments’ supporting tremendous forest biodiversity (e.g., Parmain et al., 2020) and aquatic thermal environments are still missing for global analysis (van Klink et al., 2020). Moreover, the SoilTemp database would also integrate near-surface temperatures for urban areas to analyze how the urban heat island influences temperature patterns near the ground in green areas. Urban areas can be seen as simulators of climate warming and urban surface temperature data may help to anticipate the thermal behavior of surfaces in natural areas in the future. Given the accumulating evidence that fine-scale microclimates can play a crucial role in the fate of organisms, strong collaborative work among multiple disciplines (including computer and atmospheric sciences, ecology and physiology) is likely to be needed to generate a global assessment of microclimate buffering. This empirical approach should be combined with mechanistic modeling (Maclean, 2020) in particular to generate microclimatic data for places and/or periods that were not covered by dataloggers. These advances in microclimate ecology will bring global change biology one major step ahead, notably by generating accurate and comprehensive predictive models available for ecologists.

Topics & Concepts