Introduction
Common reed [Phragmites australis (Cav.) Trin. ex Steud.] is an aggressive invader of wetlands and anthropogenically disturbed areas in the United States (Amsberry et al. Reference Amsberry, Baker, Ewanchuk and Bertness2000; Brisson et al. Reference Brisson, de Blois and Lavoie2010; Saltonstall Reference Saltonstall2002). Phragmites australis is a perennial rhizomatous grass that forms dense monocultures, allowing it to displace native plant communities and have ecosystem engineering effects on carbon storage, nutrient cycling, and hydrology (Peter and Burdick Reference Peter and Burdick2010; Rooth et al. Reference Rooth, Stevenson and Cornwell2003; Windham and Lathrop Reference Windham and Lathrop1999; Windham and Meyerson Reference Windham and Meyerson2003). This species is present on every continent except Antarctica and is separated into distinct phylogenetic lineages, which are further divided into haplotypes based on sequences of chloroplast DNA (Kettenring et al. Reference Kettenring, de Blois and Hauber2012; Saltonstall Reference Saltonstall2002, Reference Saltonstall2003). Exotic haplotypes have been problematic throughout much of the United States during the past 150 yr but have only recently presented management concerns in the state of Florida.
A cryptogenic Gulf Coast lineage with haplotype I (referred to here as the Gulf Coast type) is often considered to be native to the region; however, genetic testing suggests that this type is instead a hybrid between P. australis and the South American Phragmites mauritianus Kunth (Lambertini et al. Reference Lambertini, Mendelssohn, Gustafsson, Olesen, Riis, Sorrell and Brix2012). Regardless of origin, this haplotype has undergone a range expansion in recent years, possibly due to anthropogenic disturbance (Meyerson et al. Reference Meyerson, Lambert and Saltonstall2010). There is also an introduced lineage comprising four haplotypes, the most common of which is haplotype M (referred to here as the Eurasian type) (Kettenring et al. Reference Kettenring, de Blois and Hauber2012; Saltonstall Reference Saltonstall2002, Reference Saltonstall2003). This type, introduced to New England during the mid-1800s, was identified in Florida for the first time in 2013 (Overholt et al. Reference Overholt, Sowinski, Schmitz, Hunt, Larkin and Fant2014). Although this population has been eradicated, proximal populations of the Eurasian type in neighboring states will likely continue range expansion into Florida over the coming years, potentially co-mingling with the Gulf Coast type (Williams et al. Reference Williams, Hanson, Diaz and Overholt2012).
Management Implications
As the climate changes, land managers will need to consider the effects of warming and increased [CO2] on plant invasions. Certain species, such as Phragmites australis, are likely to see increased growth and range expansions under these conditions. In addition, changes in growth and physiology that can occur under increased [CO2] and temperature may render current management strategies, most importantly herbicide use, less effective. In this study, both the Gulf Coast and Eurasian haplotypes of P. australis showed increased growth under our elevated climate treatment (based on moderate projections for the year 2100), supporting previous findings that P. australis invasions are likely to become more problematic by the end of the century. For managers, it is critical to prioritize prevention of new P. australis invasions and eradication of those that already exist before these effects of climate change occur.
Chemical control using glyphosate was not as effective for the Gulf Coast type under our elevated climate treatment, and increased belowground biomass under these conditions indicates superior ability to regenerate following herbicide application. Effects of our climate treatments on management of the Eurasian type were negligible. The Gulf Coast type is often considered to be native but can become aggressive in disturbed habitats; managers should closely monitor this haplotype and take action as needed to prevent it from displacing desirable species. In the future, managers will need to be aware of possible changes in herbicide efficacy for P. australis and possibly other species as climatic conditions continue to change.
Climate change is projected to have major impacts on invasive species (Hellmann et al. Reference Hellman, Byers, Bierwagen and Dukes2008). Increased hydrologic fluctuations are predicted to occur in rivers and freshwater wetlands, exposing large areas of land for P. australis seed germination (Tougas-Tellier et al. Reference Tougas-Tellier, Morin, Hatin and Lavoie2015). The direct effects of increased [CO2] and temperature may also expand P. australis invasions; Eller et al. (Reference Eller, Lambertini, Nguyen and Brix2014) found that two exotic haplotypes in the Mississippi Delta region demonstrate increased salinity tolerance when grown under conditions of elevated [CO2] and temperature, which would allow the species to advance further into coastal salt marshes.
