Water and Temperature Ecophysiological Challenges of Forests Plantations under Climate Change

The impact of current and potential changes in the relationship between forests and society is critical in many parts of the world, especially considering that the value of intensively managed forest plantations for carbon mitigation has become challenged and debated as a valid mitigation strategy [1,2,3], and also due to the fact that forest plantations have been considered to be in conflict with rational land and water resource use [4,5,6,7,8]. Given the complexity of the scenario surrounding forest plantations, the forestry industry has become particularly aware of the issues and challenges relating to sustaining forest productivity under climate change (CC) [9,10,11,12]. Large areas being planted and managed for fiber and timber production have been showing increasing rates of mortality due to unexpected longer drought periods and an increase in extreme dry season temperatures.

Therefore, understanding and modeling the response of tree species to potential variations in temperature and water availability under CC is essential for forecasting changes in forest productivity and determining their true potential C sequestration contributions. Developing indicators that identify and prevent large-scale mortality of individual tree species is key to forest sustainability [13,14,15]; McDowell et al. [16] and Peltier and Ogle [17] postulated that understanding genetic, physiological, and hydraulic traits is essential to be able to manage future forests. For example, identifying species/genotypes with tolerance to drought [18,19], or identifying vulnerable species or varieties, will create effective conservation programs and policies that may maintain productivity and carbon sequestration goals [20].

Understanding the effect of CC on forest plantations is complex due to the many unknowns involved [16,21]. In the 1980s, the first hypotheses related CC with vegetation growth patterns [22,23]. Initially, it was considered that species/ecosystem responses were specific; however, this highlighted the importance of scaling up studies and the need for understanding potential global implications [24]. In the following decades, society drove tremendous scientific research interest into CC, facilitating analyses from different perspectives. Multiple studies investigated the effects of CC with spatiotemporal dimensionality, given many existing local to global uncertainties [25,26]. Bastin et al. [27], Butt et al. [20], and Adams et al. [28] mentioned that incorporating spatial (local, regional, continental or global) and temporal (historical data or future predictions) dimensionality in research implied a minimization in prediction errors, given that CC effects are not constant around the planet or over time. For example, an IPCC report showed that 58% of greenhouse gas (GHG) emissions have been generated in the last 30 years, and in the last decade, emissions have increased by 20% compared to previous projections [29]. In this regard, research into the effects of CC has been focused on different climatic parameters and understanding the response of tree species to changes in ozone (O₃), CO₂, temperature, and precipitation, and how a negative response (considered stress) influences tree growth and survival [30,31].

There is extensive current knowledge about the metabolic response of trees to changes in environmental O₃-CO₂. It has been determined that if higher CO₂ and lower O₃-CO₂ are maintained, responses across different scales may include those at the cellular level, with effects on the activity and production of RuBisCO [32], flavonoid biosynthesis [33], and protein repair and turnover [34]; those at the leaf level, with changes in photosynthesis, starch, sucrose metabolism, respiration, and leaf structure [35,36]; and those at the tree level, with changes in biomass allocation, leaf area, and reproductive output [37,38]. In the case of forest plantations, changes in gross primary production partitioning (GPP) and mortality have been the major focus [39]. A great matter of debate has arisen in regard to whether an increase in CO₂ is beneficial in the long term to forest productivity because of the “CO₂ fertilization effect on photosynthesis”. This benefit has been compared against increases in global warming and associated changes to hydrological cycles, which can have negative effects on forests [40]. Multiple studies have focused on analyzing and modeling the response of different species-genotypes to different temperature-water balance combination regimes in order to identify the ranges of adaptability and mortality rates, and how these environmental conditions may affect growth and resource use [1,20,41].

