Stable isotope approaches and opportunities for improving plant conservation

Stable-isotope ratios in plant tissues can uncover patterns of physiological responses and stress and identify plants and communities that are most resilient to changing environments. Stable-isotope ratios integrate environmental signature of plant responses in current, recent, past and paleo environments. As a consequence, stable isotopes are a powerful tool for planning, implementing and monitoring the conservation of threatened and endangered plants and for forensic evidence of environmental change as experienced by plants.

Abstract

Successful conservation of threatened species and ecosystems in a rapidly changing world requires scientifically sound decision-making tools that are readily accessible to conservation practitioners. Physiological applications that examine how plants and animals interact with their environment are now widely used when planning, implementing and monitoring conservation. Among these tools, stable-isotope physiology is a potentially powerful, yet under-utilized cornerstone of current and future conservation efforts of threatened and endangered plants. We review the underlying concepts and theory of stable-isotope physiology and describe how stable-isotope applications can support plant conservation. We focus on stable isotopes of carbon, hydrogen, oxygen and nitrogen to address plant ecophysiological responses to changing environmental conditions across temporal scales from hours to centuries. We review examples from a broad range of plant taxa, life forms and habitats and provide specific examples where stable-isotope analysis can directly improve conservation, in part by helping identify resilient, locally adapted genotypes or populations. Our review aims to provide a guide for practitioners to easily access and evaluate the information that can be derived from stable-isotope signatures, their limitations and how stable isotopes can improve conservation efforts.

Introduction

Conservation physiology is defined as ‘the study of physiological responses of organisms to human alteration of the environment that might cause or contribute to population declines’ (Cooke et al., 2013). Conservation physiology goes beyond the description of empirical patterns and instead establishes underlying mechanisms for population declines. The field of conservation physiology is well established as an important tool to shape conservation priorities in a wide range of animal taxa in aquatic, marine and terrestrial ecosystems (Cooke et al., 2013). However, linkages between physiology and plant conservation are not as well established as they are in animal conservation, although emerging opportunities to guide plant conservation through physiology are becoming more widely recognized (Richards et al., 2003; van Kleunen, 2014; Madliger et al., 2018). With almost all plant taxa, fitness (i.e. growth, survival and fecundity) is largely governed by photosynthetic processes (e.g. nutrient transport, stomatal conductance, sunlight capture) and non-photosynthetic processes (e.g. respiration, water transport, phloem transport, starch storage, synthesis of secondary metabolites). A wide range of physiological tools have been used to quantify how photosynthetic and non-photosynthetic metabolic processes respond to environmental alteration. For example, sensors are now widely used to quantify rates of water transport from soils to leaves, photosynthesis and respiration (Madliger et al., 2018). Here we examine how the use of stable isotopes complements the use of sensors and other tools to expand the physiological toolbox for plant conservation.

Stable-isotope ratios or ‘signatures’ in plant tissues, metabolites and xylem water are integrators of physical, chemical and biological processes (Fry, 2006). They are also tracers that follow the fate and transformations of resources as they are acquired by plants and transported through various trophic levels within ecosystems (Dawson et al., 2002). Plant physiologists have used stable-isotope ratios to determine photosynthetic and respiratory pathways, monitor plant stress, identify the sources of water and nitrogen used by plants, estimate water and nitrogen use efficiencies and reconstruct past climates from plant tissues (Rundel et al., 1989; Ehleringer and Dawson, 1992; Erskine et al., 1998; McCarroll and Loader, 2004; Ribas-Carbo et al., 2005; Watling et al., 2006; Wasley et al., 2012; Cernusak, 2020). Stable isotopes can shed light on plant resource acquisition under multiple stressors and across a range of spatial and temporal scales. As a consequence, stable-isotope techniques are considered among the most important empirical tools in plant physiological ecology over the last forty years (Dawson et al., 2002).

In this overview, we focus on stable isotopes of hydrogen, oxygen, carbon and nitrogen as short-term and long-term indicators of plant physiological performance, and how these indicators add to the toolbox used for the conservation of plants and their ecosystems. A full description of the basic theory, methods and terminology is beyond our scope (see Dawson et al., 2002; Fry, 2006; Gannes et al., 1998; McCarroll and Loader, 2004; Peterson and Fry, 1987, for in-depth reviews of stable-isotope theory and utility), rather we discuss applications in which stable-isotope physiology can support conservation. We provide a general framework illustrating how stable-isotope signatures in plant tissues vary across short-, mid- and long-term temporal scales and how knowledge of the signatures can support conservation efforts (Table 1). Plant tissues and biochemicals analysed for stable-isotopes include photosynthetic tissues such as whole leaf in vascular plants and whole plant in non-vascular taxa; structural tissues such as tree-ring cellulose and cactus spines; water extracted from plant xylem tissues; plant compounds such as leaf sugars, lipids and waxes; and products of photosynthesis and respiration such as respired CO₂. This framework highlights how stable-isotope signatures can be useful ecological indicators of stressors that may ultimately produce ecological consequences (Table 1). Understanding these ecological indicators can aid in developing effective adaptive management strategies. As more stable-isotope datasets become available, and advances in isotope mass spectrometry continue to emerge, stable-isotope analyses and applications will continually become more accessible to conservation biologists and managers.

Table 1.

Examples of how stable-isotope ratios can be used in conservation physiology, across increasing temporal scales

Isotopes	Increasing temporal scales
δ2H & δ18O	Seasonal plant water source use Measurement: isotopic signatures in xylem water and soil water	Forensic evidence of illegal plant trade Measurement: isotopic signatures in plant tissues and precipitation	Plant provenances & isoscape mapping Measurement: isotopic signatures in plant tissues and precipitation
δ13C	Daily response to drought and heatwaves Measurement: δ13C signatures of leaf photosynthates	Seasonal legacy effects of drought and heatwaves Measurement: δ13C signatures of whole-leaf tissue or individual cactus spines	Inter-annual legacy effects of drought and heatwaves Measurement: δ13C of annual growth rings or down moss shoots and peat layers
δ13C & δ18O	Seasonal stomatal sensitivity to drought and heat stress Measurement: isotopic signatures of whole-leaf tissue or cactus spines	Inter-annual stomatal sensitivity to drought and heat waves Measurement: isotopic signatures of annual growth rings or cactus spines	Long-term climate reconstructions Measurement: isotopic signatures of annual growth rings, or cactus spines constructed over decades, or moss and peat
δ15N δ15N & δ18O	Plant N use in habitat with isotopically distinct N sources Measurement: isotopic signatures of whole-leaf tissues or xylem water and N sources (NOy, NHx in air, soil, water)	Nitrogen cycle Measurement: isotopic signatures of plant tissues (leaves, wood) and N sources	Historic trends of N sources and cycling, paleo-environment reconstruction Measurement: isotopic signatures of plant tissues including recent, herbarium and fossils

To highlight this point, we have identified 18 questions from the list of 100 research questions in conservation physiology recently identified by Cooke et al. (2021) in which stable-isotope applications can support conservation practitioners (Table 2). Specifically, we have identified at least 17 questions that can be addressed through δ¹³C analysis of plant tissues, at least 6 questions that can be addressed through analysis of δ²H and δ¹⁸O in plant/soil/atmospheric water, 5 questions that can be addressed through δ¹⁵N tissue analysis and 3 that can be addressed through analysis of δ¹⁸O in tissues such as tree-ring cellulose (Table 2). Indeed, the list in Table 2 should not be considered exhaustive as stable isotopic tools have potential to address further questions in conservation physiology (Cooke et al., 2021), and we added two additional questions suitable for stable-isotope applications. Throughout this review, we highlight the questions of Cooke et al. (2021) and provide context for future research that couples conservation to stable-isotope physiology.

Table 2.