Phragmites australis haplotypes have demonstrated a high phenotypic plasticity in response to altered environmental conditions, indicating an ability to rapidly adapt to changes in climate (Eller and Brix Reference Eller and Brix2012; Mozdzer and Megonigal Reference Mozdzer and Megonigal2012). Increases in atmospheric [CO2] can have a significant impact on the growth and physiological processes of plants through a “fertilization effect,” particularly for C3 species such as P. australis that photosynthesize at suboptimal CO2 concentrations under current climate conditions (Eller et al. Reference Eller, Lambertini, Nguyen and Brix2014; Ziska and McConnell Reference Ziska and McConnell2016; Ziska and Teasdale Reference Ziska and Teasdale2000). Plant response to elevated atmospheric [CO2] can include increases in biomass production, water-use efficiency, photosynthesis, and altered leaf traits such as specific leaf area and stomatal density (Erickson et al. Reference Erickson, Megonigal, Peresta and Drake2007; Manea et al. Reference Manea, Leishman and Downey2011; Ziska and Teasdale Reference Ziska and Teasdale2000). Moreover, levels of plasticity vary between haplotypes, suggesting a differential response to climate change.
A growing body of research suggests that the physiological changes brought about by altered climate conditions may lessen the efficacy of herbicides on invasive plant species (Manea et al. Reference Manea, Leishman and Downey2011; Ziska Reference Ziska2010; Ziska and Teasdale Reference Ziska and Teasdale2000; Ziska et al. Reference Ziska, Faulkner and Lydon2004). This may happen through a number of mechanisms, such as altered leaf characteristics reducing herbicide uptake (decreased stomatal number, altered leaf thickness, etc.) or increased belowground biomass allowing for quicker regeneration from rhizomes following herbicide applications (Ziska and George Reference Ziska and George2004; Ziska et al. Reference Ziska, Faulkner and Lydon2004). However, there have been instances in which climate change impacts on plant growth have not yielded differences in response to herbicide (Marble et al. Reference Marble, Prior, Runion and Torbert2015). This makes it necessary to evaluate these effects on an individual species basis.
Over the next century, it seems likely that P. australis will undergo a range expansion in Florida and the Gulf Coast region that may be exacerbated by changes in atmospheric [CO2] and temperature. Given that herbicides are often the most effective and common method of controlling P. australis (Derr Reference Derr2008; Martin and Blossey Reference Martin and Blossey2013), it is important to study how climate change might impact chemical control efforts. Here, we had two main objectives: (1) to evaluate the growth response of two P. australis haplotypes (I and M) to simulated climate change conditions and (2) to evaluate the response of these haplotypes to a commonly used herbicide, glyphosate, under current and projected climate conditions. We hypothesized that the two haplotypes would show differential response to climate treatments and that plants grown under projected conditions would be less affected by the herbicide application.
Materials and Methods
Plant Material and Growth Conditions
Rhizome segments of the Gulf Coast type were collected from Lake Jesup, FL, USA. Leaf tissue samples from this population were assayed using the PCR-RFLP described by Saltonstall (Reference Saltonstall2003) for haplotype confirmation. For the Eurasian type, rhizome segments were obtained from the population in Lake Seminole, FL, USA, that was identified and sequenced by Overholt et al. (Reference Overholt, Sowinski, Schmitz, Hunt, Larkin and Fant2014). All plants were maintained in a common garden environment at a greenhouse in Gainesville, FL, for more than a year before initiation of this experiment.
