Enriched CO2 and Root-Associated Fungi (Mycorrhizae) Yield Inverse Effects on Plant Mass and Root Morphology in Six Asclepias Species

While milkweeds (Asclepias spp.) are important for sustaining biodiversity in marginal ecosystems, CO2 flux may afflict Asclepias species and cause detriment to native communities. Negative CO2-induced effects may be mitigated through mycorrhizal associations. In this study, we sought to determine how mycorrhizae interacts with CO2 to influence Asclepias biomass and root morphology. A broad range of Asclepias species (n = 6) were chosen for this study, including four tap-root species (A. sullivantii, A. syriaca, A. tuberosa, and A. viridis) and two fibrous root species (A. incarnata and A. verticillata). Collectively, the six Asclepias species were manipulated under a 2 × 2 full-factorial design that featured two mycorrhizal levels (−/+ mycorrhizae) and two CO2 levels (ambient and enriched (i.e., 3.5× ambient)). After a duration of 10 months, Asclepias responses were assessed as whole dry weight (i.e., biomass) and relative transportive root. Relative transportive root is the percent difference in the diameter of highest order root (transportive root) versus that of first-order absorptive roots. Results revealed an asymmetrical response, as mycorrhizae increased Asclepias biomass by ~12-fold, while enriched CO2 decreased biomass by about 25%. CO2 did not impact relative transportive roots, but mycorrhizae increased root organ’s response by more than 20%. Interactions with CO2 and mycorrhizae were observed for both biomass and root morphology (i.e., relative transportive root). A gene associated with CO2 fixation (rbcL) revealed that the two fibrous root species formed a phylogenetic clade that was distant from the four tap-root species. The effect of mycorrhizae was most profound in tap-root systems, as mycorrhizae modified the highest order root into tuber-like structures. A strong positive correlation was observed with biomass and relative transportive root. This study elucidates the interplay with roots, mycorrhizae, and CO2, while providing a potential pathway for mycorrhizae to ameliorate CO2 induced effects.


Study System
The aim of this study was to determine ways in which CO 2 and mycorrhizae can influence root morphology and plant performance (i.e., whole biomass). Thus, 6 species of milkweed from genus Asclepias were chosen for this experiment (Table 1). Milkweeds are excellent specimens for environmental manipulation (i.e., growth chamber) because they are adapted to a wide range of ecological niches [37], including disturbed, pristine, drought, and wetland habitats (Table 1). In addition, milkweed (hereafter referred to as Asclepias species) are important to marginal ecosystems (i.e., remnant prairie), biodiversity (i.e., monarch butterflies), and conservation interests.

Source of Asclepias spp. (Milkweed) and Mycorrhizae
Asclepias seeds were sourced from the midwestern region of the U.S.A. The 6 Asclepias spp. were A. sullivantii, A. verticillata (seed source, Prairie Moon Nursery Winona, MN, U.S.A), A. incarnata, A. syriaca, A. tuberosa, and A. viridis (seed source, Missouri Wild Flower Nursery (Jefferson City, MO, U.S.A). The relatedness of each of these species was characterized using a chloroplast DNA sequence that codes for ribulose bisphosphate carboxylase (rbcL), a conserved gene involved in CO 2 fixation. Since Asclepias spp. response to (CO 2 ) manipulation is the nature of this study, assessing Asclepias relatedness with respect to rbcL gene was appropriate. Prefers moist wet soils (wildflower.org), and is found in standing water several months of the year [38]. In addition, many stems arise from a single root stock [38].
Fine roots became increasingly coarse when colonized with mycorrhizae.
Very dense fine roots that tightly aggregated the soil together when inoculated with mycorrhizae.
Prefers well-drained loamy soils (wildflower.org). Grows large extensive rhizome systems while adjacent stems may belong to the same or different clones [38].
Primary/dominant root is modified into tuber-like structures when colonized with mycorrhizae.
Primary/dominant root is modified into tuber-like structures when colonized with mycorrhizae.

A. tuberosa
Butterfly milkweed Thick woody orange-brown tap-root that serves for C storage, attachment, and perennation. This species produces smaller lateral roots [40]. Many stems arise from a single root stock [38] Primary/dominant root is modified into tuber-like structures when colonized with mycorrhizae.