Research questions selected from the 100 questions raised by Cooke et al. (2021) and 2 additional questions that are suitable for stable isotope approaches

Question number (Cooke et al., 2021)	Research question	Research examples	Isotopic analysis/indicator	Example of expected isotopic response
1	Can the potential for rapid evolution in physiological tolerance of threatened taxa be maximized?	Analysis of physiological traits in hybrids and genotypes occurring on the edge of a species distribution	δ13C in leaf sugars, whole-leaf tissue, tree-ring cellulose	Lower δ13C in hybrids, reflecting lower stomatal sensitivity to soil water limitations and atmospheric drought
6	How do interactions among plasticity, genetic drift and adaptation affect the resilience of populations to environmental change?	Reciprocal common garden experiments conducted over multiple seasons and years	δ13C in leaf sugars, δ2H, δ18O in leaf and xylem water	Lower δ2H and δ18O in xylem water of more xeric genotypes, reflecting deeper root systems to buffer against soil water limitations
9	What physiological mechanisms determine the pace of thermal acclimation and adaptation?	Analysis of stomatal behaviour and photosynthetic processes in response to temperature extremes	δ13C in leaf sugars and respired CO2	Lower δ13C in plant taxa, reflecting lower stomatal sensitivity to thermal stress and atmospheric aridity
12	How can physiological metrics (e.g. tree growth ring analysis, metabolic rate measures in ectotherms) provide long-term predictions of organismal sensitivity to global change?	Analysis of tree rings performed over multiple decades	δ13C, δ18O in wood cellulose	Lower inter-annual variability in tree-ring δ13C and δ18O in tree taxa following drought, reflecting a lower long-term impact of drought on tree physiological performance
15	How do changes in winter climate and snow and/or ice cover influence overwintering physiology of plants and animals?	Analysis of plant water balance and drought stress over the growing season	δ13C, δ18O in wood cellulose	Lower inter-annual variability in tree-ring δ13C and δ18O in tree taxa, reflecting lower drought stress following years of low winter precipitation/snowfall
17	How will physiological systems adapt and respond to the interactive and cumulative effects of climate change?	Comparison of physiological traits among populations occurring at the core and edges of a species distribution	δ13C, δ18O in whole-leaf tissue and/or tree-ring cellulose, δ2H, δ18O of xylem and soil water	Lower intra- and inter-annual variability in leaf or tree-ring δ13C and δ18O in tree taxa occurring at the core relative to the edge of a species distribution, reflecting lower variability in stress exposure
19	To what extent does physiological resilience (or lack thereof) to environmental change of one target species affect or predict success of another?	Trait comparison of species with similar life history strategies in response to environmental gradients	δ13C in whole-leaf tissues, and/or tree-ring cellulose	Similarities in temporal patterns of δ13C among plant taxa with similar life history strategies
22	How can physiological tools be best used to improve our capacity to monitor organismal and population responses to environmental change?	Measurements of plant growth and leaf trait expression	δ13C in whole-leaf tissues, and/or tree-ring cellulose	Increasing δ13C over time, reflecting increased stress in plant taxa
27	How does human infrastructure and operations (e.g. hydropower, wind turbines, roads) affect the physiological status of wild organisms?	Comparison of organisms occurring at various distances from human infrastructure	δ13C, δ15N of whole-leaf tissues, δ2H, δ18O of xylem water relative to soil water	Gradients in leaf δ13C and/or δ15N, increases in δ2H and δ18O of xylem water downstream of water impoundments, reflecting reduced soil water availability
32	Which physiological attributes facilitate invasive species establishment and spread?	Experiments to manipulate and quantify nutrient and water availability, N sources	δ15N of whole-leaf tissues, δ2H, δ18O of xylem water	Higher δ15N in leaf tissues, reflecting N sourced from local N pollution. Lower δ2H and δ18O of xylem water, reflecting high soil water availability
35	Can physiological vulnerabilities in invasive species be identified and exploited to control them (i.e. know your enemy)?	Experiments to examine Resource limitations and competition	δ13C of whole-leaf tissues, δ2H, δ18O of xylem water	Contrasts of δ13C between invasive plants and native plants, reflecting contrasts in response to resource limitations. Contrasts in xylem water δ2H and δ18O between invasive and native plants, reflecting contrasts in rooting depth
41	Are there physiological measures that can identify or inform tipping points of drought stress in plants, particularly threatened trees and shrubs?	Comparison of tree rings in trees that survived drought versus those that succumbed to drought	δ13C, δ18O in wood cellulose	Higher δ13C and δ18O in cellulose of trees before drought in non-surviving trees relative to survivors, reflecting potential tipping points for predicting mortality of similar populations from future drought
44	What are the physiological ‘early warning signals’ of population decline or collapse?	Population-scale leaf trait analysis and/or tree-ring analysis	δ13C in whole-leaf tissue, δ13C, δ18O in wood cellulose	Marked increases in δ13C, δ18O over time, reflecting temporal decreases in stomatal conductance and photosynthetic gas exchange
71	Can physiological monitoring programs quantify the sublethal impacts of pollution and identify areas in most need of recovery efforts?	Tracing pollutants to their sources and the impact on plants	δ15N of plant tissues, such as leaves and roots	Spatial gradients in tissue δ15N along transects from pollution point sources, reflecting spatial gradients in pollution exposure
75	Do organisms at the core and edges of populations differ in physiological phenotypes such that this can be exploited to enhance management?	Reciprocal common garden experiments placed at the core and edge of population distributions	δ13C in leaf sugars or whole-leaf tissue	Lower δ13C in populations on the warm edge of a species distribution, reflecting lower stomatal sensitivity to aridity
83	How can candidate species for reintroduction be assessed to ensure that they are physiologically suited to the environment to which they would be introduced?	Measurements of leaf gas exchange, leaf trait expression and water source determination	δ13C, δ15N of whole-leaf tissue, δ2H, δ18O of xylem water	Matching δ2H and δ18O between xylem water and groundwater or other stable water source, reflecting the establishment of deep root systems
87	Which physiological functions in plants can be used to select candidates to restore landscapes following disturbances?	Measurements of growth rates and stress in plants reintroduced following disturbance	δ13C in leaf sugars, respiration and whole-leaf tissue	Low values of leaf δ13C following re-establishment of disturbed habitats, reflecting acclimation to disturbed conditions
98	What methods best predict which species will respond well to urban environments on the basis of physiological traits, compared with those that do poorly, regardless of their rarity in natural ecosystems?	Measurements of plant responses to enhanced night-time temperatures and ozone exposure	δ13C in leaf sugars, respiration and whole-leaf tissue, δ13C, δ18O in tree-ring cellulose	Low values of leaf δ13C and δ18O in tree-ring cellulose in trees established in urban habitats, reflecting acclimation to urban environmental conditions.
Additional question 1	What physiological tools can be used to identify and combat the illegal harvest of threatened and endangered species?	Forensic approaches to identify source location of illegally traded specimens	δ18O in whole-plant tissues or cellulose	Matching between δ18O of source waters and modelled δ18O of xylem water
Additional question 2	What are the best practices to monitor and manage plant–microbe symbioses in habitats impacted by disturbance and climate change?	Identify the presence of fungal mutualists following large-scale restoration projects following disturbance	δ13C in soil organic matter and soil respiration	Reduced δ13C in soil organic matter, reflecting allocation of sugars from canopy-pulse-labelled CO2

Isotope theory in biological applications

Plant physiologists developed the application of stable-isotope techniques to study the interactions between plants and their environment. Isotopes of a given element differ in the number of neutrons they contain and are classified as being either ‘heavy’ or ‘light’ depending on whether they are neutron rich or poor, respectively. Ecologists and biologists take advantage of the fact that small differences between heavy and light isotopes of a given element make their rate of reaction, or the partitioning of products, vary in physical or chemical processes depending on the energetics of the process/reaction. Therefore, the ratios between heavy and light isotopes can yield information on a wide range of environmental and physiological processes. The stable-isotope concentrations of a sample are expressed as the molar ratio of the heavy to light isotopes. Since this ratio is small, stable-isotope abundances of a sample (R_sample) are expressed relative to an internationally recognized standard (R_standard) and multiplied by 1000 to yield units in per mil (‰). Isotope ratios are described in delta (δ) notation, where δX = ((R_sample/R_standard) – 1) × 1000, such that X denotes the heavy isotope of a pair. A positive δX value means that the sample contains more of the heavy isotope than the standard; a negative δX value means that the sample contains less of the heavy isotope than the standard.