A basic aspect of understanding the adaptation of forest plantations to CC involves characterizing the degree of vulnerability of species or genotypes to climate change. Genetic improvement has usually focused on improving production, resistance to pathogens, and adaptability to nutritional restrictions [42]; however, current species and genotypes are not guaranteed to be resilient to CC scenarios [43]. Previous studies developed by Trugman et al. [44] and Schwarz et al. [45] critically considered identifying genotypes with some degree of plasticity to regulate carbon flux and water use, which guarantees resistance to heat stress and drought conditions in trees [46]. Nicotra et al. [47], Teskey et al. [48], and Sala et al. [38] considered it key to identify the degree of phenotypic plasticity to heat stress and drought conditions present among species and genotypes. Understanding stress response mechanisms and identifying physiological traits that secure tree survival at different stages of development may also be of great importance. According to Gunderson et al. [49], O’Brien et al. [50], and Pita-Barboza et al. [51], response mechanisms could be categorized as avoidance (ability to maintain a favorable internal water balance under stress), tolerance (ability to maintain growth and biomass production under stress), and resistance (a sum of avoidance and tolerance mechanisms).

O’Brien et al. [50] synthesized morphological, physiological, hydraulic, and biochemical response mechanisms to water stress. Morphological traits consider absolute and relative growth efficiency per individual or stand biomass allocation (e.g., root system architecture, root maximum depth, and root: shoot relationship), anatomical traits at leaf level (e.g., stomatal density, leaf area, and leaf thickness), and xylem structure (e.g., density and diameter of vessels and wood density). Silva et al. [52] found that Eucalyptus genotypes with greater resistance to drought reduced their leaf area as an adjustment mechanism to avoid excessive transpiration and regulate photo assimilation. Similarly, Corrêa et al. [53] identified stress-tolerant Eucalyptus species that develop complex root architecture, allowing trees a greater probability of survival under droughts, but negatively affecting wood production. Worbes et al. [54] also concluded that vessel size is the limiting variable regulating hydraulic conductivity and tree water loss, and that wood density is correlated with water transport and loss for plantation forests containing tropical species.

Physiological variables such as photosynthetic rate, transpiration rate, stomata conductance, carboxylation efficiency, leaf water potential, and non-structural carbohydrates (NSC) have also been investigated as proxies to evaluate drought resistance [55]. Photosynthesis and transpiration are highly susceptible to changes in climatic conditions, so it has been suggested that species with greater vulnerability to stress will experience sudden changes in these processes [56]. Aspinwall et al. [57], investigating the response of Eucalyptus tereticornis genotypes to heat stress, observed a significant reduction in, and a lower physiological variation of, photosynthesis, transpiration, stomatal conductance, and leaf relative water content for resistant genetic materials, and concluded that physiological traits may allow for discriminating genotypes that could withstand stress conditions.

Hydraulic traits such as the hydraulic safety margin, water potential at P50 or P88, and percentage loss of conductance (PLC) have been directly related to mortality under drought conditions. For example, Anderegg et al. [58] found that PLC was highly correlated to mortality for Pinus spp. and that wood anatomical traits were critical in this context. Kursar et al. [59] recommended the use of xylem hydraulic failure as a predictive factor for drought-induced tree mortality. Similarly, Mantova et al. [60] postulated that relative water content (RWC) and PLC could provide novel and valuable information surrounding the prediction of tree mortality, and thus forest dieback, under drought conditions. O’Brien et al. [50] recommended that hydraulic traits be analyzed in greater depth in tree species from different regions and conditions. Evidence of the mechanistic link connecting hydraulic failure with living cell damage and tree death is yet to be identified, and this compromises our ability to predict mortality events. Some studies have focused on identifying tolerance traits through analyzing the resilience of materials to hydraulic failure and carbon starvation (Figure 1) [50,61].

Biochemical traits, such as the activity and production of RuBisCO [32], flavonoid biosynthesis [33], antioxidant activity, carotenoid concentration, and protein repair and turnover [34] have been correlated well with morphological and physiological adaptation traits to stress. For example, at the leaf level, it has been identified that changes in photosynthesis and stomatal closure processes show a significant relationship with RuBisCO and ABA concentration [35,36].