Two- to three-node rhizome segments were planted in commercial potting soil (Professional Growing Mix, Sun Gro Horticulture Canada, Agawam, MA, USA) with slow-release fertilizer (Osmocote® Plus 15-9-12, Scotts Miracle-Gro, Marysville, OH, USA) and grown in climate-controlled greenhouse chambers set to one of two treatments: (1) an ambient climate treatment of 400 ppm atmospheric [CO2] and temperature of 32/21 C (actual values of 457.0±1.2 ppm [CO2], 33.1 ± 0.2/22.3 ± 0.2 C as recorded during the experiment); or (2) an elevated climate treatment of 650 ppm atmospheric [CO2] and temperature of 35/24 C (actual values of 651.1 ± 0.15 ppm [CO2], 35.6 ± 0.1/24.1 ± 0.04 C as recorded during the experiment). All chambers were maintained under a 14-h photoperiod. The elevated climate treatment was chosen to reflect projected springtime temperatures in the southeastern United States during the year 2100, based on representative concentration pathway 4.5 from the latest Intergovernmental Panel on Climate Change report (IPCC 2013). As there can be an interactive effect of CO2 and temperature on plants (as demonstrated by Eller et al. Reference Eller, Lambertini, Nguyen, Achenbach and Brix2013), and these two factors are projected to increase simultaneously, we focused on their combined effects rather than individual effects. There were two greenhouse chambers per climate treatment, each with 14 pots per haplotype.
Initial Measurements
After 6 wk of growth, initial measurements were made on morphological and physiological traits. For each plant, height (cm) and stem number were recorded. Photosynthetic measurements were made on the second-highest fully extended leaf of the tallest culm, using an LI-6400 XT infrared gas analyzer (IRGA; Li-Cor Biosciences, Lincoln, NE, USA). Measurements were made on 4 plants per haplotype, per greenhouse chamber. Photosynthesis (A) versus leaf intercellular [CO2] (C i) curves were plotted for each plant, with a light level of 1,800 µmol m−1 s−1, relative humidity of 50 ± 10%, and a flow rate of 500 µmol s−1. Block temperature was adjusted to the temperature of the room (32 C for the ambient treatment, 35 C for the elevated treatment). Plots were used to solve for V cmax (the maximum carboxylation rate of Rubisco) and J max (the maximum electron transport rate of RuBP regeneration) following the methods of Farquhar et al. (Reference Farquhar, Caemmerer and von, Berry1980), with measurements corrected to 25 C (per Bernacchi et al. [Reference Bernacchi, Singsaas, Pimentel, Portis and Long2001, Reference Bernacchi, Pimentel and Long2003] and Long and Bernacchi [Reference Long and Bernacchi2003]). Stomatal conductance (g s) and leaf transpiration (E) were autologged by the LI-6400 XT as it recorded the A–C i curves; values recorded at 400 ppm CO2 were used for data analysis. The second-tallest fully extended leaf was then harvested from all plants, and leaf area was measured using an LI-3100C Area Meter (Li-Cor Biosciences). Leaves were oven-dried at 60 C for 72 h and weighed. Leaf area and leaf weight were then used to calculate specific leaf area (SLA).
Herbicide Application and Final Measurements
Following initial measurements, herbicide applications were made on 7 plants per haplotype, per climate chamber (14 total per climate treatment). Glyphosate (560.4 g ae ha−1) and a nonionic surfactant (0.25% v/v) were applied using a backpack sprayer at a rate of 187.1 L ha−1. This sublethal rate of glyphosate was chosen to allow detection of subtle differences in haplotype response to treatment. Height (cm), stem and lateral branch number, and visual injury (%) were recorded 30 d after treatment (DAT). Aboveground biomass was then harvested, oven-dried at 60 C for 72 h, and weighed. Plants were allowed to regrow for an additional 30-d period, after which height, stem and lateral branch number, and above- and belowground dried biomass were measured.
Data Analysis
Data from each haplotype were analyzed separately using RStudio v. 1.0.136 (RStudio, Boston, MA, USA). An ANOVA was used to determine the effect of climate treatment on initial growth and photosynthetic characteristics, with mean separation at P < 0.05. A two-way ANOVA was used to determine the effect of herbicide application and climate treatment on measured characteristics at 30 DAT and 60 DAT. Residuals were tested for all model assumptions, and data were subjected to logarithmic transformation when necessary. Each pair of climate chambers was considered an experimental run. There was no significant run effect (P ≥ 0.05), so data were pooled between experimental runs.
