Abstract
Background Urban trees provide many environmental benefits but often face challenging growing conditions like waterlogged soils. How tree root systems respond to waterlogging impacts tree performance and survival, yet this has received little attention. Our goal was to identify how the roots of temperate urban tree species respond and recover to waterlogging.
Methods We monitored the responses and recovery of 2 contrasting maple and magnolia species pairs that differ in their reported waterlogging tolerance to a 2-week waterlogging period, measuring belowground stress indicators, fine-root mortality, and aboveground responses including leaf-level photosynthesis, leaf loss, and stem growth.
Results Though silver maple experienced a temporary reduction in photosynthetic activity during waterlogging, it exhibited no fine-root mortality, and photosynthetic activity recovered after a 10-day recovery period. In contrast, sugar maple showed high fine-root mortality, decreased photosynthetic activity, and significant leaf loss, with no recovery in fine-root growth or photosynthetic activity after the recovery period. Both magnolia species showed high fine-root mortality and reduced photosynthesis during the waterlogging period. However, after the 10-day recovery period, both magnolias also showed new fine-root growth and increased photosynthetic activity.
Conclusion The species studied here showed a wide range of fine-root response and recovery strategies to waterlogging, and this was mirrored in their aboveground performance. Future work clarifying the mechanisms driving these different strategies, such as silver maple’s ability to maintain fine roots and mitigate internal tissue damage, will help us to further understand species differences in waterlogging tolerance and better inform urban tree selection for repeatedly flooded soils.
Introduction
Urban trees provide many ecosystem services and contribute to making urban environments environmentally sustainable, resilient, and biodiverse. Urban trees provide flood water management, capture air pollutants, reduce the urban heat-island effect, provide habitats for wildlife, and increase property values (Akbari 2005; Nowak 2006; Grote et al. 2016; Hirons and Thomas 2017). There is also strong evidence for a connection between the presence of urban trees and the health of residents. Examples of this connection include lower rates of morbidity and mental health ailments and an increased sense of community where trees are more abundant (Nowak 2006; Ulmer et al. 2016; Turner-Skoff and Cavender 2019). The above environmental, economic, and public health impacts are not sustained when urban tree health declines, and this is an important consideration as urban trees often face challenging growing conditions such as flooding, water stress, high temperatures, extreme light intensity, air pollution, mineral deficiencies, and soil compaction that often lead to premature death (Roman et al. 2016; Hirons and Thomas 2017).
Major stressors experienced by urban trees often involve water stress from drought and waterlogged soil. Drought is an important constraint on tree health in many urban areas and has received considerable research attention (e.g., Dale and Frank 2017; Wang et al. 2019; Haase and Hellwig 2022). In contrast, excess water, and specifically waterlogged soils (Hirons and Thomas 2017), is also an important factor that can limit tree growth and lead to mortality in urban trees, yet this has received much less attention. Following precipitation events in urban environments, waterlogging can be the result of poor soil drainage (often due to compaction and high clay content), sometimes arising from improper site design. When soil surfaces are inundated for long durations (e.g., days to weeks), oxygen is displaced from soil pore spaces, and this causes anoxic conditions (Baldwin and Mitchell 2000). Anoxic conditions impact soil chemical properties like pH and inhibit aerobic respiration by root systems, leading to tree stress, damage, and death (Pezeshki and Chambers 1985; Kreuzwieser and Rennenberg 2014). Furthermore, the continued rise in global temperature is projected to alter hydraulic regimes (Christensen and Christensen 2007; Kundzewicz et al. 2014). Some regions are expected to experience increased frequency and severity of flooding and waterlogging events, making identification of mechanisms driving species responses to waterlogging, and their capacity for recovery a critical focus for urban tree selection.
Tree species may demonstrate wide variation in their responses and tolerance to waterlogging conditions (Glenz et al. 2006). Variation in waterlogging tolerance is driven by species biogeography and evolutionary history, which have led to adaptive traits like lenticels and modified root tissue that enhance gas exchange in inundated soils (Buchel and Grosse 1990; Yamauchi et al. 2013). Investigations into species thresholds for waterlogging stress have identified tolerant species that can withstand months-long periods of waterlogging during the growing season (Frye and Grosse 1992; Glenz et al. 2006). In contrast, other species are noted for having very little tolerance to waterlogging, with severe damage occurring in only a few days (Glenz et al. 2006). Although we know the relative tolerance levels of many temperate tree species, we know very little about the root traits and physiological processes associated with waterlogging tolerance.