Using traits associated with stress tolerance or resistance allows, in the short term, improvement in the adaptability of forest plantations to adverse events related to CC. Identifying stress resistant or tolerant traits at the nursery stage may significantly affect the selection process of genotypes to be planted according to the environmental conditions of each site and may provide room to facilitate reduced intraspecific competition for water resources [62]. For example, Müller et al. [63] and García et al. [64] differentiated resistance to water stress in Eucalyptus nursery plants, finding that plant height growth, root: shoot relationship, predawn leaf water potential, and instantaneous water efficiency use (iWUE) were key features in determining such resistance. These results suggest that genotype selection for greater resistance to environmental limitations may reduce mortality and growth loss risks in the early stages of plant development after planting.

Phenological development may also provide another driver for the selection of species and genotypes. Understanding species-specific and/or genotypic responses to temperature extremes and the duration of such stresses is critical for predicting plantation responses under climate change. Detailed phenological studies, such as those developed by Watt et al. [65] and Queiroz et al. [66], that evaluate optimal, minimum, and maximum performance temperature limits, or cardinal temperatures, for several tropical and Mediterranean eucalyptus species, may provide a basis for understanding key CC-related effects on growth performance of tree species and genotypes (Figure 2). Major changes in vegetation phenology have been observed in several ecosystems due to earlier warmer temperatures throughout the seasons, causing an earlier start to the growing season and increasing stand growth and carbon fixation [67,68]. However, these changes may also affect site carbon fluxes, in particular in tissues and soil respiration, and may also limit water resource availability, affecting site productivity and carbon sequestration by increasing seasonal stand evapotranspiration and water loss ahead of potential critical periods of increased atmospheric demand [69]. Temperature may also significantly affect absolute WUE, but responses may be positive or negative given specific site interactions [70].

Trees maintain relatively safe water levels through strict stomatal control; exposure to lethal water stress thresholds can increase the risk of carbon depletion and the chance of plant hydraulic failure. Sala et al. [38] and Allen et al. [71] investigated the interactions between storage and capacitance (water content × water potential) and showed the strong relationship of this interaction to tolerance to water stress and its value as a key factor in the design of plant strategies to cope with the environment [27,72]. Water storage can buffer rapid fluctuations in water status caused by severe environmental events, such as drought, and capacitance plays an important role in the efficiency and safety of water transport through the xylem [73].

Tree water stress response to cope with the inhibition of photosynthesis depends on the genetic potential of the species [74]. Indirect water use efficiency evaluation using carbon isotopes (e.g., δ C¹³, δ C¹²) has been proposed as a strategy for selection in forest breeding programs. However, investigations have shown mixed results in evaluating the response to water stress among species and genotypes. The reasons for this may be related to differential carbon partitioning mechanisms (e.g., leaf area adjustment to sapwood area, root:shoot ratio adjustment, and leaf photosynthetic capacity or respiration) that may affect this relationship [75,76]. Opportunities to forecast the risk of death for species/genotypes under CC may also require a clear understanding of a tree’s hydraulic safety margin between observed minimum xylem water potential and cavitation water potential [77]. However, the tradeoff between selecting better-adapted trees under CC requires a compromise: tolerant species or genotypes that sustain higher stomatal conductance to maintain C fixation will often lose water and risk hydraulic failure (anisohydric or drought-tolerant behavior). This contrasts with resistant species/genotypes that are very sensitive to closing stomata under drought (reducing C fixation and growth), but possess secure xylem hydraulic conductivity (isohydric or drought-avoider species or genotypes) [78].

The growth response to water stress over time and its genetic variation in Eucalytpus genotypes has been analyzed using the water stress integral concept [79,80] (Figure 3). Low productivity varieties showing higher levels of accumulated stress response over time may represent the responses of sensitive species to seasonal water stress (Figure 3a) [81]. For example, Waghorn et al. [82], researching juvenile Pinus radiata, concluded that late-season drought (higher WSI levels) is more detrimental to diameter growth and biomass accumulation compared to early-season drought (Figure 3b). The commercial use of forest materials that resist higher WSI levels may allow maximizing of production at early ages in drought-prone regions.