A. viridis Spider milkweed
Prefers moist soils (wildflower.org). Root cardenolides in this species are four times higher than those of common milkweed [41].
Primary/dominant root is modified into tuber-like structures, but, in some cases, the tuber-like root was less elongated and more gall-like when colonized with mycorrhizae.
Similar to Asclepias spp., mycorrhizal isolates were also sourced from the midwestern region of the U.S.A. Fungal consortia featured 4 isolates of mycorrhizae derived from Kankakee sands prairie reserve in Indiana, USA. Isolates included several species (Claroideoglomus claroideum, Racocetra fulgida, Funneliformis mosseae, and Claroideoglomus lamellosum) that have comparable levels of colonization [42]. In addition, these species are representative of mycorrhizal phylogenetic diversity [43,44], and include isolates of high and low experimental usage rate [45].

Preparing Mycorrhizal Inoculate
Mycorrhizae are lab isolates that were first isolated from prairie soil to start lab cultures (Table S1). Identity of spores were previously determined using morphological and phylogenetic species concept [43]. Since then, these isolates have been of standard use in the lab [42,46,47]. Prior to the experiment, fungal cultures were prepared and bulked on sorghum roots for a full growing season under glasshouse conditions [46]. Briefly, after a full growing season, aboveground sorghum tissue was removed, while the belowground soil and mycorrhizal root mix was stored at 4 • C prior to experimental use.
At the time of experimental use, mycorrhizal fungal inocula (fungal spores and sorghum root mix) was used to inoculate in between the top and bottom layers of heat sterile background soil (Kansas clay loam-sand (1:1)). The background soil was mixed with sand, primarily because sand enhances drainage in pots, reduces nutrient levels, and enhances plant's mycorrhizal functional response. The volume of the added inocula (i.e., fungal spore-root mix) was 50 cm 3 (+ mycorrhizae). Meanwhile, sterile inocula void of live mycorrhizae was added to control pots (− mycorrhizae), also at a volume of 50 cm 3 . This assured that the soil structure and nutrient ratio of all pots was maintained. Finally, Asclepias seedlings were transplanted to cone-tainer pots (Stuewe & Sons Inc., Tangent, Oregon, U.S.A.), which were positive or negative for mycorrhizae.

Growth Chamber and Atmospheric Conditions
Prior to transplanting milkweed to experimental conditions (i.e., −/+ mycorrhizae and CO 2 manipulation), Asclepias spp. were germinated on heat sterile potting soil (Berger bark mix BM 7, www.berger.ca. access date: 9 September 2019) After five weeks, 5-weekold seedlings were transferred to pots that were positive or negative for mycorrhizae. The dimensions of the pots were 6.8 cm in diameter by a depth of 35.56 cm. In a random block design, pots were then placed into 4 Conviron CMP 6050 growth chambers (Controlled Environments Limited, Winnipeg, Manitoba, Canada). Two of the chambers were set to ambient (CO 2 ) at 400 ppm, while the other two chambers were set to 3.75× ambient levels. The 3.75× ambient level (hereafter referred to as enriched CO 2 ) was chosen because geological time records suggest (CO 2 ) > 1500 ppm reduces plant stomatal index [48], which is of detriment to photosynthesis and Earth's biosphere. Growth chamber settings were 23 • C, 70% humidity (% RH), and a lighting intensity of 1000 unit micromoles. CO 2 concentrations were set to 400 ppm (ambient) and 1500 ppm (enriched). Plants were fertilized with nitrogen (0.2 g/L) periodically (average of once a month) at an application volume of 50 mL. Similar to Malik et al. (2016) [46], this (N) was determined and applied by using the atomic mass of N in the NH 4 NO 3 compound, which was weighted to control for N mass. Using this ratio, we determined how much NH 4 NO 3 is needed in 1 L of water to create 0.2 g/L of N.

Harvest
After a duration of 10 months, the effects of mycorrhizae and CO 2 on Asclepias spp. were observed. Roots were washed and scanned for root morphological responses. Briefly, roots were blot dried and positioned next to a scale, and an EPSON scan was used to make morphological observations. Roots were then stored in 75% ethanol and boiled in KOH, prior to trypan blue staining [49]. Staining and qualitative microscopic examination confirmed the mycorrhizal treatments. Finally, plant performance was assessed as whole dry weight.