Research investigating tree responses to waterlogging has focused on aboveground responses. Well documented signs of waterlogging sensitivity include leaf necrosis, shoot die-back, and impaired growth and biomass accrual (Lopez and Kursar 2003; Kreuzwieser and Rennenberg 2014; Bhusal et al. 2022). Despite the frequent focus on aboveground tree responses to waterlogging, fine roots (i.e., the most distal portion of the root system responsible for nutrient and water uptake) are the first point of contact for waterlogging stress. Lack of oxygen impedes normal fine-root metabolic processes that drive fine- root tissue maintenance and growth (Armstrong 1980). As damage to the root system accrues, there are declines in essential functions such as nutrient and water uptake, solute transport (e.g., sugars, micronutrients, phytohormones) to root apical zones and aboveground organs, and hydraulic conductivity throughout the root system and whole plant (Saglio 1985; Schmull and Thomas 2000; Martínez-Alcántara et al. 2012; Ow et al. 2019). The decrease in hydraulic conductivity is subsequently met with a droughtlike response in aboveground organs, as described previously. This leads to impaired stomatal conductance and decreases in photosynthetic functioning that result from obstructed gas exchange and damage to other cellular photosynthetic machinery (Dreyer 1994; Herrera et al. 2008; Parent et al. 2011; Li et al. 2015; Wiström et al. 2023).
The limited research investigating belowground responses and recovery to waterlogging shows that trees undergo both structural and physiological changes to cope with waterlogging stress. Structural changes may include the growth and/or increase of adventitious roots, aerenchyma, pneumatophores, and hypertrophied lenticels (Armstrong 1980; McDonald et al. 2008; Yamauchi et al. 2013; Takahashi et al. 2014). Additionally, trees may adjust their rooting depth or form suberized epidermal tissues on the external layers of fine roots to prevent radial oxygen loss (Armstrong 1980; Colmer 2003). These adaptations increase gas exchange and allow trees to continue respiration, water and nutrient transport, photosynthesis, and growth during periods of inundation (Armstrong 1980; Col- mer 2003; McDonald et al. 2008; Yamauchi et al. 2013; Kreuzwieser and Rennenberg 2014; Takahashi et al. 2014). Importantly, structural changes generally require time to resolve and often necessitate new growth, potentially making them less viable strategies for handling acute waterlogging stress. In contrast, physiological changes can be relatively rapid, maintain energy production, and limit buildup of harmful free radicals like reactive oxygen species (ROS) that can damage cellular membranes and organelles (Kreuzwieser and Rennenberg 2014).
Studies investigating tree metabolic and physiological changes in response to waterlogging have identified different mechanisms that are responsible for some species’ high tolerance levels. Due to the inhibition of oxidative phosphorylation caused by anoxic conditions during waterlogging, aerobic respiration is decreased or completely halted (Kreuzwieser and Rennenberg 2014). Trees may cope with this by increasing glycolysis, switching to fermentative pathways, and upregulating production of enzymes to neutralize free radicals. Tolerant species increase glycolysis at higher rates and/or upregulate genes involved in fermentative pathways (Salvatierra et al. 2020; Habibi et al. 2023). As a result of increasing fermentation, by-products of anaerobic pathways (e.g., lactate, ethanol, acetates, carbon dioxide, hydrogen gas) can accumulate and cause cytosolic pH imbalances which can lead to damage of the lipids composing cell membranes (i.e., lipid oxidative damage)(Salvatierra et al. 2020; Habibi et al. 2023). It is therefore presumed that tolerant species have a way to recycle or scavenge these harmful by-products using neutralizing molecules such as antioxidants and other enzymes (e.g., peroxidase, ascorbic acid, glutathione, catalase, and superoxide dismutase)(Salvatierra et al. 2020; Habibi et al. 2023). However, the above studies have focused on a limited number of commercial fruit trees, and the mechanisms driving tree responses to waterlogging need to be investigated in a broader range of species, especially in common urban trees that may be more prone to experiencing waterlogging stress.
Our study investigates physical and physiological fine-root and whole-plant responses by subjecting 4 tree species to a waterlogging and recovery period. Interestingly, previous studies have noted strong differences in waterlogging tolerance even among closely related species and/or cultivars (Zwack et al. 1999; Parent et al. 2011; Li et al. 2015; Wiström et al. 2023). Accordingly, we selected congeneric contrasts that were expected to differ in their waterlogging tolerance based on prior observations of aboveground performance. Specifically, we used the congeners sugar maple (Acer saccharum) and silver maple (A. saccharinum ), which are considered to be strongly intolerant and strongly tolerant to waterlogging, respectively. Similarly, the congeners of saucer magnolia (Magnolia × soulangeana) and star magnolia (M. stellata) are expected to differ in their tolerance, with star magnolia considered more tolerant than saucer magnolia. Within each congeneric contrast, the 2 species possess similar fine-root traits, suggesting that root morphology alone cannot explain species differences in tolerance. In addition, preliminary results from prior waterlogging trials suggest that silver maple may employ fine-root strategies to cope with waterlogging stress that differ from strategies used by certain species of magnolia (McCormack ML, unpublished).