Water use by forest plantations has been strongly debated in the scientific community, and implications under CC are critical for ecosystem sustainability [83]. Several reports emphasize the idea that the water consumption of natural forests is lower than forest plantations [84,85] and assume a larger resilience of natural forests compared to plantations under drought events [86]. White et al. [87] showed potential effects on water ecosystem resources due to larger water consumption of perennial planted exotic species compared to natural senescent forest species during seasonal leaf-off periods. This perennial vs. senescent vegetation implies different phenological mechanisms, affecting ecosystem water balance but not vegetation responses during spring and summer’s drier months. Schwärzel et al. [7] also showed that understory, as opposed to overstory, may be an unexpected key major water consumer regulator. In situ water balance studies have shown differences among vegetation types, and a major common finding indicates that water use is a function of leaf area as a key driver, triggered by atmospheric demand [62,88,89]. This suggests that effects on ecosystem water resources are mainly physical, and forest plantations may not differ from a broad range of natural forests that sustain similar leaf area levels under the same conditions of atmospheric demand. Thus, both are highly susceptible to CC event impacts.

The effects of forest plantations on water resources, considering erosion, sedimentation, and groundwater recharge, have been summarized by van Dijk and Keenan [90]. Comparisons between forest plantations and natural forests at the watershed level have usually provided assessments that do not account for differences in past land use (e.g., well-developed soil profiles are usually maintained under natural forests). Additionally, underlying geological differences may be large under forest landscapes, creating biased comparisons [91]. Interestingly, Cassiano and Ferraz [12] showed that management of landscape forest mosaics may provide a better balance, in terms of stable annual streamflow and maintenance of wood production, at the watershed scale level. Current research studies suggest that assessment of the potential effects of plantations on water resources requires a site-specific planning approach [91,92,93]. A minimum forest cover that may provide a balance between ecosystem NPP for productive/carbon fixation purposes and water sustainability has been proposed by Tarigan et al. [94]. Hakamada et al. [95] showed practical opportunities for reducing stand WU, considering lower initial planting densities or initial stockings, and Medhurst et al. [96] also suggested a leaf area reduction approach to estimate thinning effects on stand water use. In this regard, thinning and pruning have also been suggested as silvicultural strategies for reducing potential impact on transpiration and interception, improving site water yield. However, studies have shown that the effects are transient depending on thinning intensity and final time of recovery of leaf area levels of the overstory and understory [97]. The strong relationship between the plantation leaf area and understory leaf area (site leaf area) relies on the key role of the understory in transpiration, and its direct effects on interception and soil evaporation. Nevertheless, atmospheric demand is the major driver of site water use of forests and accompanying vegetation. A comprehensive approach modeling effects on water resources, combining crop water balance and environmental variables, has been synthesized by White et al. [91] for South-East Asia. Water sustainability under CC for intensively managed plantations may consider stand management strategies, such as mosaics, reduced planting densities with or without changes in plantation design, and species or genotype selections that may provide sustained ecosystem-level effects, in order to minimize impacts.

Understanding carbon and water balance is key for understanding ecosystem sustainability under CC. The strong relationships reported between GPP and ET for forest plantations suggest a connection between carbon fixation and water use [98]. In this regard, forest plantations, similar to natural forests and other vegetation types, show a strong positive correlation between productivity and rainfall [70], but higher net ecosystem productivity rates and higher absolute water use efficiency compared to other vegetation types [98]. Interestingly, plantations have been reported to be more sensitive to drought compared to natural forests due to natural adaptation and more resistant than natural forests due to physiological adaptation advantages [70,85,99]. It is found that these responses are local due to site-specific environments when comparisons are made among studies [100,101].

A synthesis of studies accounting for annual water use and forest stand carbon annual production suggested an increase in water use when increasing the amount of C being fixed by forest plantations [91,102,103]. Comparable results analyzing forest productivity across a rainfall gradient have been reported by Zhang et al. [104] for Eucalyptus spp. in Australia, Eucalyptus spp. and Pinus spp. in Brazil [103], and Pinus radiata in Chile [102]. Worldwide studies have shown a strong linear relationship between low-productivity forests and low water use efficiency, but this relationship changes to an asymptotic behavior with larger variability in absolute water use efficiency as higher levels of site carbon fixation are reached with smaller increments in water use (Figure 4). Interestingly, the linear portion of this relationship agrees with the linear relationship between leaf area and water use evaluated for several Eucalyptus species by Hatton et al. [105], but does not explain the higher absolute WUE at higher productivity sites. This WUE relationship has several implications for climate change effects on productivity and suggests that severe and mild water-limited environments may show severe and significant reductions in yield and carbon sequestration. On the other hand, high-productivity forest sites (e.g., >6 Mg C ha⁻¹ yr⁻¹) that maintain higher WUE levels may provide sustained levels of forest productivity and carbon sequestration unless extreme shortages in water resource availability affect site productivity (Figure 4).