Here we focus on the stable-isotope ratios of (i) hydrogen ²H/¹H (²H is also called deuterium and then ratio is presented as δD, but is the same as δ²H); (ii) ¹⁸O/¹⁶O; (iii) carbon ¹³C/¹²C; and (iv) nitrogen ¹⁵N/¹⁴N. The natural abundance levels of these isotopes, as well as studies using artificially enriched stable-isotope tracers, can provide information to guide conservation efforts (Peterson and Fry, 1987; Dawson et al., 2002). Changes in the relative abundance of heavy versus light isotopes between a source substrate and the product(s) are called isotopic fractionation (see Table 3 for the definition of common terms). There are two primary types of fractionation: equilibrium and kinetic fractionation. Equilibrium isotope fractionation occurs during reactions that change phases (e.g. going from water vapour to liquid water) and are temperature dependent. These reactions can go both forward and backward depending on the energetics of the phase changes. In the example of water going from vapour to liquid, the heavier isotopes become enriched in the liquid water relative to the vapour. Kinetic fractionation is generally unidirectional, and the reaction rates are mass dependent. Many biological processes are unidirectional and, therefore, isotopic fractionations in biotic systems are usually caused by kinetic fractionation. Many biological kinetic fractionations are mediated by an enzyme that discriminates against an isotope in a mixture (for more details, see Dawson et al., 2002; Kendall and Caldwell, 1998; Peterson and Fry, 1987). For example, in C₃ plant photosynthesis the enzyme Rubisco discriminates against ¹³C, thus producing metabolites and tissues that are depleted in δ¹³C relative to the atmosphere (current atmospheric CO₂ approximately −8.5‰). Carbon stable-isotope ratios (δ¹³C) in plant tissues vary among the three main pathways of photosynthesis—C₃, C₄ and CAM (Winter and Holtum, 2002) (Fig. 1)—with each carboxylation route creating a different fractionation. For example, a sugar from C₄ sugarcane might have a δ¹³C of −10.5‰, as during carboxylation, ¹³C is assimilated more slowly than ¹²C. Sugar from beet, a C₃ plant, typically shows a greater ¹³C depletion of −25‰.

Table 3.

Definitions of key terms used throughout the paper

Term	Definition
Isotopic fractionation	The relative partitioning of heavier and lighter isotopes between two coexisting phases (e.g. water in the liquid and vapour phases)
Isotopic enrichment	Isotopic fractionation that increases the abundance of the heavier isotope relative to the lighter isotope
Water-use efficiency	The amount of photosynthetic carbon gain per unit transpirational water loss or stomatal conductance
Isotopic mixing models	Models that calculate the contributions of different isotopic sources in a sample
Isoscapes	Spatially explicit predictions of isotope ratios that are produced from process-based models
Isotope chronologies	A time series of stable isotope signatures, typically (for plants) measured in tree growth rings, cactus spines and moss or peat layers

Figure 1.

The δ¹³C values recorded from 506 different plant species from nine families. Note that CAM species are relatively small in number (mainly restricted to xeric habitats) and are rather similar in their metabolism and fractionations to C₄ plants. The data were redrawn from Winter and Holtum (2002).

Stable isotopes in water

Isotope ratios of oxygen (δ¹⁸O) and hydrogen (δ²H or δD) in precipitation vary due to temperature, rainfall amount, elevation and distance from the coast (Dansgaard, 1953; Gourcy et al., 2005; Bowen, 2010). The isotopic composition of precipitation reflects fractionation processes during evaporation, transport, condensation and ultimately precipitation itself (Sprenger et al., 2016). There is a global linear relationship between precipitation δ¹⁸O and δ²H first described by Craig (1961) as the Global Meteoric Water Line. Subsequent research investigated variations in Local Meteoric Water lines, which represent site-specific control on hydroclimatic processes (Rozanski et al., 1993; Putman et al., 2019). Precipitation that infiltrates into the soil and becomes plant-available water is further subject to fractionation caused by evaporation. Soil water generally becomes isotopically enriched (i.e. less negative) because of the preferential evaporation of the lighter isotopes (see Fig. 2, modified from Sprenger et al. (2016)). This effect varies by soil depth with generally more enrichment of shallow soil water and less with depth due to evaporative processes in shallow soils (Ehleringer and Dawson, 1992; Dawson and Ehleringer, 1998). Furthermore, groundwater that was recharged at higher elevation or during cool winter months may be more similar to precipitation and less enriched, leading to isotopic variation within the soil profile and capillary fringe associated with groundwater. Because plants generally do not substantially fractionate water through root uptake and xylem transport to the canopy, xylem water can be used to determine where within the soil profile plants are obtaining their water (see Cernusak et al., 2016; Ehleringer and Dawson, 1992). Although, this may not be applicable in all habitats, such as dry or saline habitats (Lin and Sternberg, 1993; Ellsworth and Williams, 2007), wet-dry or even mesic habitats (Evaristo et al., 2017; Barbeta et al., 2019; Vega-Grau et al., 2021).

Figure 2.

Redrawn from Sprenger et al. (2016). Depiction of the processes affecting the pore water stable isotope composition in the vadose zone during summer and winter in a temperate climate. The plus sign indicates an isotopic fractionation process leading to enrichment in heavy isotopes; the minus sign represents depletion in heavy isotopes; and the zero is a sign of non-fractionating processes. However, during water uptake fractionation may occur in some root systems, especially at the root tips. The text indicates the labels of the closest two arrows. Detailed information about spatiotemporal variations of each process are given in Sprenger et al. (2016).

Similar to water in shallow soils, water in leaves is subject to evaporative enrichment during transpiration as stomata open to acquire CO₂ and/or cool leaves (Cernusak et al., 2016). The amount that leaf water is enriched relative to xylem water depends primarily on the leaf-to-air vapour pressure gradient and leaf water turnover rates that are governed by leaf morphology (e.g. leaf size, vein structure) and stomatal regulation of leaf water fluxes. The enriched leaf water is ultimately used in photosynthesis to construct sugars from CO₂ that diffuses into the leaf (Farquhar et al., 1998; Farquhar et al., 2007). Thus, the δ¹⁸O in sugars used to build plant tissues reflects the isotopic composition of soil water taken up by the roots and the amount of evaporative enrichment of leaf water during transpiration, plus post-photosynthetic exchange (Yakir and DeNiro, 1990; Roden et al., 2000; Barbour et al., 2004).

Carbon isotopes in plant tissues

Within a given photosynthetic pathway, δ¹³C in plant tissues mainly reflect the balance between the supply of CO₂ into a leaf, controlled by stomatal conductance, and the demand for CO₂ by photosynthetic enzymes (Fig. 3a; Farquhar et al., 1989). For C₃ plants, δ¹³C ranges from −35‰ to −20‰; for C₄ plants, from −15‰ to −7‰; and for CAM plants, from −22‰ to −10‰ (Osmond et al., 1982; Ehleringer, 1989; Ehleringer and Osmond, 1989). For C₃ plants, δ¹³C values are a reasonable proxy of water-use efficiency (WUE), with more enriched (less negative) δ¹³C values indicating higher WUE and lower δ¹³C indicating lower WUE (Farquhar et al., 1982). WUE is often defined as the ratio of photosynthesis (A, assimilation) over stomatal conductance (g_s), which reflects the supply of CO₂ relative to the demand by the photosynthetic enzymes (with carbon fixation accompanied by kinetic fractionation). Thus, δ¹³C of leaf tissue and tree rings can be used to assess changes in WUE, in response to soil water availability, heatwaves and other stressors (e.g. van der Sleen et al., 2017). Furthermore, measuring δ¹⁸O in tree rings can help resolve whether variation in δ¹³C was due to changes in stomatal conductance or changes in photosynthetic capacity (Shestakova et al., 2014; de Boer et al., 2019). The δ¹⁸O of cellulose should reflect differences in stomatal conductance, vapour pressure deficit (VPD) and transpiration, which lead to differences in evaporative fractionation in leaves that would then be recorded in tree-ring cellulose (Barbour et al., 2004; Barbour, 2007; de Boer et al., 2019). This dual isotope approach produces a general relationship for C₃ plants, where a positive relationship between δ¹³C and δ¹⁸O generally indicates that changes in δ¹³C are due to larger changes in stomatal conductance relative to photosynthetic capacity, while a negative relationship indicates larger changes to photosynthetic capacity relative to stomatal conductance (Scheidegger et al., 2000; Barnard et al., 2012). However, studies of this dual isotope approach have had varied results due to uncertainties in the interpretation of δ¹⁸O that arise from variation in source water and atmospheric δ¹⁸O and species-specific differences in the effect of relative humidity on stomatal conductance. Therefore, these studies should be interpreted with caution (for details, see Roden and Siegwolf, 2012); Roden and Farquhar, 2012).