Our objectives were to elucidate (1) how fine roots respond and recover to waterlogging stress, and (2) how fine-root response and recovery strategies interact with aboveground responses to influence whole-tree waterlogging tolerance. After subjecting trees to a 2-week waterlogging period followed by a 7-week recovery period, we measured fine-root growth, mortality, and common stress indicators (peroxidase activity [POX] and lipid oxidative damage [LOD]) in fine-root tissues to observe possible mechanisms causing and/or mitigating fine-root tissue damage. Fine- root growth, mortality, and tissue stress were measured using tracing and imaging, dissection of fine roots, and colorimetric assays for assessing POX and LOD, respectively. Further, aboveground measurements of photosynthesis rates, leaf loss, and stem diameter were taken as additional proxies for species’ waterlogging tolerance. The species chosen for our study are common urban tree species, and understanding how their fine roots respond to and recover from waterlogging is important for proper tree selection in frequently flooded urban sites.
Materials and Methods
Plant Material and Waterlogging Procedure
Sugar and silver maple saplings were purchased as 3-year-old bareroot saplings in 2022 while saucer and star magnolia were purchased as 3-year-old potted saplings in 2020, making them 3- and 5-years-old at the start of the study. All saplings were seed grown. Prior to the start of the study, saplings were of roughly similar size and were transplanted to 7.7-L and 11.4-L nursery pots (maples and magnolias, respectively), with height × width dimensions of 23 cm × 24 cm and 27 cm × 24 cm, respectively. Pots were filled with a mix of 50% by-volume potting soil (60% pine bark fines, 20% peat moss, 10% perlite, 10% HydraFiber® EZ [Profile Producs LLC, Buffalo Grove, IL, USA]) and 50% sand mixture in 2022 April. The soil mixture was roughly 20-cm deep. Osmocote® (Bloomington Brands LLC, Bloomington, IN, USA) 19-6-12 was added at a rate of 24 g per 3.8 L of potting media to meet sapling nutrient requirements. Prior to potting, 4 flaps (11 cm × 16 cm) were cut into each pot, and clear acetate was glued into the cutouts. This allowed us to assess fine-root growth and death throughout the experiment as well as sample roots of known ages (Figure S1). Saplings were maintained in a greenhouse for one month and moved outdoors when outdoor temperatures stabilized above freezing. Nonetheless, some star magnolia and sugar maple exhibited burned leaves after being moved outdoors. Sugar maple recovered quickly, and the star magnolia recovered at a slower rate, but all plants had fully recovered prior to waterlogging initiation. Saplings were watered on a regular basis until root growth was observed in the windows and waterlogging was initiated in late July.
In total, 48 trees were used in our experiment, with 12 plants of each species divided into control and treatment groups (n = 6 per group). For the waterlogging treatment, 6 replicates of each species were placed in small (1.2 m × 1.2 m × 0.3 m) pools and water was added until pots were submerged to the soil line. The waterlogging period lasted 14 days, with control pots watered every 2 to 3 days. Waterlogged saplings were rearranged within the pools 3 times during this period to avoid potential edge effects. The pool water was not changed during the 2-week trial, but small additions of water were added to maintain complete inundation. Shade cloths blocking 90% of the sunlight were placed directly over the pools to prevent solar radiation from warming the water to high temperatures. Three replicate plants from the control and treatment groups of each species were destructively harvested immediately after waterlogging, leaving three replicates (n = 3) for measurement during the recovery period.
Belowground Measurements
Fine-Root Growth and Mortality Assessments
Fine roots (< 2 mm) of all 48 trees were imaged and traced once every 2 weeks. Tracing began approximately 4 weeks before waterlogging treatments and continued for 7 weeks following the cessation of waterlogging to monitor new fine-root growth and/or fine-root recovery. Three of the four plastic windows were designated for tracing using permanent paint pens. Paint pens of different colors were used to indicate tracing dates which could then be used to identify age classes of fine roots during harvests. Initially, 7 tracing sessions led to a total of 7 age classes. However, age classes were later pooled into 3 groups (roots produced before waterlogging period, roots produced during waterlogging period, and roots produced after waterlogging) to ensure sufficient root tissue (40 mg) for our stress test analyses. In contrast to the 3 traced windows, the fourth window was left untraced and was instead used for imaging. Images were collected using a Nikon D750 (Nikon, Minato City, Tokyo, Japan) camera with a 60-mm f/2.8-wide aperture mounted 1 meter from the pot window to allow for standardized, repeat imaging of the root window. Finally, fine-root mortality assessments were made immediately following waterlogging by randomly selecting and cutting 2 windows on opposite sides of the pot and closely inspecting individual fine roots touching the window. Visual, olfactory, and turgidity cues were used to identify dieback and decay. Roots that were dark in color, squishy, and had degraded cortical tissues were considered dead, while those with lighter color and firm, intact tissues were considered alive. Estimates from 2 windows per pot were then averaged to estimate percent fine-root mortality for each individual tree.
