Mapping physiology: biophysical mechanisms define scales of climate change impacts

Abstract

The rocky intertidal zone is a highly dynamic and thermally variable ecosystem, where the combined influences of solar radiation, air temperature and topography can lead to differences greater than 15°C over the scale of centimetres during aerial exposure at low tide. For most intertidal organisms this small-scale heterogeneity in microclimates can have enormous influences on survival and physiological performance. However, the potential ecological importance of environmental heterogeneity in determining ecological responses to climate change remains poorly understood. We present a novel framework for generating spatially explicit models of microclimate heterogeneity and patterns of thermal physiology among interacting organisms. We used drone photogrammetry to create a topographic map (digital elevation model) at a resolution of 2 × 2 cm from an intertidal site in Massachusetts, which was then fed into to a model of incident solar radiation based on sky view factor and solar position. These data were in turn used to drive a heat budget model that estimated hourly surface temperatures over the course of a year (2017). Body temperature layers were then converted to thermal performance layers for organisms, using thermal performance curves, creating 'physiological landscapes' that display spatially and temporally explicit patterns of 'microrefugia'. Our framework shows how non-linear interactions between these layers lead to predictions about organismal performance and survivorship that are distinct from those made using any individual layer (e.g. topography, temperature) alone. We propose a new metric for quantifying the 'thermal roughness' of a site (RqT, the root mean square of spatial deviations in temperature), which can be used to quantify spatial and temporal variability in temperature and performance at the site level. These methods facilitate an exploration of the role of micro-topographic variability in driving organismal vulnerability to environmental change using both spatially explicit and frequency-based approaches.

Keywords: Biomechanics; ecophysiology; environmental heterogeneity; intertidal zone; microclimate; microhabitat.

Figure 1

Systematic workflow of the integrative framework to analyse thermal landscapes. Following the framework steps: (i) create DEM through (ia) image acquisition with sUAS and (ib) image processing through photogrammetry. (ii) Estimate surface temperatures with heat budget model. Finally, (iii) estimate relative performance with thermal performance curve.

Figure 2

Fort Beach in Marblehead, MA USA (42.508° N, 70.843° W).

Figure 3

DEM of Fort Beach. Colours represent elevation across the model, with max elevation at 4.3 m above mean lower low water. Rq, the surface roughness, is calculated to be 0.96 m at Fort Beach. Note that blank spaces represent coverage by water at low tide.

Figure 4

SVF of Fort Beach. SVF is calculated by using 32 search vectors per pixel and a maximum search radius of 5 m per vector. SVF ranges from 0 to 1 for each pixel of the map, 0 being 100% shaded and 1 being 100% exposed to the sky. Note that blank spaces represent coverage by water at low tide.

Figure 5

Surface temperature models at Fort Beach. Figure presents the low-tide period during the hottest day of 2017 (22 June) from (A) 14:00 EDT to (F) 19:00 EDT (solar noon at this location is at 13:00 EDT). Surface temperature distribution is highly variable through space and time. Note that blank spaces represent coverage by water at high tide.

Figure 6

A typical TPC for intertidal organisms. TPC is divided into different performance categories to identify the levels of thermal stress. This TPC has its sub-optimal range from 0°C to 21.2°C; optimal range is from 21.2°C to 30.1°C with optimal temperature (T_opt) at 28°C; sub-lethal range from 30.1°C to 35°C and finally high lethal temperature is >35°C.

Figure 7

Relative performance models at Fort Beach. The figure describes variation in performance distribution spatially as well as throughout a low-tide period. Performances are categorized into sections for sub-optimal (green), optimal (blue), sub-lethal (orange) and lethal (red). Colours correspond to the categories in Fig. 6, and blank spaces correspond to water at high tide.

Figure 8

Frequency distribution of surface temperature across the six timestamps on 22 June 2017 shown in Fig. 5. Surface temperature distribution range from 22°C to 41°C (from left to right at each panel). Panels describe variations in the peaks and shape at each timestamp. The patterns of surface temperature distribution can be quantify through thermal roughness (RqT), where the higher RqT means higher distribution of temperatures across the site.

Figure 9

Frequency distribution of relative performance across the six timestamps on 22 June 2017 shown in Fig. 7. Relative performance distribution range from 0 to 1 (from left to right at each panel). Panels describe variations in the peaks and shape at each timestamp. The patterns of performance distribution can be quantify through performance roughness (RqP), where the higher RqP means higher distribution of performances across the site.

Figure 10

Surface temperature distribution on 22 June 2017 at 16:00 EDT. This figure compares the total area (greyscale) during low tide to micro-refugia area shown in black (in this scenario, micro-refugia are defined as sites where surface temperature < 28°C). This figure shows how spatial models can reveal the scarcity of micro-refugia at the rocky shores during a heat event.

Figure 11

Changes in optimal performance areas within 1 h on 23 June 2017. Blue–brown represents optimal performance areas at 14:00 EDT, while blue represents optimal performance areas at 15:00 EDT. Orange represents areas of sub-lethal or lethal performance. Relative performance is based on an intertidal organism TPC with a T_opt of 28°C. Temporal comparisons of relative performance can be an opportunity to describe thermal behaviour for mobile species. The difference of optimal performance area between timestamps represents potential thermal corridors for mobile species.

Figure 12

Habitat structure based on the relative performance of interacting species. Blue represents a prey species with a TPC of 28°C; purple represents a predator or dominant species with a TPC of 25°C. Species blue is bounded by not only the habitat of the predator but also its own upper thermal limits, shown in orange.