Figure 3.

Schematic representation of physiological processes that determine carbon isotopic signatures in higher plants and mosses (A). Fractionation against ¹³C occurs during diffusion through stomatal pores in leaves, or water layers over mosses (B), and assimilation by Rubisco. In higher plants, assimilates are transported and incorporated into other parts of the plants. This does not happen in plants that lack a vascular system, including most mosses. Downstream modification can also occur, e.g. to create lipids or waxes. Lower panel. (B) shows how a water layer over mosses creates a barrier to diffusion of CO₂. This means that moss δ¹³C can be used as a proxy for water availability in the moss’s micro-environment. For example, δ¹³C in samples of the Antarctic moss, C. purpureus, reveal a positive δ¹³C gradient with wetness; from dry (more negative) at the top of the slope to wet (more positive) at the lake’s edge (inset graph, sketch from Bramley-Alves et al., 2015). Photo reproduced with permission from Jessica Bramley-Alves.

Nitrogen isotopes in plant tissues

Applications of stable nitrogen (N) isotopes center on N sources (reviewed by Marshall et al., 2007), with δ¹⁵N considered both as an integrator of the N cycle and plant stress responses (reviewed by Robinson, 2001). Plant tissues and xylem water can be depleted or enriched relative to atmospheric N₂ (defined as 0‰). A ~ 3‰ isotopic enrichment with each trophic level (Minagawa and Wada, 1984; Perkins et al., 2014) causes distinct δ¹⁵N signatures in plants that access N from different trophic levels. Plants relying on biologically fixed N₂ can mirror atmospheric N₂ (Shearer et al., 1983), while plants relying on soil N acquire enriched N because the transformations in soil (e.g. microbial denitrification) cause the loss of isotopically light N. However, a strict dichotomy of δ¹⁵N of N₂-fixing and non-fixing plants is often not observed as plant signatures reflect root specializations and N source preferences (Robinson, 2001; Marshall et al., 2007) with numerous processes influencing δ¹⁵N: discrimination against ¹⁵N during uptake, mycorrhizal fungi passing on isotopically light N, rooting depth and isotopically distinct N sources (e.g. isotopically enriched ammonium compared with nitrate). Like other stable isotope pairs, plant parts can differ in δ¹⁵N. For example, δ¹⁵N of root xylem water emerged as a more direct indicator of biological N₂-fixation than foliar δ¹⁵N, which overlapped in N₂-fixing and non-fixing trees (Soper et al., 2015b). As succinctly put by Robinson (2001), the ‘information derived from δ¹⁵N is often circumstantial rather than definitive. δ¹⁵N can provide clues about, or reveal the “footprint” of a process, but not necessarily deliver conclusive evidence for its cause’. δ¹⁵N signatures when carefully applied, can show robust trends ideal to applications in conservation.

Water isotopes and resource acquisition

Short-term δ ¹⁸ O and δ ² H

Conservation of riparian tree species and groundwater-dependent systems

Stable isotopes of oxygen (δ¹⁸O) and hydrogen (δ²H) can be used to determine plant water sources by comparing the stable isotopic values of plant xylem water to the stable isotopic values of all potential source waters (Ehleringer and Dawson, 1992). Water in shallow soils is often more isotopically enriched by evaporative fractionation than deep soil water, stream water and groundwater. Using isotopic mixing models, the fractional contribution and its associated probability can be determined for each water source (Phillips and Gregg, 2003; Evaristo et al., 2017). These studies generally are short-term and track plant water use through a growing season or several growing seasons, by sampling plants at various times throughout the year. Short-term δ²H and δ¹⁸O surveys provide snapshots of plant water use in response to drought, heatwaves or other stressors, such as human alterations to river flows or declining groundwater levels (Flanagan et al., 2019). These studies can aid conservation and restoration practitioners on the proper selection of species and genotypes and/or necessary modifications to restore hydrologic function of ecosystems such as riparian or wetland systems (Schook et al., 2020) (e.g. Question 27 in Cooke et al., 2021; Table 2).

One of the most ecologically important and threatened forest types is riparian forests that are often characterized by a mix of obligate phreatophytes (deep-rooted plants) and facultative phreatophytes. Obligate phreatophytes rely on growing their roots to the capillary fringe above groundwater, while facultative phreatophytes can use both shallow and deep roots to obtain water for transpiration. Alteration of hydrologic regimes due to dams, groundwater pumping and water diversions can alter the availability of streamflow and groundwater, while weather variability and changing climate can alter the availability of precipitation. Successful restoration efforts thus require knowledge of the various dependencies of riparian species on these different water sources. For example, in the southwestern United States, Fremont cottonwood (Populus fremontii) and Goodding’s willow (Salix gooddingii) are important components of riparian systems (Stromberg, 2001). Analyses of δ¹⁸O of xylem water determined that P. fremontii occurring along the San Pedro River—one of the few free-flowing rivers in the western United States—relied on groundwater during dry periods but could opportunistically use shallow soil water from monsoon rains. Alternatively, S. gooddingii was strictly deep-rooted and relied exclusively on groundwater throughout the growing season (Snyder and Williams, 2000). Another study on the riparian species box elder (Acer negundo) found a similar opportunistic use of rainwater during rainy periods and reliance on groundwater during dry periods (Kolb et al., 1997). There is also evidence for a strong genetic component in root distribution patterns within phreatophytic species that may be detected with stable-isotope analysis (Hultine et al., 2020b). For example, P. fremontii genotypes sourced from intermittent streams produce deeper roots than genotypes sourced from perennial reaches (Hultine et al., 2020a). Rooting depth and root elongation rates may be detected experimentally with stable-isotope analysis as trees extend roots through soil profiles with contrasting water isotopic signatures. Thus, the use of stable isotopes in water has the potential to impact riparian restoration by uncovering rooting habits and allow practitioners to match genotypes with predicted future hydrological conditions (e.g. Questions 17, 22 in Cooke et al., 2021; Table 2).

Woody plant responses to ecological drought in dryland regions

Vegetation in drylands is particularly vulnerable to climate change because soil moisture may become increasingly limited as temperatures and potential evapotranspiration increase (IPCC, 2014; USGCRP, 2017). Ecohydrological simulation models indicate that feedbacks between climate change and vegetation change will exacerbate soil water limitations in most dryland regions over the next century (Schlaepfer et al., 2017; Tietjen et al., 2017). Stable isotopes in water may be a powerful tool to identify woody plant species that are most resilient to the effects of ecological droughts. In turn, resilient plant assemblages may limit the synergistic impacts of climate change and vegetation change on ecological drought.

In the deserts of the United States as well as in other ecosystems, summer and winter rainfall can have distinct isotopic signatures due to differences in the formation and precipitation history of rainfall during the growing season in comparison with precipitation during the winter. These signatures have been valuable in understanding how plant species react to changes in climate that affect the seasonality of precipitation (White et al., 1985; Ehleringer et al., 1991; Donovan and Ehleringer, 1994; Phillips and Ehleringer, 1995; Dawson and Ehleringer, 1998). Williams and Ehleringer (2000) studied three dominant woody species along a gradient of increasing monsoon rainfall and sampled δ¹⁸O and δ²H over two growing seasons. Gambel oak (Quercus gambelii) used only deep soil water along the monsoon gradient, while two-needle pinyon (Pinus edulis) and Utah juniper (Juniperus osteosperma) used a large proportion of shallow soil water after monsoon rain events highlighting species-specific responses to changes in the seasonality of precipitation. Coupled with other measurements of plant growth and stress, stable isotopes can be used to elucidate the responses of species to changing environmental conditions. Furthermore, in the above-mentioned study, irrigation with deuterium-labelled water suggested that, for P. edulis, high soil temperatures may limit activity of shallow roots, thus reducing the ability of some shallow-rooted species to use monsoon rainfall (Williams and Ehleringer, 2000). Increasing soil temperature highlights the need to identify species and genotypes that can sustain roots over a long-enough duration to forage for soil moisture, particularly at the seedling stage of development. Coupling stable isotopic measurements in water with ecohydrological simulation modelling is a potentially robust approach for identifying resilient plant traits in a changing climate (Brinkmann et al., 2018) (e.g. Question 17 in Cooke et al., 2021; Table 2).