Selecting water-use efficient (WUE) tree species as a strategy for forest plantations has been challenged. For example, Hodge and Dvorak [107] considered that selecting species/genotypes based on their productivity in a drought environmental gradient may indirectly result in the selection of more WUE trees. However, higher productivity may also be attained via different physiological mechanisms, and more importantly, rates of water use (Figure 4). In all cases, selecting sustainable water use species/genotypes for intensively managed fast-growing forest plantations may provide an opportunity to maintain forest productivity in regions where climate change scenarios will reduce water resource availability [93,107,108,109].

Maintenance of productivity and survival may strongly depend on carbon fixation and partitioning between aboveground and belowground components [110]. Partitioning to belowground components conveys an important tradeoff that may favor survival by accessing limited water or nutrient resources, but investment in non-photosynthetic tissues may reduce carbon fixation at the expense of growth required for survival and growth rate maintenance [111]. Campoe et al. [112] evaluated the carbon balance of five clonal Eucalyptus across four contrasting sites representing a water and temperature stress gradient in Brazil and Uruguay. The study observed large differences in gross primary production across sites where belowground partitioning increased, with increasing site mean annual temperature and soil water deficit reducing aboveground wood production. Results from other studies and species have also shown comparable results on a broader scale [113]. Interestingly, in this study, clones showed similar belowground partitioning, but larger differences in partitioning regarding wood. These differences among clonal materials may provide guidance for selecting genotypes better adapted to CC scenarios and improving the parametrization of process-based models for evaluating species and genotypes [114,115]. Considering more tolerant, rather than more resistant, genotypes for selection in order to maintain productivity under CC may demand a tradeoff in terms of risk of sudden death or failure under drought.

The resistance characteristics of selected genotypes may secure survival, but this is usually at the expense of reduced growth rates, as these characteristics typically rely on the avoidance of water stress mechanisms that may reduce productivity. These mechanisms include premature foliage senesce and drop in leaf area, the development of larger and deeper root systems at the expense of growing aboveground biomass, or having faster or larger sensitivity of stomata to water deficit. On the other hand, tolerant genotypes will maintain site productivity and carbon sequestration in forest plantations, as their key physiological, morphological, and phenological mechanisms of response to water stress do not strongly affect carbon fixation or partitioning to aboveground components. However, extremely tolerant materials showing high growth rates may be susceptible to higher risks under short-term droughts (Figure 5). Appropriate matching of species/genotypes to sites is required as a strategy to reduce such risks. Additionally, species selection is critical for defining silvicultural management strategies and the plantation rotation period. For example, older P. radiata plantations are more conservative in terms of WU and show higher WUE levels compared to younger stands (Figure 5b).

Despite the complexity of carbon fluxes and balance assessments at an individual tree or stand level, understanding these responses for improved species/genotypes/clones under water availability or temperature environmental gradients, or under manipulated experimental conditions, may provide a base understanding to guide the assessment of mortality risks and prediction of potential growth rate declines and sustainability of future plantations. The implications of these key elements will provide better certainty surrounding the use of current process-based model projections and their effects on ecosystem net carbon sequestration potential, as well as the role of forest plantations at large-scale levels [116,117].