Results and Discussion
Initial Growth Response
For the Gulf Coast type (Table 1), there was no effect of climate treatment on height, stem number, SLA, or on the photosynthetic traits Vcmax and Jmax. Although there was no significant difference in specific leaf area between treatments, there was an effect of climate on leaf area, with plants having smaller leaves when grown under elevated conditions (38.38 ± 2.12 cm2 compared with 44.35 ± 1.8 cm2 for the elevated and ambient climate treatments, respectively). Plants had higher gs under elevated climate conditions (127.4 ± 33.86 and 196.26 ± 79.2 mmol H2O m−2 s−1 for ambient and elevated climate treatments, respectively), as well as higher E (3.65 ± 1.28 and 5.96 ± 2.04 mmol H2O m−2 s−1 for ambient and elevated climate treatments, respectively). This response is unusual in elevated [CO2] conditions; typically, leaf area (as well as overall biomass production) for C3 species is higher when atmospheric [CO2] is increased, while gs and E decrease (Ainsworth and Rogers Reference Ainsworth and Rogers2007; Morison and Gifford Reference Morison and Gifford1984). However, these results are consistent with the effects of increased temperature. Plants under heat stress have been shown to produce smaller leaves, and in well-watered conditions often increase gs and E to regulate leaf temperature through evaporative cooling (Crawford et al. Reference Crawford, McLachlan, Hetherington and Franklin2012; Murata and Mori Reference Murata and Mori2014). These responses may have implications for herbicide efficacy; for example, a lower leaf area can limit the amount of foliar-applied herbicide that is taken up by plants.
Table 1 Initial differences between plants of the Gulf Coast type of Phragmites australis grown under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments.a
a Means with SEs are displayed for height, stem number, leaf area, specific leaf area (SLA), the maximum carboxylation rate of Rubisco (Vcmax), the maximum electron transport rate of RuBP regeneration (Jmax), stomatal conductance (gs) at 400 ppm CO2, and leaf transpiration (E) at 400 ppm CO2. Statistically significant values in bold:
* P < 0.05.
For the Eurasian type (Table 2), Vcmax was significantly higher under ambient climate conditions (87.43 ± 5.11 compared with 71.64 ± 4.92 µmol m−2 s−1 for ambient and elevated climate treatments, respectively). The same relationship was found for Jmax (191.76 ± 16.68 compared with 147.78 ± 8.99 µmol m−2 s−1 for ambient and elevated climate treatments, respectively). There was no effect of climate on any other measured physical or physiological traits for this haplotype. Again, this is an unusual response for a C3 species under increased atmospheric [CO2]; typically, these values both increase under elevated climate conditions, and a previous study has demonstrated this with the Eurasian haplotype of P. australis (Eller et al. Reference Eller, Lambertini, Nguyen and Brix2014). Increased atmospheric [CO2] can exacerbate nitrogen-deficiency symptoms and lower Vcmax and Jmax, and it is possible that this occurred in our study (Miglietta et al. Reference Miglietta, Giuntoli and Bindi1996). However, lowered photosynthetic capacity can also result from exposure to high temperatures. Increased atmospheric [CO2] can sometimes mitigate the effects of heat stress, although for some species this is not the case (Wang et al. Reference Wang, Bunce, Tomecek, Gealy, McClung, McCouch and Ziska2016; Yu et al. Reference Yu, Chen, Xu and Huang2012).
Table 2 Initial differences between plants of the Eurasian type of Phragmites australis grown under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments.a
a Means with SEs are displayed for height, stem number, leaf area, specific leaf area (SLA), the maximum carboxylation rate of Rubisco (Vcmax), the maximum electron transport rate of RuBP regeneration (Jmax), stomatal conductance (gs) at 400 ppm CO2, and leaf transpiration (E) at 400 ppm CO2. Statistically significant values in bold:
* P < 0.05.