Longer-term studies of water isotopes

Using Iso-forensics for plant conservation

Stable isotopes have been used in forensic applications to isolate the provenances (origins) of various plant materials. This has helped determine the potential geographic sources of counterfeit currency, plant-derived illegal drugs, adulterated food and beverages, sources of explosives and even to identify the origins of human remains (Ehleringer et al., 1999; Ehleringer et al., 2008; Bartelink et al., 2016). The origins of forensic research efforts were based in ecological studies to understand hydrological inputs, plant water sources, animal migration patterns and the relation of flora and fauna to climate (Gat and Gonfiantini, 1981; Kelly et al., 1991; Ehleringer and Dawson, 1992; Comstock and Ehleringer, 1992a; Hobson, 1999; McGuire and McDonnell, 2007). Spatially explicit maps were created by evaluating the natural abundance of stable isotopes over ecological and geological systems to generate predictive models that incorporate spatial variables to construct isotope landscapes, also known as isoscapes (Bowen, 2010). For example, the stable isotopic composition of precipitation has been used to identify the provenance of plant products (West et al., 2007b). Likewise, an isoscape of the composition of tap water has been used to identify regions of possible habitation for missing persons using tooth enamel and hair from human remains (Bartelink et al., 2016).

There is a potential to use the isoscape framework to target illegal trade of rare plants, which leads to an additional conservation question (Table 2): what physiological tools can be used to identify and combat the illegal harvest of threatened and endangered species? For example, one potential, yet untested, application of the isoscape framework is confronting the illegal international trade of endangered cactus plants. Cacti are among the most threatened taxonomic groups on the planet with over 30% of all species of cacti considered threatened or endangered (Goettsch et al., 2015; Hultine et al., 2016b). The illegal trade of cacti is recognized as a primary contributor to the endangered status of this taxon (Goettsch et al., 2015; Hultine et al., 2016a). Because many cactus species are endemic to small areas, their tissues should only record a small range of values of δ¹⁸O and δ²H in the wild and should be predictable using isoscape modelling approaches (West et al., 2007b). Thus, stable isotopes of cacti tissues, such as spines, could be analysed from plants obtained from points of entry along international borders to detect whether plants are collected illegally from wild populations or cultivated legally in greenhouse facilities. However, using isoscapes to identify illegally traded cacti needs further refinement. Water cycling in cactus stems, particularly giant cacti such as giant saguaro (Carnegiea gigantea) can be slow with water turnover rates of potentially more than a year, leading to large seasonal and inter-annual swings in tissue stable-isotope values (English et al., 2007; Hultine et al., 2019). These large swings limit the capacity for iso-forensic applications to pinpoint a source location for an individual plant, and thus could limit the capacity for these data to stand up under legal scrutiny for giant cacti. Nevertheless, isotopic variation in small-stemmed species of cacti—such as those that are primarily subjected to illegal trading—is likely much smaller than that reported in giant cactus species. Consequently, there is considerable potential to use isoscape approaches as a much-needed tool to combat the illegal trade of this ecologically important plant family (e.g. Additional Question 1, Table 2).

Limitations and on-going questions in water isotope research

Although water isotopes have been used to determine water sources, it is still a field of active research. Plant xylem water in dual isotope space (δ²H and δ¹⁸O) can be more evaporatively enriched than rain, snow melt, stream water, groundwater and even soil water (Brooks et al., 2010; Goldsmith et al., 2012; McDonnell, 2014; Evaristo et al., 2015; Bowling et al., 2017; Berry et al., 2018; Penna et al., 2018). The observed phenomenon whereby plant xylem water is isotopically distinct from water inputs that rapidly recharge streamflow and groundwater (Brooks et al., 2010; Goldsmith et al., 2012; McDonnell, 2014; Evaristo et al., 2015) has been attributed to a fast-moving pool of ‘mobile waters’ versus an ‘immobile’ pool of plant-available water that is tightly bound in soil pores. These observations challenged previous assumptions that soil water was well mixed following precipitation events or that new water uniformly displaced antecedent soil water deeper into the soil profile (Brooks et al., 2010; Sprenger and Allen, 2020). In some cases, xylem water was isotopically more evaporatively enriched than the soil moisture evaporation line (Evaristo et al., 2015; Evaristo et al., 2016), highlighting that other processes deserve further consideration. Potential processes include the following: fractionation within the plant that could be more widespread than expected due to water exchanges at the Casparian strip and mycorrhiza that may be involved in water uptake; plant embolisms that change liquid water to vapour within the xylem; and the transfer of evaporatively enriched phloem water to xylem water (Berry et al., 2018). A review (von Freyberg et al. (2020) suggests ways to move forward including designing experiments that (i) take advantage of extreme variation in water sources, (ii) take advantage of sampling precipitation after extremely dry periods, (iii) use labelling experiments (see section on labelling below), (iv) use potted plant experiments and (v) increase the temporal and spatial resolution of isotopic sampling.

Carbon isotopes and resource acquisition

Short-term carbon isotope ratios

Stable carbon isotopic composition of fully developed leaves in C₃ plants generally reflect their WUE. Therefore, measurement of δ¹³C of leaves to determine WUE can be useful to assess the following: which plant species or genotypes can respond to changes in precipitation patterns and rising temperature; plant materials suitable for restoration of functionally changed ecosystems; and plant species or genotypes suitable for assisted migration. The ability of plants to survive increased temperatures and extended periods of drought is predicted to govern changes in species ranges as a consequence of climate change (Wilson et al., 2005; Vitt et al., 2010; Crimmins et al., 2011; Pauli et al., 2012). In a review of key traits for assessing adaptive phenotypic plasticity across a range of species, WUE assessed through stable isotopes was 1 of 11 key traits identified (Nicotra et al., 2010). The effects of climate change may be amplified in high elevation hotspots of biodiversity (IPCC, 2013), which are often characterized by rare and endemic species (Casazza et al., 2005; Casazza et al., 2008). However, widely distributed species that span a range of temperature and moisture gradients and are critically important for regional productivity are also at risk depending on the ability of these populations to adapt to increased temperatures and drought. We provide two examples, a rare endemic species and widely distributed forest species, to illustrate how stable isotopes can guide management decisions regarding conservation. Heterotheca brandegeei is a perennial pincushion plant endemic to rocky high elevation outcrops in Baja California, Mexico. Winkler et al. (2020) used growth chambers to determine the effects of drought and predicted temperature increases on the early life stage responses and phenotypic plasticity of H. brandegeei. Leaf-level WUE, measured using the δ¹³C of leaves, did not change in response to warming or drought alone; however, drought in combination with warming significantly increased WUE. This increase in WUE was partly explained by maternal lineage, indicating a role of genetic variance in adaptive plasticity that is detectable using measurements of leaf δ¹³C ratio (e.g. Questions 6 and 17 in Cooke et al., 2021; Table 2).

Aleppo pine (Pinus halepensis) is the most common coniferous tree in the Mediterranean and spans a wide range of temperature and moisture gradients. Intraspecific genetic differentiation in the various climatic regions has been proposed as likely to shape this species’ response to the magnitude and timing of droughts (Voltas et al., 2008). Voltas et al. (2008) used a common garden experiment at two sites with 25 P. halepensis populations that spanned the species’ geographic range. They analysed the δ¹⁸O and δ¹³C of hemicellulose in tree rings of 3-year-old P. halepensis branches. Trees from moist regions had lower WUE and grew more rapidly, while trees from dry regions had greater WUE and grew more slowly. The positive relationship between δ¹⁸O and δ¹³C for P. halepensis indicated that variation in δ¹³C was mainly driven by differences in stomatal regulation. Specifically, VPD, the length of seasonal drought and total rainfall were climate characteristics that influenced WUE. Measuring how WUE shifts in response to changes in climate is a potentially powerful tool for guiding successful conservation and restoration success of rare/endemic and widespread species (e.g. Questions 17, 22 and 75 in Cooke et al., 2021; Table 2).