Debate on the positive effects of forest plantations on site carbon sequestration has been raised in regard to the comparison of plantation carbon gain versus the restoration of natural forests, under global forestation plans, for carbon mitigation [115,118]. Considering the relevance and caution of sustaining forest biodiversity, robust scientific studies have shown the importance of considering fast-growing forest plantations to reach high C sequestration rates, and at the same time, the importance of maintaining long-term large C stocks accumulated over decades or centuries by old and natural forests. From a soil carbon sequestration point of view, and in particular considering degraded or low fertility soils with agricultural abandonment, plantations with fast-growing species undoubtedly show faster and larger positive effects on soil C sequestration and stocks [81,119]. The C sequestration potential of a given soil depends strongly on belowground biomass growth and decay, up until reaching a physical maximum potential of organic matter accumulation associated with soil porosity. This short to medium-term soil organic carbon (SOC) accumulation potential is, as may be expected for most degraded soils, strongly linked to the maximum aboveground biomass attained at a given site for a particular soil-site condition (Figure 6).

Reforestation efforts being made around the world with fast-growing plantations need to consider climate change risks in order to achieve mitigation objectives [119,120,121,122,123]. Several reports from intensively managed forest plantations worldwide have indicated lower survival and reduced growth rates due to CC [15,124,125]. Increases in maximum and/or minimum absolute temperatures, reductions in precipitation, and extended drought periods have affected the sustained productivity of intensively managed plantations [126,127,128]. However, productivity gains have been frequently reported due to longer growing seasons, or increases in seasonal temperature, rainfall, and/or atmospheric CO₂ [129,130,131]. Interestingly, the gains in productivity observed in several reports rely on a CO₂ fertilization effect that, when removed, may result in productivity decreases at sites with water and/or nutrient resource limitations or environmental temperature constraints that may elicit unknown results [123,132]. Forest ecosystem responses under elevated CO₂ may be affected by interactions with water and nutrient availability that depend strongly on site-specific characteristics [133]. Despite there being some evidence of improved water use efficiency when nutrients are not limited, any other environmental limitation (water, nutrients, temperature, oxygen, light) at different stages of stand development could reduce the productivity and carbon sequestration potential of forest plantations, limiting the expected results under an elevated CO₂ enrichment scenario.

The selection of forest tree species for large-scale forestation planting programs targeted for climate change mitigation may provide interesting opportunities for increasing C sequestration, as well as the establishment of mixed-species plantations [120,134,135,136,137,138,139]. Silveira et al. [140] evaluated several Eucalyptus spp. and Pinus spp. based on an extensive literature review and ranked their potentials in terms of carbon sequestration and storage. Their results suggested that in 60% of all comparisons, Eucalyptus species performed better than Pinus spp., but some broadly planted pine species such as P. patula, P. halepensis, and P. radiata showed better performance in limited environments. In fact, selecting fast-growing or better-adapted species may reduce the land area required for a given target mitigation goal and/or provide larger C stocks in a shorter spam of time. Similarly, Bush et al. [139] evaluated several potential new species and sustainable product alternatives for Australia under CC scenarios, considering the impacts of biotic and abiotic stresses on survival and productivity. As a corollary of these studies and the observed CC biotic and abiotic challenges, the need to evaluate new potential species for the establishment of productively managed plantations should be considered a strategic goal of countries with highly developed forest industries and/or those considering target sequestration scenarios.

Appropriate management of plantations, and associated silvicultural activities, such as removing competing vegetation and fertilization and soil preparation, are required to maintain, improve, and sustain the productivity and carbon sequestration of forest plantations, and cement their role in mitigating climate change effects [140,141,142,143,144]. Intermediate management activities such as pruning, thinning, or variable retention harvesting may also provide opportunities for reducing severe water stress effects on plantations [145,146,147].

Understanding the physical characteristics of soils and their effects on species or genetic material may gain importance under reduced water availability CC scenarios. Several studies have shown critical differences in root system architecture that are affected by soil properties at the species, taxa, and variety levels. Soil dynamic properties, such as soil moisture and soil resistance, may impose limitations on root growth that in some instances, if not appropriately manipulated [148], may limit seasonal plant development and growth. In fact, despite most modern fast-growing stands, silviculture considers that intensive soil preparation [108], drier soil conditions, and enhancing high soil resistance to penetration may provide unknown effects on belowground plant growth.