Stress Tests
Fine-root tissues were collected at 3 time points (immediately before and after waterlogging and 7-weeks postwaterlogging) to assess fine-root POX and LOD. During the first 2 collections, fine roots were harvested from 1 window in each of the 6 control and 6 waterlogged individuals per species (n = 6). Following the second collection, 3 individuals from each treatment were destructively harvested for analysis of nonstructural carbohydrate content (data not included here because of issues with the protocol). Therefore, the third collection of fine roots had fewer replicates per treatment (n = 3). For each collection, one plastic window was cut per pot and fine roots were grouped by age class (as described above). Using forceps and scissors, fine roots were removed carefully, cleaned with water, and flash frozen using liquid nitrogen before transferring to a –80 °C freezer where they were stored until analysis. Fine roots were briefly removed from the freezer prior to stress test extractions, when they were ground using a mortar and pestle with liquid nitrogen and weighed to a uniform target of 40 mg per sample. Each sample was then divided into 2 samples of 20 mg (1 for each stress test), and then transferred back into liquid nitrogen and a –80 °C freezer.
POX was measured following a protocol modified from MacAdam et al. (1992). Briefly, proteins were extracted from approximately 20 mg of ground fine-root tissue using an extraction buffer consisting of 50 mM Na-PO4 and 5% (w:v) polyvinylpolypyrrolidone (PVPP) at pH 6.8. An assay buffer was made using the same materials, excluding the PVPP. Extraction buffer (200 μl) was added to each sample of fine roots and sonicated in an ice bath for approximately 3 minutes before centrifuging (20,000 × g, 1 minute) to separate the protein extract from the plant material. Protein extracts were transferred to clean microcentrifuge tubes prior to analysis.
For analysis, a substrate solution was made using an assay buffer and 0.25% guaiacol and 0.375% H2O2. The solution was made fresh, protected from light, and used the same day as prepared. Assay buffer (10 μl) and substrate solution (190 μL) served as a standard. For the samples, protein extracts (10 μL) and substrate solution (190 μL) were combined in individual wells of a 96-well plate. Two replicate wells of each sample were prepared. Protein extract absorbance was measured at 470 nm using an absorbance plate reader with BioTek Gen5 Software (Agilent Technologies, Inc., Santa Clara, CA, USA). Measurements were adjusted for kinetic reading using 10-second read intervals and an incubation temperature of 30 °C.
LOD was measured using a thiobarbituric acid reactive substance assay (TBARS)(Zhang et al. 2008). An extraction solution containing 0.6% thiobarbituric acid and 10% trichloroacetic acid was prepared and 500 μl was added to microcentrifuge tubes containing 20 mg of fine-root tissue. Samples were vortexed until tissues were fully soaked, after which they were left to incubate at room temperature for 20 minutes. Following incubation, samples were centrifuged at 21,000 × g, and 400 μl of supernatant was transferred to new microcentrifuge tubes and incubated in a 95 °C oven for 30 minutes. Immediately after incubation, samples were placed in an ice bath for 5 minutes to cool. Samples were centrifuged again before analysis.
Absorbance at 532 nm was measured as described above for POX. The reaction solution without any additions served as a standard. For samples, 150-μL supernatant from the centrifuged solutions was transferred (150 μL) into 2 replicate wells for analysis.
Aboveground Measurements
Leaf Gas Exchange Measurements
Carbon assimilation (A) and stomatal conductance (gsw) were measured as indicators of gas exchange rates on 5 individual treated and control trees for each species using a LI-6800-01A Portable Photosynthesis System with a multiphase flash fluorometer (LI-COR Biotechnology, Lincoln, NE, USA). Settings were adjusted to irradiance of 1,500 μmol m-2 s-1, 420 ppm of CO2, 27.5 °C, and 70% relative humidity. Measurements were taken from 9:00 to 12:00 CST on cloudless days to limit environmental variation across measurements. Gas exchange was measured roughly every 2 weeks (9 sessions total) over the course of the experiment, but weekly around the waterlogging period. These sessions included dates immediately before the waterlogging period (the day of), during waterlogging (7 days after initiation), and after waterlogging ended (14 days after initiation). During each session, 3 mature and undamaged sun leaves from each individual were separately placed in the leaf chamber and analyzed. Measurements continued for 7 weeks after the waterlogging period ended. However, control leaves of all 4 species indicated a steady pattern of photosynthetic decline during measurements taken in September, which we attribute to progressive autumn leaf senescence. We therefore present photosynthesis-related data for 10 days following the waterlogging period to assess recovery during active growing periods only.