WUE and its interpretation is a key concept in plant ecology with debate as to whether high or low WUE confers a fitness advantage because it is highly dependent on the timing and severity of water stress (Heschel and Riginos, 2005) and may vary contextually by biome (see review by Nicotra and Davidson, 2010). In the case of the endemic perennial pincushion example, higher WUE coupled with mortality in the most severe treatments was indicative of high WUE reflecting greater stress and not greater fitness. In dry environments, selection favouring high WUE was sometimes associated with increased fitness, while other times lower WUE was associated with increased fitness and, in some cases, there was no correlation between WUE and fitness (Nicotra and Davidson, 2010). These results indicate that there may be opportunities to use δ¹³C measurements as proxies of WUE to predict fitness of individual plants before they are planted ex situ to restore plant populations and communities.

There are complexities and potential pitfalls with δ¹³C analysis of leaves that need to be considered. For example, in some species there are differences in young versus old leaves—with young leaves switching from being dependent (heterotrophic) on other leaves or carbohydrate storage until photosynthetic capacity develops in the new leaf (autotrophic) (Cernusak, 2020). The stable carbon isotope ratio of some species may also reflect the environment when the leaves were formed and not the entire growing season, and for some species this reflects periods of high water availability in the spring during leaf flushes (Comstock and Ehleringer, 1992a,b; Graham et al., 2014). For some deciduous species, the growth of new leaves relies upon stored carbohydrates fixed during the previous growing season, thus creating time lags between the current resource environment and past environmental conditions (Ehleringer et al., 1992). Therefore, the interpretation of leaf δ¹³C is not straightforward because the timing of leaf formation relative to the acquisition of carbon is not constant. One way to reduce temporal lags between leaf formation and current conditions is to measure leaf photosynthates (i.e. soluble sugars) that usually have a 24–72-hour turnover rate (Brugnoli et al., 1988; Fravolini et al., 2005; Hultine et al., 2013). However, leaf soluble sugar extraction is fairly labourious compared with analysing whole-leaf tissue.

Additional challenges arise from the variation in leaf δ¹³C that occurs at a variety of scales within leaves, within individual plants, within a species, among species and among species across climatic gradients (for global review, see Cornwell et al., 2018). This variation has been attributed to variation in stomatal aperture (Farquhar et al., 1982; Farquhar and Richards, 1984), the photosynthetic biochemistry of a leaf (Virgona et al., 1990) and the difference in diffusion of CO₂ within the mesophyll (Evans et al., 1986; Barbour et al., 2010). Fractionation can continue to occur as more complex compounds are produced within plants (Fig. 3a). For example, fractionation during the formation of lipids and waxes means that gradients in δ¹³C from the epidermis to the leaf center can be considerable, e.g. up to 4.3‰ in CAM plants (Robinson et al., 1993). Importantly, variation in mesophyll conductance can at least partially decouple measurements of leaf δ¹³C from WUE. For example, mesophyll conductance can vary substantially across a broad range of species and result in up to a 3-fold difference in carbon isotope discrimination among species (Warren and Adams, 2006). This means that care must be taken to compare similar tissue types and in making inferences from leaf δ¹³C.

Long-term responses of tree-ring stable-isotope ratios δ ¹³ C and δ ¹⁸ O—the need for climate proxies

Tree-ring widths from a composite of several trees have been used by dendrochronologists to successfully reconstruct past climates (Cook et al., 2010). Stable isotopes of oxygen and carbon in tree rings have also been used to infer past climatic variables such as temperature, relative humidity, the occurrence of drought and tree physiological responses to droughts and soil moisture (Saurer et al., 1997; McCarroll and Loader, 2004; Sarris et al., 2013; Lavergne et al., 2017; van der Sleen et al., 2017). The processes that affect the stable-isotope ratios δ¹⁸O and δ¹³C in tree rings include canopy temperature, transpiration and photosynthetic uptake of CO₂. These processes are directly linked to meteorological variables such as air temperature, VPD, solar irradiance and available soil moisture from precipitation, which make stable isotopes of tree rings useful to evaluate species-specific physiological responses, as well as useful proxies of past climates (Hartl-Meier et al., 2015). However, the biophysical processes that regulate the synthesis of sugars and starches, transport processes between phloem and xylem and the ultimate conversion of photosynthate into lignin and cellulose in tree rings are complex and still not completely resolved, in particular regarding post-photosynthetic fractionations. These fractionation processes can decouple leaf processes recorded in stable-isotope ratios of transpiration and photosynthesis from the stable-isotope ratios recorded in tree rings, though much progress has been made (see review by Gessler et al., 2014).

Furthermore, tree-specific strategies of adjusting individual physiologies can be revealed in stable isotopes of tree rings such as long-term responses to climate (McCarroll and Loader, 2004; van der Sleen et al., 2017). Longer-term analyses of δ¹³C and δ¹⁸O in different tree genotypes can provide important information on mitigating the effects of extreme drought by identifying drought-resistant genotypes. Additionally, surveys of stable-isotope chronologies in a forest experiencing mortality can yield predictive tools to assess where mortality may be greatest and identify priorities for restoration planning (Maier et al., 2019). For example, Schook et al. (2020) compared the tree-ring width and δ¹³C of riparian cottonwoods (Populus angustifolia and P. angustifolia x P. trichocarpa) along a partially dewatered river reach to watered reaches over a 50-year time series. Tree-ring width took decades to respond to lowered water availability, while δ¹³C reflected dewatering immediately. Thus, δ¹³C provided an early indication of branch and tree mortality (see also Rood et al., 2013) (e.g. Questions 12, 41 and 44 in Cooke et al., 2021; Table 2).

Stable isotopes in tree rings and other long-lived plant tissues such as cactus spines (and even bryophytes; see below) have the potential to be powerful tools in conservation physiology. The vertically arranged chronological series of cactus spines provides a non-invasive, sensitive and high-resolution time-series of the physiological and metabolic responses of cacti plants to environmental variability that are similar to tree-ring isotope records (English et al., 2007; Hultine et al., 2019). For example, a 58-year-old δ¹⁸O spine chronosequence in the threatened and long-lived giant cactus, cardón (Echinopsis atacamensis), on the Bolivian Altiplano revealed varying degrees of plant water evaporation and stem recharge over time (English et al., 2021). Moreover, minimum annual δ¹⁸O in spines had a strong positive relationship to the annual mean of minimum monthly temperatures (i.e. lower annual mean minimum temperatures lead to lower VPD, lower evaporation and lower δ¹⁸O values in spines). These results show that during the cool/cold night temperatures of the Altiplano (4000 masl), small increases in temperature can have an exponential effect on VPD, leading to higher δ¹⁸O in spines (i.e. more evaporated cactus stem water) in years with higher minimum (night-time) temperatures. Thus, spine δ¹⁸O chronologies could reveal changes in temperature that may impact the water balance of this culturally important yet threatened cactus species (e.g. Questions 9, 12, 15 and 17 in Cooke et al., 2021; Table 2).

Long-term responses of bryophyte stable-isotope ratios δ ¹³ C and δ ¹⁸ O

In addition to these examples of large woody plants, stable isotopes have been used to better understand the response of some of the smallest, non-vascular plants to climate and landscape change. Some mosses contribute to the development of peat banks, which record thousands of years of climate history and in which both carbon and oxygen isotopes are used to infer changes in temperature, water availability and water sources over millennia (Björck et al., 1991; Royles et al., 2012; Royles et al., 2013a; Royles et al., 2013b; Royles and Griffiths, 2015). Recently, stable isotopes have also been applied to living shoots of long-lived (>100 years) moss plants, providing information on recent changes in local environments (Clarke et al., 2012; Robinson et al., 2018). Such data can be particularly useful in alpine and polar regions where mosses tend to be dominant components of the flora and to sequester considerable amounts of carbon.