Previous research on P. australis response to climate change has found significant differences in growth and photosynthesis under altered climate regimes (Caplan et al. Reference Caplan, Wheaton and Mozdzer2014; Eller et al. Reference Eller, Lambertini, Nguyen and Brix2014; Mozdzer and Megonigal Reference Mozdzer and Megonigal2012). Mozdzer and Megonigal (Reference Mozdzer and Megonigal2012) found that both the Eurasian type and a native North American haplotype exhibited increased productivity due to elevated atmospheric [CO2] and simulated nitrogen pollution (although in this study there was not a temperature treatment), with the Eurasian type showing a greater overall response than the native. In addition, they showed the Eurasian type to have reduced specific leaf area and increased nitrogen productivity under elevated [CO2] conditions. In this study, we found few initial differences between plants grown in ambient (400 ppm CO2, 32/21 C) or elevated (650 ppm CO2, 35/24 C) climate conditions for either haplotype.
It is possible that our [CO2] treatment was not high enough to overcome the effects of temperature. Eller et al. (Reference Eller, Lambertini, Nguyen and Brix2014) also included increased temperature in their elevated climate treatment, but the magnitude of change between climate treatments was different than in ours (+310 ppm [CO2] and +5 C in their study compared with +250 ppm [CO2] and +3 C in ours). We chose a moderate climate projection for the year 2100; it is possible that had our climate treatment been more severe, we would have seen more differences related to the effects of increased [CO2]. However, these results suggest that P. australis response to moderate changes in climate may be largely determined by temperature rather than [CO2].
Initial Herbicide Response
Overall, initial response to glyphosate was limited for both haplotypes at 30 DAT. Injury rates were low, which is indicative of the low rate of glyphosate used in this study. Despite the low injury, there was still a significant effect of glyphosate application on most measured traits for both haplotypes (Table 3). For the Gulf Coast type, plants treated with glyphosate were significantly shorter, had fewer stems, lower aboveground biomass, and higher injury ratings than untreated plants regardless of climate treatment (Figure 1). For the Eurasian type, plants treated with glyphosate had higher injury ratings and lower height than untreated plants; however, there was no effect of glyphosate on aboveground biomass, and stem number was significantly increased by herbicide application at 30 DAT (Figure 2). These data suggest that the Gulf Coast type was more susceptible to glyphosate than the Eurasian type at 30 DAT. This is in contrast to a previous study by Cheshier et al. (Reference Cheshier, Madsen, Wersal, Gerard and Welch2012), which found no significant differences in response to herbicide treatment between these two haplotypes; this may be due to differences in application rates between the two studies. Low rates of glyphosate have been shown to stimulate growth in certain plants, and the low rate used in this study may have had this effect on the Eurasian type (Velini et al. Reference Velini, Trindade, Barberis and Duke2010).
Table 3 F-ratios resulting from a two-way ANOVA of plant traits at 30 DAT, with glyphosate treatment (0.57 kg ae ha−1) and climate treatment (ambient: 400 ppm CO2, 32/21 C; or elevated: 650 ppm CO2, 35/24 C) as factors.a
a Results are shown for both the Gulf Coast and Eurasian haplotypes of Phragmites australis. Statistically significant values in bold:
* P < 0.05;
** P < 0.01;
*** P < 0.001.
Figure 1 The response of the Gulf Coast type of Phragmites australis to glyphosate application at 30 d after treatment under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments. Displayed values represent the mean and SE of (A) height, (B) visual injury, (C) stem number, (D) lateral branch number, and (E) aboveground biomass.
Figure 2 The response of the Eurasian type of Phragmites australis to glyphosate application at 30 d after treatment under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments. Displayed values represent the mean and SE of (A) height, (B) visual injury, (C) stem number, (D) lateral branch number, and (E) aboveground biomass.
For the Gulf Coast type, there was no effect of climate on most measured plant traits (Table 3). There was a significant effect of climate on lateral branch number posttreatment, as well as a significant interaction between climate and herbicide application (Table 3). Only plants treated with glyphosate produced lateral branches, and those grown under elevated climate conditions produced significantly more (21.1 ± 2.8) than those under ambient conditions (13.5 ± 1.3) (Figure 1D). Similarly, there was a significant effect of both climate and glyphosate application (as well as their interaction) on lateral branch production for the Eurasian type (Table 3). Unlike the Gulf Coast type, however, Eurasian type plants treated with glyphosate produced more lateral branches when grown under the ambient climate treatment (6.4 ± 1.4 compared with 2.7 ± 0.7 for ambient and elevated climate conditions, respectively) (Figure 2D). There was a significant effect of climate and glyphosate application on plant height as well, although the interaction was not significant; plants grown under the elevated climate treatment were taller than those under the ambient treatment, and plants treated with glyphosate were shorter than untreated plants (Figure 2A).