Stem and Leaf Characteristics
Leaf counts of control and treated individuals were recorded manually before waterlogging and after a 7-week recovery period ( n = 3 for each species). All undamaged leaves were counted, including immature leaves, to determine percent leaf reduction for each individual tree:
Stem diameter measurements of control and treated individuals were recorded manually before waterlogging and after a 7-week recovery period (n = 3 for each species). Measurement locations were 5 cm above the soil line and were marked with a paint pen. Individuals with more than 1 leading stem within the first 5-cm section had both stems marked and measured. Using the marked line as a reference point, stems were measured from 3 different positions around the circumference and averaged to get the most accurate diameter. We then calculated the percent change in stem diameter for each individual as follows:
Statistical Analysis
One nonparametric and six independent ANOVA tests were used to assess the effects of waterlogging on fine-root mortality, POX, LOD, gas exchange, leaf loss, and stem growth. Between species comparisons were also assessed, specifically focusing on comparisons within congener pairs. For fine-root mortality, control and treated individuals were compared at one time point immediately after waterlogging. The residuals of the fine-root mortality assessments did not follow a normal distribution, therefore treatment and species effects, and differences between groups, were assessed using the Kruskal-Wallis rank sum test with Dunn’s multiple comparison test as a post hoc analysis. The effects of treatment, species, and session (collection period), on POX, LOD, gas exchange rates, and leaf loss were assessed using 4 independent 3-way ANOVAs. For POX and LOD, 3 time points were compared (immediately pre- and postwaterlogging and following a 7-week recovery period). For gas exchange rates, the same time points were used except for the third point, which followed a 10-day recovery period. Finally, the effects of treatment and species on stem growth were assessed using a 2-way ANOVA. For all tests, outliers were assessed using the geom_boxplot() function within the ggplot2 package (Wickham 2016). Values above Q3 + 1.5 × IQR or below Q1 – 1.5 × IQR were removed. Assumptions of normality and homogeneity of variances of the standardized residuals were tested using Shapiro-Wilk’s test and Levene’s test, respectively. Accordingly, POX values were log transformed and leaf counts were square root transformed to better meet assumptions of normality. Any measurements made on more than one point per tree (e.g., leaf gas exchange) were considered as subreplicates and averaged to the individual tree level with individual trees then being considered as the unit of replication for all statistical analyses. All analyses were performed using R version 4.2.1 (R Foundation for Statistical Computing, Vienna, Austria).
Results
Belowground Impacts: Fine-Root Growth, Mortality, and Stress Indicators
Visual observations and images indicated that fine- root growth ceased in all waterlogged trees. Fine-root mortality was elevated after waterlogging in all species except silver maple (P < 0.05)(Figure 1, Table S1, Figure S3), which had near 0 (< 1%) mortality. In contrast, sugar maple trees exhibited high fine-root mortality after waterlogging (71%) but relatively low fine-root mortality under control conditions (3%) (Figure 1). Further, sugar maple showed no fine-root growth following the recovery period. Star and saucer magnolia exhibited high fine-root mortality during and immediately after waterlogging (78% and 82%, respectively), whereas control trees showed little to no mortality (< 1%)(Figure 1). Despite this, both magnolia species had an abundance of new fine-root growth following the 7-week recovery period. However, new fine-root growth was only observed after destructively harvesting the magnolia trees, and could not be seen in the pot windows (i.e., new fine roots were not long enough to extend to the windows)(Figure 2, Figure S2).
Fine-root mortality in control and waterlogged individuals of 4 tree species. Asterisks indicate significant (P < 0.05) differences due to treatment. Dots indicate outliers, and numbers indicate the number of individual trees for each measurement.
Apparent mortality during waterlogging and subsequent regrowth of fine roots during the 7-week recovery period. Examples of waterlogged (left) and control (right) plants are shown for saucer magnolia. Note the much smaller root system of the waterlogged plant that resulted from high-mortality rates of the fine roots during waterlogging. Also note the proliferation of new, healthy root tips emerging around the root ball. Similar amounts of new fine-root growth were observed on all waterlogged saucer and star magnolia individuals (Figure S2), but not for sugar maple (Figure S4).