In Antarctica, stable isotopes of carbon have been shown to reflect the extent to which mosses are growing in water-saturated or dry, exposed microclimates (Fig. 3b). As nonvascular plants with reduced movement of assimilates between cells and tissues, many mosses lay down cellulose in sequential layers (similar to vertical tree rings; Clarke et al., 2012; Robinson et al., 2018). Because mosses lack stomata, isotopic discrimination is largely determined by diffusion into the plant including across the cell wall and then enzymatic fractionation by Rubisco. If the photosynthetic surface of the moss is covered with water, then access to carbon dioxide is limited by diffusion, and the photosynthetic enzyme Rubisco has less opportunity to discriminate against the isotopically heavier ¹³CO₂ (Fig. 3b). This means that where moss is growing in predominantly wet hollows, or under thick montane clouds, the organic matter will be relatively enriched in ¹³C (Bramley-Alves et al., 2015; Royles and Griffiths, 2015; Horwath et al., 2019). Conversely, if the moss is growing in a drier and more exposed microclimate like a hummock, its photosynthetic surface will be exposed to air, greater discrimination will occur and cellulose will be more depleted in ¹³C (Bramley-Alves et al., 2015; Royles and Griffiths, 2015). Furthermore, analysis of ¹³C and carbon’s radioactive form ¹⁴C—to establish a date in which a sample was formed using accelerator mass spectrometry—can reveal how moss respond to climate variation overtime. For example, δ¹³C in moss shoots collected from the Windmill Islands, East Antarctica, were correlated negatively with wind speed and positively with air temperature (Clarke et al., 2012). This dual isotope technique was then used to show that ozone depletion and climate change (Robinson and Erickson III, 2015) are drying these East Antarctic terrestrial communities with consequent changes in species favouring survival of cosmopolitan species like Ceratodon purpureus over the endemic moss, Schistidium antarctici (Robinson et al., 2018). This technique allows scientists to monitor environmental changes in polar vegetation communities across space and time using carbon isotopic signatures (Clarke et al., 2012; Royles et al., 2013a; Bramley-Alves et al., 2015; Royles et al., 2016; Robinson et al., 2018).

Isotopic analysis can be used to determine plant communities at risk from (Robinson et al., 2018), or favoured by (Royles et al., 2016), climate change at broader scales or those impacted by buildings that can affect hydrology at local scales (Robinson et al., 2018; Brooks et al., 2019). Combining stable isotopic measurements (¹³C and ¹⁸O) can help to integrate photosynthesis, growth and water supply in polar moss species. However, more research, including growth under controlled environmental conditions (Bramley-Alves et al., 2015) is needed to fully understand the oxygen and deuterium signals (Barbour, 2007; Royles and Griffiths, 2015).

Given that trees are absent from many high polar and alpine regions, this use of isotopes in mosses could provide proxy data to constrain climate models (Royles and Griffiths, 2015; Robinson et al., 2018). An alpine example of such an application is seen in tropical montane cloud forest mosses and liverworts where isotopic composition of epiphytic bryophytes can reveal the position of the cloud immersion zone (Horwath et al., 2019). The stable-isotope composition (δ¹³C and δ¹⁸O) of these canopy-dwelling bryophytes reflects diffusive limitations due to surface water and, along with C/N content, provides a generic index for the extent of cloud immersion. From lowland to cloud forest, δ¹³C increased from −33‰ to −27‰, while δ¹⁸O increased from 16.3‰ to 18.0‰. Changes in stable isotopes thus have the potential to show where the cloud base is shifting upslope with climate change. Since changes in isotopes were apparent prior to diversity changes, they could be used to track epiphytic communities at risk, as well as to identify new areas suitable for growth. Combined radio- and stable-isotope analysis of bryophyte shoot cores could thus provide conservationists with a climatic proxy data to track changes in Antarctic terrestrial communities and potentially other bryophyte-dominated ecosystem change (e.g. Questions 12 and 41 in Cooke et al., 2021; Table 2).

Unfortunately, not all moss species are suitable as climate proxies. Plant life form and anatomical differences can influence the extent to which water layers form and/or rates of diffusion of CO₂ into leaflets of different moss species, and this will then be reflected in their effectiveness as climate proxies (Royles and Griffiths, 2015; Perera-Castro et al., 2020). It should also be noted that mosses track climate in very localized microclimates and so multiple independent samples over a wide area are needed to infer macro climatic trends (Robinson et al., 2018).

Nitrogen isotopes and resource acquisition

Plants growing in ecosystems that recycle proportionally more N in the vegetation-soil-vegetation loop are less enriched in ¹⁵N than in ecosystems with a more open N cycle and loss of isotopically light N (Fig. 4). Across biomes, foliar δ¹⁵N signatures span mostly from −8‰ to 8‰ from mesic to arid habitats, which broadly mirrors the increasing isotopic enrichment of whole-soil N (Handley et al., 1999). Similarly, within a biome, a more open N cycle in the drier ecosystem (e.g. wetter vs. drier rainforests) results in isotopic enrichment of soil and plants (Handley et al., 1999; Houlton et al., 2006). Within an ecosystem, ecological guilds (early pioneer to late successional species) can have distinct foliar δ¹⁵N indicative of different N sources and N physiologies (Aidar et al., 2003). In systems that have extreme isotope signals, δ¹⁵N can be used as a natural tracer. Seabird guano (δ¹⁵N ~ 10‰) fractionates into isotopically depleted gaseous N and isotopically enriched liquid and solid N, allowing tracing of N sources with foliar δ¹⁵N ranging from −10‰ to 20‰ in landscapes with large seabird rookeries (Erskine et al., 1998; Schmidt et al., 2004; Wasley et al., 2012) (Fig. 4).

Figure 4.

Plant δ¹⁵N signatures integrate N relations. Lower δ¹⁵N signatures occur in mesic plant communities and systems with a more closed N cycle (less N loss). Higher δ¹⁵N signatures characterize systems with more open N cycle (greater N loss), such as arid environments and systems with high N input. Plant δ¹⁵N mirrors distinct N sources with isotopic signatures ranging from neutral (air N₂), depletion or enrichment in natural ecosystems, and generated anthropogenically. Other effects on plant δ¹⁵N, such as environmental stresses or root specializations, are not shown.

Stable-isotope ratios of N can be used for analysis of plant–plant interactions, for example the impact of invasive plant species on native species at the ‘isoscape’ landscape level. In a dune system, the δ¹⁵N signature of native species Corema album identified the influence of an invasive N₂-fixing Acacia species with isotopic enrichment in the native species together with a doubling in leaf N content (Hellmann et al., 2016). Stable isotopes of N can also support interpreting stress physiology. For example, genotype comparisons of wild barley found that drought- and N-stress tolerant genotypes had the lowest δ¹⁵N values, reflecting the extent to which stress tolerant plants better retained N (Robinson et al., 2000). Similarly, among taro (Colocasia esculenta) genotypes, the highest WUE correlated with the lowest δ¹⁵N (Gouveia et al., 2019) (e.g. Questions 32 and 83 in Cooke et al., 2021; Table 2).

Short-term plant nitrogen uptake and δ ¹⁵ N

In the short term, N uptake via roots is most immediately reflected in xylem water. At wetter sites along a continental moisture gradient, foliar δ¹⁵N of N₂-fixing Acacia species overlapped with non-fixing Eucalyptus species, while Acacia xylem water was depleted in δ¹⁵N and distinct from δ¹⁵N-enriched xylem water of Eucalyptus (Soper et al., 2015b). At drier sites, both tree genera had similar xylem δ¹⁵N signatures indicative of soil N use rather than biological fixation. Similarly, moribund vs. actively growing Acacia trees had isotopically enriched vs. depleted xylem water (Soper et al., 2015b), and, by using xylem water (but not foliar δ¹⁵N), biological N₂-fixation could be identified in Prosopis glandulosa in grassland-to-woodland transition (Soper et al., 2015a). These patterns indicate that measurements of δ¹⁵N in xylem water can detect rapid shifts in plant N sources that, in turn, could preclude other measures of plant stress or changes in the availability of N in the ecosystem (e.g. Question 71 in Cooke et al., 2021; Table 2).