Lateral branching is a common symptom of sublethal glyphosate applications, indicating that for the Gulf Coast type, plants under the elevated climate conditions were showing a greater initial response than those in the ambient conditions. This could have one of two causes. First, glyphosate is translocated to areas of active growth; plants grown under the elevated climate conditions tended to produce more stems and had greater aboveground biomass than those in the ambient conditions (although this was not significant at 30 DAT), and it is possible that increased growth stimulated movement of the herbicide through the plant. Alternatively, this response could have resulted from heat stress; increased temperature has been correlated with increased herbicide efficacy in certain species (Hammerton Reference Hammerton1967; Johnson and Young Reference Johnson and Young2002). If the primary effect of the elevated climate treatment was due to increased temperature rather than [CO2], it could have also resulted in increased herbicide symptoms.
For the Eurasian type, lateral branching was also increased by glyphosate, but to a lesser extent than for the Gulf Coast type. Additionally, lateral branching for this type was greater under ambient conditions. This differential response could be linked to photosynthesis. Photosynthesis affects movement of systemic herbicides such as glyphosate, and we observed lowered photosynthetic rates in the Eurasian type under elevated climate conditions; this may result in slower or more limited herbicide efficacy under these conditions (Varanasi et al. Reference Varanasi, Prasad and Jugulam2016; Waltz et al. Reference Waltz, Martin, Roeth and Lindquist2004). It may also be that for P. australis, the Eurasian type is less sensitive to increased temperatures compared with the Gulf Coast type. Genetic variability has been shown to play a role in plant response to climate change conditions; a recent study has found that rice (Oryza sativa L.) accessions respond differently to elevated atmospheric [CO2] and temperature, with some accessions being more impacted by increases in temperature than others (Wang et al. Reference Wang, Bunce, Tomecek, Gealy, McClung, McCouch and Ziska2016). Morphological and ecophysiological differences are known to exist between P. australis haplotypes, and it is not unlikely for these two haplotypes to respond differently to environmental stress (Eller and Brix Reference Eller and Brix2012; League et al. Reference League, Colbert, Seliskar and Gallagher2006; Mozdzer and Zieman Reference Mozdzer and Zieman2010; White et al. Reference White, Hauber and Hood2004).
Effects on Regrowth
For the Gulf Coast type, there was an effect on regrowth height by both climate treatment and glyphosate application, as well as an interactive effect of the two at 60 DAT (Table 4). Plants treated with glyphosate had less regrowth than untreated plants; however, under the elevated climate treatment, this difference (81.8 ± 6.2 cm for untreated plants, 59.8 ± 9.1 cm for treated plants) was not significant. Plant height was more affected by herbicide application under the ambient climate treatment (79.1 ± 4.4 cm for untreated plants, 26.0 ± 6.0 cm for treated plants) (Figure 3A). This indicates that plants that had been treated with glyphosate were exhibiting a greater stress response under the ambient climate conditions. Stem number was affected by climate treatment, with plants in elevated climate conditions having more stems than those in the ambient conditions, although there was no effect of herbicide application (Table 4).
Table 4 F-ratios resulting from a two-way ANOVA of plant traits at 60 DAT, with glyphosate treatment (0.57 kg ae ha−1) and climate treatment (ambient: 400 ppm CO2, 32/21 C; or elevated: 650 ppm CO2, 35/24 C) as factors.a
a Results are shown for both the Gulf Coast and Eurasian haplotypes of Phragmites australis. Statistically significant values in bold:
* P < 0.05;
** P < 0.01;
*** P < 0.001.
Figure 3 The response of the Gulf Coast type of Phragmites australis to glyphosate application at 60 d after treatment under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments. Displayed values represent the mean and SE of (A) height, (B) stem number, (C) lateral branch number, (D) aboveground biomass, and (E) belowground biomass.