POX was highly variable in all species and was not consistently nor significantly influenced by waterlogging or collection date (Table 1, Table S2, Figure 3). However, the 2 maples showed distinctly lower peroxidase levels (0.02 to 0.86 mmol mg-1 tissue) compared to magnolia species (0.35 to 4.15 mmol mg-1 tissue). Neither waterlogging nor collection date influenced LOD, but there was an interaction between waterlogging and species that impacted the waterlogged sugar maple (Table S3). Sugar maple showed a significant increase in LOD in fine roots collected after the 7-week recovery period compared to immediately after waterlogging (approximately 5.7 nmol g-1 tissue and approximately 3.2 nmol g-1 tissue, respectively; Figure 3). However, only one waterlogged sugar maple replicate (of fine roots) was collected after the recovery period because most fine roots were dead at this stage of the experiment.
Peroxidase activity (POX) and lipid oxidative damage (LOD) in control and waterlogged fine roots of 4 tree species measured immediately after a 2-week period of waterlogging and again after a 7-week recovery period (9 weeks total). Significant differences between control and treatment for a given species and sample date are indicated by an asterisk (P < 0.05). Dots indicate outliers, and numbers indicate the number of individual trees for each measurement, respectively.
Summary of P-values from analysis of variance for measurements of fine-root stress (POX, LOD) and leaf and stem characteristics between control and waterlogged individuals for all species. Numbers in bold are significant at P < 0.05. Samples for POX and LOD were collected immediately after waterlogging, while numbers associated with assimilation and stomatal conductance signify measurements taken immediately after waterlogging (e.g., Assimilation-1 and Conductance-1) and following a 10-day recovery period (e.g., Assimilation-10 and Conductance-10). POX (peroxidase activity); LOD (lipid oxidative damage).
Aboveground Impacts: Gas Exchange, Leaf Loss, and Stem Growth
Waterlogging significantly reduced carbon assimilation and stomatal conductance in all species (Table 1, Table S4, Table S5). Following waterlogging, maple species showed similar reductions to each other in assimilation (71% to 73%) but were more variable in stomatal conductance (59% to 78%)(Figure 4). After the 10-day recovery period, only silver maple’s photosynthetic rates recovered to the extent that they were similar to those in the control group. Sugar maple was the only species that continued to decline until the end of the experiment (Figure 4). At the end of the recovery period, assimilation rate and stomatal conductance reductions for sugar maple were 93% and 84%, respectively. The 2 magnolia species showed similar reductions in assimilation (88% to 93%) and stomatal conductance (92% to 96%) after the 2-week waterlogging period (Figure 4). Only star magnolia’s stomatal conductance recovered to the extent that they were like those in the control group after the 10-day recovery period.
Reductions in carbon assimilation and stomatal conductance for waterlogged individuals compared to control individuals measured immediately after a 2-week period of waterlogging and again after a 10-day recovery period (3.5 weeks total). Error bars represent standard error. Asterisks indicate significant (P < 0.05) differences between the treatment and control for a given species and sample date.
Waterlogging significantly increased leaf loss in all species but silver maple and had no effect on stem diameter (Table 1, Table S6, Table S7). Waterlogged maple species showed a range of responses in leaf loss, with sugar maple trees exhibiting strong leaf loss and silver maple exhibiting slight leaf loss (Figure 5). Waterlogged saucer magnolia trees exhibited high leaf loss, but star magnolia had notably greater leaf loss in the control individuals compared to the other species (Figure 5, Table S6). This was likely the result of leaf burning during the transition from the greenhouse to outdoors. There was no significant effect of waterlogging on stem diameter. However, both control and waterlogged silver maple trees had significantly more stem growth than all other species (Figure 5, Table S7).
Leaf loss and stem growth of control and waterlogged trees of 4 tree species. Trees were measured before a 2-week waterlogging period and again after a 7-week recovery period. Asterisks indicate significant (P < 0.05) differences between the treatment and control trees for a given species.
Discussion
Our study aimed to investigate the fine-root responses and recovery of tree species with different degrees of waterlogging tolerance by measuring fine-root growth, mortality, and tissue stress in saplings before, during, and after a 2-week waterlogging trial and a 7-week recovery period. We also incorporated whole-tree responses and recovery by taking aboveground measurements of gas exchange, leaf loss, and stem growth throughout the experiment. Our observations of fine-root growth and mortality showed no fine-root death in silver maple during the waterlogging period, indicating that this species is well adapted to waterlogged soil conditions. This observation is consistent with previous reports (Niinemets and Valladares 2006) and reflects its natural habitat in moist bottomlands and riparian zones prone to flooding (Sargent 1922; Saeki et al. 2011). The other 3 species in this study all experienced near-complete fine-root mortality and cessation of fine-root growth during the waterlogging period, a reported consequence of waterlogging (Fujita et al. 2020; Repo et al. 2020). Specifically, sugar maple exhibited high root mortality and no fine-root growth even after the 7-week recovery period, highlighting the variation in waterlogging tolerance between congeners. The magnolia species also showed high root mortality, but their postwaterlogging fine-root growth contrasted sharply with that of silver and sugar maple. Both magnolia species showed substantial fine-root growth during the 7-week recovery period, which may indicate unique belowground strategies (involving fine-root production poststress) between genera in response and recovery to anoxic conditions. Further, for the star magnolia, the increased fine-root production also coincided with recovery of stomatal conductance aboveground. This indicates that, despite severe declines in metabolic activity during waterlogging, some magnolia species may be able to recover from waterlogging by rapidly rebuilding root systems following reoxygenation. Species relying on different strategies, such as the silver maple and the magnolia species in our study, may therefore show variation in their ability to persist through acute, chronic, or repeated periods of waterlogging.