Longer-term responses—the nitrogen cycle and δ ¹⁵ N

Nitrogen is in the global spotlight because reactive N in the biosphere has more than doubled over the past century. Of the ~120 Tg N that are annually fixed synthetically, half is lost from agricultural soils and, together with N derived from manures, other wastes, urban and industrial processes, impacts global ecosystems. Nitrogen deposition is considered the third most important driver of plant biodiversity loss in terrestrial ecosystems (Midolo et al., 2019). Plant δ¹⁵N signatures can reflect the distinct isotopic signals of pollutant N sources in industrial, agricultural and urban environments. Tracking historic trends, aerial rainforest epiphytes showed a pronounced shift from minor δ ¹⁵N depletion (−3‰ to −1‰) in herbarium specimens to strong depletion (−10.9‰) in contemporary epiphytes exposed to air pollution from petro-chemical and fertilizer industries, while δ¹⁵N of contemporary epiphytes at remote locations resembled herbarium specimens (Stewart et al., 2002). Exposed to gaseous ¹⁵N-depleted NH₃ and NO_x in agricultural landscapes with animal husbandry and crops, plant δ¹⁵N averaged −11.2‰, exposed to ¹⁵N-enriched NO_x of vehicle traffic, plant δ¹⁵N was 13.3‰, while in natural ecosystems plant δ¹⁵N averaged 5‰ (Díaz-Álvarez et al., 2018).

Thus, plant functional types can be explored for different investigations: mosses as ideal biomonitors for wet deposition, vascular epiphytes for dry deposition and terrestrial plants to monitor N saturation of soil. At the plant community level, a more closed N cycle with lower N turnover has less isotopic enrichment than an open N cycle, which can manifest in lower plant δ¹⁵N signatures, higher plant diversity and more efficient N use (Kleinebecker et al., 2014), due to less isotopically enriched N sources and because plants discriminate against ¹⁵N in the presence of excess N (Marshall et al., 2007). In oligotrophic alpine regions vulnerable to N enrichment, N and O isotopes in plant tissues could quantify atmospheric input of nitrate (Bourgeois et al., 2019).

Nitrogen isotopes allow forensic investigation and mass balance calculations of N inputs at a landscape level. Plant δ¹⁵N mirrored the isotopic signal of municipal effluent in a production forest and wetland, which intercepted 65% of the applied N (Tozer et al., 2005) while most of the remaining N (29%) entered a stream and could be tracked as isotopically distinct nitrate. Illustrating the detailed information derived from N isotope analysis, ¹⁵N enrichment of effluent-fertilized vegetation of up to 20‰ could be tracked with different stem parts that had distinct δ¹⁵N signatures, enabling temporal resolution of N sources (Tozer et al., 2005). This is of value to conservation managers as plant δ¹⁵N signatures integrate the isotopic composition of atmospheric, soil and water sources and document changing N relations over temporal and spatial gradients. Stable-isotope ratios of N can be a standalone tool or accompanied by other responses at ecosystem and organism levels, supporting observations of changing species composition, tissue N content, growth rate, phenology or mortality rate (e.g. Question 71 in Cooke et al., 2021; Table 2).

Isotopic pulse labelling

A primary utility of stable isotopes is to identify plant resiliency (or susceptibility) to environmental stress caused by drought, heat waves, herbivory/disease, episodic disturbance or a combination of stress factors. Variation in the natural abundance of two isotopes of a given element can often uncover patterns of resilience when measured in plants and the surrounding environment. However, under many conditions, natural abundances do not contain enough variation to detect plant responses to environmental stress. Therefore, more sophisticated isotopic labelling approaches can yield critical information on plant resilience.

One of the most straightforward and effective labelling approaches in stable-isotope ecology is enriching irrigation water with a deuterium spike that can be traced in xylem water or leaf water. Previous deuterium ‘pulse-labelling’ experiments have uncovered contrasts in shallow water versus deep water exploitation among co-occurring dryland plant species (Williams and Ehleringer, 2000; Schwinning et al., 2002; West et al., 2007a), among different periods of the growing season (Williams and Ehleringer, 2000), on different geomorphic surfaces (Fravolini et al., 2005), as well as different transit times of groundwater recharge and plant water uptake (Evaristo et al., 2019). Experiments using isotopically enriched water have also been used to instantaneously track the regeneration of leaf waxes (Gao et al., 2012) and water turnover rates in succulent-stemmed plants (English et al., 2007). Combined, these experiments provide a template of how plant conservation may be benefitted through isotopic labelling of water taken up by plants. For example, isotopic irrigation experiments could identify genotypes or species that are most resilient to changing seasonal precipitation patterns. Similarly, deuterium-labelling experiments can be used to detect whether certain plants can best regenerate and maintain leaf waxes that are critical for preventing passive water loss from leaves during heat waves, reduce intense UV exposure and minimize foliage herbivory (e.g. Question 22 in Cooke et al., 2021; Table 2).

Pulse-labelling experiments can also be used to track the fate of assimilated CO₂ following photosynthesis by fumigating a plant, or part of the plant canopy with ¹³C-labelled CO₂ (Dawson et al., 2002; Staddon, 2004; Epron et al., 2012). Pulse-labelling of CO₂ can help quantify resource partitioning under varying environmental conditions (Joseph et al., 2020) and help determine carbon allocation strategies and carbon transport to fungal mutualists (Grimoldi et al., 2006) that are often critical for plant fitness and conservation, which leads to an additional conservation question (Table 2): what are the best practices to monitor and manage plant–microbe symbioses in habitats impacted by disturbance and climate change? Stable isotopes can identify the presence of fungal mutualists following large-scale restoration projects following disturbance. For example, a pulse ¹³C-labelling experiment conducted in an old-growth Scots pine forest revealed that belowground carbon allocation to microbes is strongly modulated by soil moisture (Joseph et al., 2020). A recent study on Scots pine corroborates earlier studies that indicate that the transfer of recently acquired photosynthates to belowground sinks slows as soil-water content decreases (Epron et al., 2012). These results show the extent to which drought can disrupt plant-microbial mutualisms that may be important to successfully restore ecosystems impacted by disturbance or competition from invasive species (Meinhardt and Gehring, 2012; Grove et al., 2017). However, belowground allocation is also strongly governed by seasonality, temperature and plant phenology (Epron et al., 2012). For example, a pulse-labelling experiment revealed that belowground carbon allocation in 20-year-old beech peaked in midsummer but shifted to labile carbon storage under cooler temperatures at the end of summer (Epron et al., 2011). Thus, the seasonal timing in which ¹³C-labelling experiments are conducted is an important consideration (e.g. Additional Question 2, Table 2).

Ongoing advances in isotopic labelling approaches coupled with advances in genetic sequencing are providing new avenues for understanding basic plant biology in ways that can guide conservation. Among these are advances in the fields of metabolomics, transcriptomics and proteomics. Carbon isotope labelling experiments coupled with metabolomic, transcriptomic and proteomic profiling can be used to identify ways to improve nitrogen use in plants or uncover metabolic pathways for plant defence against microbes (Pang et al., 2018; Zhang et al., 2018) by characterizing metabolic fluxes, pathways and networks (Chokkathukalam et al., 2014; Nakabayashi and Saito, 2020). As with pulse chase experiments using δ²H of water or δ¹³C of air, compound-specific isotope labelling coupled with omics analysis can be challenging, labour intensive and costly to implement and analyse. However, rapid technical advances promise to provide new opportunities to develop conservation strategies of plants and ecosystems threatened with environmental changes.

Future directions

Over the past several decades, stable-isotope ecology has evolved from the basic exploration of isotopic variation in nature to being recognized as a powerful tool for plant and ecosystem responses to environmental change. Consequently, as conservation physiology continues to grow as an emerging discipline of plant ecology, stable-isotope techniques are increasingly gaining acceptance within the plant conservation physiology toolbox (Madliger et al., 2018). Advances in mass spectrometry and tunable diode lasers (e.g. Schaeffer et al., 2008) are not only increasing the spatial and temporal resolution of field-collected data, they are also leading to reduced costs associated with stable isotopic inquiry. In turn, these advances are leading to new avenues to couple stable-isotope analysis with other high-data capacity measurement tools including whole-genome sequencing (Chhetri et al., 2019), remote sensing (Lavergne et al., 2019) and phenology studies using automated camera sensors (Brown et al., 2016). As conservation physiology evolves to include a broader spectrum of physiological tools and concepts (Cooke et al., 2020), stable-isotope analysis has the potential to be on the forefront of conservation. To maximize this potential, we advocate for the construction of publicly accessible databases of isotopic abundance in plant tissues, soils and air (see Hayden et al., 2014), continued standardization of laboratory protocols and standards and frequent horizon scanning to identify emerging challenges and opportunities for stable-isotope ecology and physiology to enhance conservation.