There was again an effect of both herbicide application and climate treatment, and an interactive effect on lateral branch production for the Gulf Coast type (Table 4). Glyphosate-treated plants grown in elevated climate conditions produced fewer branches (2.5 ± 0.8) than did those in ambient conditions (7.1±1.6) (Figure 3C). Aboveground biomass was significantly affected by both climate and herbicide application as well (Table 4). Plants produced more aboveground biomass under elevated climate conditions, and less when treated with glyphosate (Figure 3D). For belowground biomass there was no effect of climate, only of herbicide application (with treated plants showing lower biomass than those that were untreated) (Figure 3E).
For the Eurasian type, there was no effect of climate or glyphosate application on plant height (Table 4). Stem number was significantly higher under elevated conditions, and plants produced more stems when treated with glyphosate (Figure 4B). There was no effect of climate on lateral branch production, although plants treated with glyphosate produced more branches than the untreated controls (Figure 4C). Plants produced significantly lower aboveground and belowground biomass under ambient climate conditions, but there was no effect of glyphosate application for either trait (Table 4).
Figure 4 The response of the Eurasian type of Phragmites australis to glyphosate application at 60 d after treatment under ambient (400 ppm CO2, 32/21 C) and elevated (650 ppm CO2, 35/24 C) climate treatments. Displayed values represent the mean and SE of (A) height, (B) stem number, (C) lateral branch number, (D) aboveground biomass, and (E) belowground biomass.
Plant regrowth at 60 DAT showed more effects of climate treatment than in the initial measurements. For the Gulf Coast type, stem number was higher under elevated climate conditions. Additionally, the elevated climate conditions increased belowground biomass production for both haplotypes, an effect of increased [CO2] that has been observed for P. australis in other studies (Caplan et al. Reference Caplan, Wheaton and Mozdzer2014; Mozdzer and Megonigal Reference Mozdzer and Megonigal2012). High belowground biomass facilitates shoot regeneration in rhizomatous species like P. australis, allowing for greater recovery potential following herbicide application. This resulted in plants growing in the elevated climate conditions having greater aboveground biomass than those in the ambient conditions for the Gulf Coast type.
Our data suggest that overall, increased temperature and atmospheric [CO2] had a greater long-term effect on the Gulf Coast type’s response to glyphosate, while there was only a limited initial impact on the Eurasian type. At the time of herbicide application, photosynthesis was lower for the Eurasian type under elevated climate conditions; although this may have slowed down glyphosate translocation, it was not enough to limit long-term efficacy. For the Gulf Coast type, differences in leaf area likely contributed to the differential effect of glyphosate under our two climate treatments; altered leaf traits are thought to be a primary mechanism by which climate change affects plant response to herbicides (Ziska Reference Ziska2008). Here, plants of the Gulf Coast type had a lower leaf area under elevated climate conditions, reducing the amount of herbicide taken up by the plants. There were no significant differences in leaf area between climate treatments for the Eurasian type.
The Gulf Coast type has been expanding its range in certain parts of the Gulf Coast region, including anthropogenically disturbed wetlands in Florida, where it frequently requires management. Glyphosate is currently one of the most effective tools for managing P. australis invasions; if climate change reduces its efficacy, this haplotype may become increasingly problematic in the coming years. Here, we conducted an initial study to determine whether elevated carbon dioxide concentrations and temperature have an effect on glyphosate efficacy for P. australis, and we used a sublethal rate to detect subtle differences. Further research is needed to evaluate alternative climate scenarios, lethal rates of glyphosate, and alternative herbicides and management strategies.
Acknowledgments
We thank the Florida Fish and Wildlife Conservation Commission (FWC) Invasive Plant Management Section for providing funding for this research. No conflicts of interest have been declared. We thank Carrie Reinhardt Adams and William Overholt of the University of Florida for assistance with haplotype identification and C. J. Greene and Mike Sowinski of FWC for facilitating collection of plant material. Finally, we would also like to thank Jason Ferrell of the Center for Aquatic and Invasive Plants at the University of Florida for his advice.