Root Image Analysis
For image analysis, roots were washed and scanned to assess morphological response to CO 2 and mycorrhizae. Root morphological response was quantified with ImageJ 1.46r [50], as pixels were calibrated to centimeters to assess diameter. Root diameter was scored at the midpoint of the highest order root (transport root), as well as the midpoint of the lowest order root (absorptive root). Percent difference in transport versus absorptive root diameter (relative transport root) was estimated.
Relative transport root, or percent difference between transport and absorptive root, was appropriate because root diameter has been shown to vary across plant groups [1], including tap-and fibrous root species. Relative transport root is calculated according to Equation 1, where transport (T) and absorptive (A) root diameters are divided by transport root (T) diameter. This value is then multiplied by 100.
Relative transport root = T − A T * 100

Statistical Analysis for Root Morphology and Biomass
This study was analyzed in R version 4.0.2. Shapiro-Wilk and Levene's Test, as well as diagnostic plots, were used to assess homogeneity in variance and normality for dependent variables (i.e., relative transport root and biomass). CO 2 level, mycorrhizae, species and growth chamber effects were set as predictors, while both relative transportive root and biomass were set as response variables. MANOVA analysis was employed, as relative transport root and biomass were response variables from the same set of plants. Briefly, for MANOVA analysis, relative transport root and biomass were made into a vector, and simultaneous inferences were made by the predictor variables. The analysis was followed by a set of ANOVAs that independently predict biomass and root morphology. Interactions among the predictors were also included in the model, this was critical for determining whether or not growth chamber(s) were a confounding factor. Post-hoc analysis included Step AICc, Tukey HSD, and a priori contrast using the 'multcomp' package. The 'multcomp' package allowed simultaneous test for general linear hypotheses [51]. Finally, a correlation between relative transport root and biomass was assessed using the 'rcorr' and 'cor.test' function via 'Hmisc' package [52]. Correlations were then examined with 95% CI ellipses, using 'ggplot 2'. The ellipses enabled graphical analysis for outcome and experimental variables 2.9. Phylogenetic Analysis: Asclepias Relatedness Using CO 2 Fixation Gene Amino acid (AA) sequence for rbcL locus was aligned using Muscle, which assesses distance using kmer distance for unaligned pairs and kimura distance for aligned pairs (Edgar, 2004). Evolutionary analysis was inferred using Maximum Likelihood Method and the JTT matrix-based model [53]. The phylogenetic tree for the 6 Asclepias species was constructed according to the highest log likelihood. This enabled a tree that was drawn to scale with branch lengths measured as the number of substitutions per site. While this analysis involved 6 AA sequences, there were a total of 475 positions in the final dataset. This allowed evolutionary analysis to be conducted with MEGA X [54].

Correlation with Relative Transport and Biomass
A significant correlation was also detected with biomass and root diameter, as these two responses were positively associated (Pearson correlation, r = 0.59, n = 191, p < 0.001, Figure 3). In addition, 95% CI ellipse revealed that the association with relative transport root and biomass was increasingly positive in the presence of mycorrhizae, which led to a small overlap between + mycorrhizae and − mycorrhizae (Figure 4a). The CO 2 regime (i.e., ambient versus enriched) produced a 95% CI ellipse with a high degree of overlap and similar shape. However, the CO 2 ellipses were not congruent as the ambient CO 2 ellipse was more positive (Figure 4b). As it relates to species, both fibrous root species, A. incarnata and A. verticillata produced ellipses shapes that were unique and not angled at 45 degrees, unlike the tap-root species (Figure 4c). These differences are even more apparent when ellipses are assorted by root system (Figure 4d).

Asclepias Species Relatedness and Root System
With respect to the rbcL locus (i.e., gene involved in CO 2 fixation), a tree was constructed to the highest log likelihood (−1398.00) to examine relatedness of the six Asclepias species. Interestingly, the root system helped explain relatedness. Specifically, the two fibrous root species, A. incarnata and A. verticillata were sister taxa, and an outgroup, that was distant to the four other tap-root species ( Figure 5). versus enriched) produced a 95% CI ellipse with a high degree of overlap and similar shape. However, the CO2 ellipses were not congruent as the ambient CO2 ellipse was more positive ( Figure 4b). As it relates to species, both fibrous root species, A. incarnata and A. verticillata produced ellipses shapes that were unique and not angled at 45 degrees, unlike the taproot species (Figure 4c). These differences are even more apparent when ellipses are assorted by root system (Figure 4d).

Asclepias Species Relatedness and Root System
With respect to the rbcL locus (i.e., gene involved in CO2 fixation), a tree was constructed to the highest log likelihood (−1398.00) to examine relatedness of the six Asclepias species. Interestingly, the root system helped explain relatedness. Specifically, the two fibrous root species, A. incarnata and A. verticillata were sister taxa, and an outgroup, that was distant to the four other tap-root species ( Figure 5).