For additional metrics of fine-root responses and recovery to waterlogging, we focused on 2 physiological indicators of oxidative stress in root tissues. First, we investigated whether waterlogging led to an increase in POX, a very common antioxidant mechanism in plants. Increased POX would suggest that the root tissues were experiencing increased oxidative stress as a result of the anoxic waterlogged conditions. We did not find strong evidence to suggest that POX is significantly greater in waterlogged roots of these species, and these results were somewhat surprising, as elevated POX is a nearly ubiquitous response in stressed plants (Welinder 1992; Jaiswal and Srivastava 2018). This finding was particularly interesting for silver maple, as this species maintained fine roots during waterlogging and would have presumably shown signs of anoxic stress and responses to cell-damaging molecules like ROS during this period. However, our data show that scavenging of ROS via peroxidase does not appear to be a major strategy used to mitigate fine-root tissue damage in this species, and alternative enzymatic and nonenzymatic antioxidant mechanisms should be investigated (e.g., ascorbic acid, glutathione, catalase, and superoxide dismutase)(Salvatierra et al. 2020; Habibi et al. 2023).
The second physiological indicator we measured was LOD (i.e., TBARS), which provides a quantitative measure of oxidative damage to cellular membranes. In agreement with our results from the peroxidase test, we did not find strong evidence to suggest that LOD is significantly greater in the waterlogged roots of these species. The only species that showed significant differences in LOD between treatments was sugar maple, which exhibited higher oxidative damage following the 10-day recovery period compared to the period immediately following waterlogging. This suggests that oxidative damage was still occurring in fine roots after waterlogging ended. The lack of significant results in all other cases was initially surprising. However, it should be noted that we sought to compare only living roots between treatments with our sampling approach. In the cases of the magnolia and sugar maple trees, most roots were either dead or dying at the end of the waterlogging period and were therefore not sampled. By only selecting the few remaining viable fine roots, samples were biased towards tissue which likely had the lowest cellular membrane damage. Future studies may consider a midtreatment harvest (e.g., after 3 to 7 days) to assess fine-root stress levels to better identify physiological responses.
Photosynthetic rates of all species declined sharply at the initiation of waterlogging, but one species from each congeneric pair (silver maple and star magnolia) was able to recover to levels exhibited under control conditions. Silver maple’s ability to recover was likely aided by their maintenance of a functioning root system throughout the waterlogging period. This rapid recovery of photosynthetic activity may also explain how silver maple was able to roughly double their stem diameter in the single growing season, despite the waterlogging stress. In contrast, photosynthetic rates of the sugar maple trees continued to decline throughout the postwaterlogging recovery period, possibly associated with the lack of significant production of new fine roots postwaterlogging. Alternatively, the continued impairment in sugar maple’s photosynthetic activity could be due to waterlogged-induced damage of leaf tissue (Shi et al. 2023), or the persistence of long lasting chemical signals (e.g., abscisic acid) preventing stomata from opening even after waterlogging ceased (Christmann et al. 2013). Ultimately, the inability of sugar maple to produce new fine roots following waterlogging may have been due to a depletion of, or an inability to remobilize or catabolize, nonstructural carbohydrate reserves (Camisón et al. 2020). The photosynthetic rates of both waterlogged magnolia species increased postwaterlogging, but only star magnolia’s stomatal conductance recovered to control rates postwaterlogging. The ability of magnolia species to resume photosynthetic activity following waterlogging is likely supported by the new fine-root growth we observed but was also likely influenced by their maintenance of functioning leaves (data not shown).