Asclepias Species Relatedness and Root System
With respect to the rbcL locus (i.e., gene involved in CO2 fixation), a tree was constructed to the highest log likelihood (−1398.00) to examine relatedness of the six Asclepias species. Interestingly, the root system helped explain relatedness. Specifically, the two fibrous root species, A. incarnata and A. verticillata were sister taxa, and an outgroup, that was distant to the four other tap-root species ( Figure 5). Figure 5. Provided here is a maximum likelihood tree for 6 the Asclepias species. This tree was constructed using highest log likelihood (−1398.00) for the rbcL amino acid sequence (rbcL is a gene involved in CO2 fixation). According to this tree, fibrous root species, A. verticillata and A. incarnata, are distant from the 4 tap-root species.

Chamber Effects
Predictors (i.e., CO2, mycorrhizae, Asclepias species) did not interact with CO2 chambers (Table S2 and S3, p > 0.05). This provides confidence that results were not confounded by the experimental apparatus. In addition, Step AICc revealed that response variables Figure 5. Provided here is a maximum likelihood tree for 6 the Asclepias species. This tree was constructed using highest log likelihood (−1398.00) for the rbcL amino acid sequence (rbcL is a gene involved in CO 2 fixation). According to this tree, fibrous root species, A. verticillata and A. incarnata, are distant from the 4 tap-root species.

Chamber Effects
Predictors (i.e., CO 2 , mycorrhizae, Asclepias species) did not interact with CO 2 chambers (Table S2 and S3, p > 0.05). This provides confidence that results were not confounded by the experimental apparatus. In addition, Step AICc revealed that response variables including biomass and relative transport root were best explained when growth chamber interactions were removed from the model.

Discussion
Enriched CO 2 and mycorrhizae may be additive or opposing forces helping shape plant eco-physiological responses. CO 2 flux is relevant to root life span, production, and diameter [55][56][57], as well as ecosystem processes [58][59][60], and plant fitness [61]. Interestingly, the targeted role of roots may be enhanced by mycorrhizae [62], but our findings reveal that enriched CO 2 reduced plant biomass by 25% (Figure 2a). We speculate that the mechanism at play may be CO 2 regulation of stomatal density, conductance, and aperture [63], as enriched CO 2 can lead to stomatal closures, which can reduce carbon capture and net CO 2 assimilation [64,65]. Together, this may explain negative CO 2 -induced effects on Asclepias spp., particularly biomass. Contrarily, mycorrhizae improved biomass by 12-fold while modifying root morphology (Figures 1 and 2b,d). Significant interactions with CO 2 and mycorrhizae were detected for biomass and relative transport roots (Tables S2 and S3). This study elucidates potential mechanisms in which CO 2 and mycorrhizae may yield asymmetrical outcomes on plant eco-physiology.

Biomass, CO 2 and Mycorrhizae
Carbon is captured from the atmosphere, and used to build biomass and organs (i.e., roots), but this would not be sustainable without root absorption, subsurface foraging, nutrient transport, and organic content storage [66][67][68]. While root functionality can be stimulated by CO 2 [69], enriched CO 2 can have negative outcomes on fine lateral roots and plant mass in agroecosystems [70,71]. As it relates to Asclepias species, which are essential to specialist herbivores (e.g., monarch butterflies), biodiversity, and prairie ecosystems; enriched CO 2 decreased plant biomass by 25% (Figure 2a). These differences were most apparent in five of the six Asclepias species (A. incarnata, A tuberosa, A. sullivantii, A. viridis, A. syriaca, Figure 2e-j (but not Figure 2i)). Fibrous root species, A. verticillata, was the only species in which the combination of mycorrhizae and enriched CO 2 was observed to have an additive effect on biomass (Figure 2i). This may be explained by the fact that A. verticillata is considered to have some of the most divergent traits in genus Asclepias [39]. Hence, the disparate thin pointy leaves and very fine shallow grass-like roots makes A. verticillata distinct, and may have a role in explaining A. verticillata's unique response to mycorrhizae and CO 2 .
Interestingly, the CO 2 effect on biomass and root morphology may be context dependent. At a CO 2 level of 650 ppm, plant growth and percent mycorrhizal colonization was reportedly increased [72]. Perhaps suggesting that when CO 2 stimulates plant growth, percent mycorrhizal colonization is also stimulated. While percent mycorrhizal colonization was not quantified in this study, it may be the case that since enriched CO 2 depressed plant growth (Figure 2a), percent mycorrhizal colonization was also depressed. This may help explain the significant interaction with CO 2 and mycorrhizae on biomass (Table S2). However, the outcome of CO 2 on plant-mycorrhizal nutrient exchange may be species or cultivar specific [24].

CO 2 -Mycorrhizae Interaction and Root Response
Although elevated CO 2 can lead to roots of high tissue density [73], root morphology has also been found to be associated with aboveground traits [1]. However, the precise mechanism as to how aboveground environmental cues affect belowground physiology remains obscure. In the present study, CO 2 did not influence root morphology (Figure 2c), as these findings differed from a recent study [69]. It may be the case that CO 2 can indirectly affect root morphology by increasing mycorrhizal root length colonization [74]. Interestingly, this hypothesis would rely on root systems that are highly responsive to mycorrhizae, as well as atmospheric-edaphic environment interactions.
Irrespective of the main effect of CO 2 , we observed an interaction with CO 2 and mycorrhizae on root response (p = 0.03, Table S3). This may provide insight as to why the effect of plus mycorrhizal treatment depended on the CO 2 level in A. verticillata roots. Specifically, mycorrhizae and enriched CO 2 depressed A. verticillata's relative transport root (Figure 2o). This sort of interaction was not observed in A. incarnata, the other fibrous root species. Differences among these two fibrous root species (i.e., A. verticillata and A. incarnata) may be due to divergent life histories and localized adaptation. In particular, A. verticillata thrives in xeric soils, while A. incarnata thrives in hydric soils (Table 1). Despite this, A. incarnata and A. verticillata (two fibrous root species), were more closely related than the other four tap-root species.
Tap-versus fibrous root systems may have been a good predictor of Asclepias spp. relatedness, as this observation does not appear to be confounded by rhizome trait(s). Hence, when rhizome traits were mapped onto a pre-existing phylogeny [75,76], it was determined that Asclepias spp. relatedness did not assort according to rhizome morphology. However, as it relates to the present study, Asclepias spp. assort according to the root system ( Figure 5), as relatedness was characterized using the rbcL locus (Table S4)

Mycorrhizae, Root Modification, and Carbon Storage
Root morphology can vary across plant groups (i.e., lilies, forbes, grasses) of the same community [1]. This may be explained by root metabolomics [77], which have been shown to be altered by mycorrhizae [78]. Despite this, mycorrhizal implications may depend on the root system and morphology [21]. The outcome of mycorrhizae on root morphology was most exaggerated in tap-root systems ( Figure S1), irrespective of [CO 2 ]. In most cases, tap-roots were modified into tuber-like structures (Figure 1), perhaps re-iterating the notion that tap-root systems are more likely to be mycorrhizal responsive. In contrast, fibrous root systems may antagonize mycorrhizae, as fibrous root systems have been shown to have negative effects on mycorrhizal hyphal length [22]. In addition, mycorrhizal dependence has been reported to decrease with the increase in root fibrousness [21].
As it relates to the ecosystem function, mycorrhizae may use different mechanisms to promote carbon storage depending on the plant root system. Our findings suggest that in tap-root systems, mycorrhizae may directly promote root carbon storage by helping to modify tap-roots into tuber-like structures ( Figure 1). However, mycorrhizae may complement fine roots by effectively promoting microaggregate stabilization [22], which can affect C sequestration and nutrient cycling [79][80][81].

Conclusions
Mycorrhizae enhanced relative transportive root and biomass, while enriched CO 2 had the opposite effect. Interactions with CO 2 and mycorrhizae were observed for both biomass and relative transport root. The findings of this study shed insight on how CO 2 and mycorrhizae may interact and influence plant eco-physiology. In particular excess CO 2 (i.e., 1500 ppm) can depress plant productivity, while mycorrhizae can act as a countering force to improve productivity.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/plants10112474/s1, Table S1. Code of Isolates. Table S2. Biomass explained by CO 2 , mycorrhizae, species, and growth chamber. Significant codes are for alpha thresholds. Table S3. Root morph. explained by CO 2 , mycorrhizae, species and growth chamber. Significant codes are for alpha thresholds. Figure S1 The effect of CO 2 and mycorrhizae on fibrous root systems. Irrespective of CO 2 regime, +mycorrhizae yielded a larger effect on tap-root systems. Table S4. GenBank Accession table.