Two contrasting fine-root response and recovery strategies to waterlogging were observed in our study. Each strategy was related to postwaterlogging aboveground performance. Silver maple trees, as excellent tolerators belowground, were able to recover photosynthetic functioning quickly after waterlogging, likely due to their ability to maintain a mostly intact fine- root system. In contrast, species like the star magnolia recovered more slowly following a steady flush of new fine-root growth postwaterlogging. Previous research testing the root response and recovery strategies of various species (e.g., Quercus spp., Acer spp., Pinus spp.) to stressful site conditions (e.g., waterlogging, physical breaking, drought, and fire) also report differences in species’ capacities for fine-root regeneration (Meinen et al. 2009; Montagnoli et al. 2016; Fujita et al. 2020; Repo et al. 2020). However, the mechanisms underlying the fine-root responses of these different species are not always clear. Additionally, the studies mentioned here, as well as our own, used young trees to test the effects of stressors on root growth, and it is known that tolerance thresholds to certain stressors can shift with age (Niinemets 2010). Therefore, more research on the physiological drivers of fine-root responses and recovery, and experiments evaluating response and recovery strategies of trees spanning a wider age range than used here, will help with species selection for urban sites experiencing different stressors and soil conditions.
Conclusion
Understanding how the fine roots of different tree species employ various strategies to respond to and recover from waterlogging is a key step in urban tree selection. Researchers should consider experiments subjecting species that both maintain and lose their fine-root systems during inundation to repeated and standalone waterlogging events to see how different belowground strategies influence whole-tree tolerance to waterlogging when temporal factors are manipulated. For example, if it is shown that species like magnolia have only enough carbohydrate reserves to recover from occasional flooding events, it would not be the best practice to plant them in a seasonally flooded site. Specific questions may address the ability of species with different strategies to recover from repeated events, longer durations of waterlogging, and how such strategies are impacted by the timing of waterlogging (i.e., the interplay between fine-root phenology and season of waterlogging). Finally, communication between researchers and green industry professionals regarding proper tree and site selection can also aid in the survival of our urban trees and maximize the ecosystem services they provide. For example, lists of species’ flooding tolerances ranging from “very low” to “very high” (such as that provided in Glenz et al. 2006) are already utilized by some arborists. Future studies can work to elucidate where other species exist on this spectrum, and arborists can aim to be specific about site conditions (i.e., how often and for how long sites are typically flooded) before selecting species to plant. Through continued focus on the belowground responses and recovery strategies of trees to waterlogging (which significantly varied by species in our study), and collaboration between researchers and green professionals, we can help care for our urban trees and contribute towards the sustainability, resilience, and biodiversity, of our cities.
Conflicts of Interest
The authors reported no conflicts of interest.
Acknowledgements
This research was funded by the National Science Foundation, award #1851961, and The Morton Arboretum. We would like to thank the following members and volunteers in the Root Lab at The Morton Arboretum for their contributions to this publication: Newton Tran, Sarah Romy, Renee Ramos, and Isabella Vergara, for their help with root tracing, imaging, and leaf measurements throughout the duration of our experiment; Bill Prescott and Don Brown, for taking leaf counts and helping with pot construction; finally, Veta Bonnewell, for assisting in data analysis.
Appendix
Effects of treatment and species on fine-root mortality assessed with Kruskal-Wallis test.
Results from 3-way ANOVA assessing the main and interactive effects of treatment, species, and root age on POX (peroxidase activity) in fine-root tissue.
Results from 3-way ANOVA assessing the main and interactive effects of treatment, species, and root age on LOD (lipid oxidative damage) in fine-root tissue.
Results from 3-way ANOVA assessing the main and interactive effects of treatment, species, and session (collection date) on leaf carbon assimilation.
Results from 3-way ANOVA assessing the main and interactive effects of treatment, species, and session (collection date) on stomatal conductance.
Results from 3-way ANOVA assessing the main and interactive effects of treatment, species, and session (collection date) on leaf loss.
Results from 2-way ANOVA assessing the main and interactive effects of treatment and species on stem growth.
“Rhizopots” were constructed from standard nursery pots: four flaps (11 cm × 16 cm) were cut into each pot, and clear acetate was glued into the cutouts. Green and yellow colors indicate different tracing sessions; differentiating in this way allowed us to monitor root growth and sample roots produced before and after waterlogging. When pots were not being actively traced or imaged, the flaps were closed, and each rhizopot was placed inside a second, uncut pot to prevent light penetration through the windows.
Apparent mortality during waterlogging and subsequent regrowth of fine roots during 7-week recovery period. Examples of control (left) and waterlogged (right) plants are shown for star magnolia, silver maple, and sugar maple, respectively. Note the substantially smaller root system of the waterlogged star magnolia and sugar maple that resulted from high mortality of fine roots during waterlogging (see Figure 1). However, also note the proliferation of new, healthy root tips emerging around the root ball of the star magnolia, and the low root mortality in the waterlogged silver maple.
- © 2024 International Society of Arboriculture
Literature Cited
Jump to section
Related Articles
Cited By...
- No citing articles found.