Diel patterns of stem CO₂ efflux vary among cycads, arborescent monocots, and woody eudicots and gymnosperms

ABSTRACT

The diel patterns of stem carbon dioxide efflux (E_s) were determined for cycads, monocots, and woody eudicot and gymnosperm tree species. Stem E_s at a height of 30–40 cm was measured every 2 h throughout 31-h campaigns. Our range of E_s was 1.5–4.0 µmol·m⁻²·s⁻¹ for cycads, 1.0–3.5 µmol·m⁻²·s⁻¹ for arborescent monocots, and 1.5–4.5 µmol·m⁻²·s⁻¹ for woody eudicot and gymnosperm trees species. Time of day did not influence E_s of cycads or monocots. In contrast, the woody stems of eudicots and gymnosperms exhibited diurnal E_s that was 36% to 40% greater than nocturnal E_s. The established literature based on E_s of woody tree species cannot be used to estimate habitat carbon cycles in habitats which contain cycad or monocot trees. Time of day must be included for accuracy of research on E_s of woody tree species. Failures to account for the spatiotemporal differences of E_s may explain some of the disparity in outcomes of published stem respiration studies.

Introduction

Habitat-level respiration is a critical component of ecosystem carbon cycling and is the sum of soil, stem, and leaf respiration. Considerable attention has been given to stem respiration as a source of atmosphere CO₂.^1,2 Determining the contribution of stem respiration to these processes typically involves the measurement of CO₂ efflux from stem surfaces, extrapolating this to the entire stem surface area, then upscaling the flux to the stand level. The accuracy of this upscaling step is hindered by inadequate understanding of spatiotemporal variations in E_s. Temporal variations of E_s include seasonal changes as influenced by abiotic mediators such as temperature or drought.^3-6 Temporal variations of E_s also include diel changes, with the greatest E_s reported during the diurnal^4,7-9 or nocturnal^10-15 period. Biotic factors such as stem growth increment, primary productivity, or carbohydrate supply have been invoked to explain cause and effect relationships in temporal changes of E_s.^3,4,6,11,12 Much of the diel variation in E_s is also explained by the influence of sap flow on internal CO₂ movement.¹⁶ The contribution of root-respired CO₂ to stem CO₂ efflux may be substantial as a result of sap flow.^13,15,17 The differences of E_s in space and time may explain some of the disparity in outcomes of published stem respiration studies.

The E_s literature is dominated by studies on lignophyte eudicot and gymnosperm tree species that produce dense xylem from bifacial secondary vascular cambium. Cycads, palms, and other arborescent monocot species produce pachycaulous stems that do not possess secondary vascular cambium.^18-24 The E_s traits of these thick-stemmed species have been little studied to date. We are aware of only one publication that reported E_s of palm species,²⁵ and only two publications that reported E_s of a cycad species.^26,27 These reports did not consider the influence of time of day on E_s. Therefore, the established literature on diel E_s patterns has been restricted to lignophyte eudicot and gymnosperm trees.

An understanding of changes in E_s throughout the day is required for upscaling stem E_s data to stand level carbon cycle estimates. Unfortunately, the current knowledge of the diel variation of E_s is not reliable for upscaling in habitats that contain cycad, palm, or other arborescent monocot tree species. We addressed these issues by comparing the diel cycle of E_s for cycad, monocot, and lignophyte trees in three study sites. The monocot stem form included representatives from four families with dissimilar stem construction approaches.

We predicted that E_s of cycads and most arborescent monocots would exceed that of woody tree species, primarily because most of the volume of a lignophyte stem is dead heartwood that does not respire, and most of the volume of cycad and monocot families is respiring ground tissue. We also predicted that the diel pattern of E_s would be similar within each stem growth form category, but would vary among the growth forms.

Materials and methods

Our study sites and dates included Nong Nooch Tropical Botanical Garden, Chonburi, Thailand from 6 to 7 July 2019; Mangilao, Guam from 1 to 2 December 2019; and Angeles City, Philippines from 15 to 16 December 2019. For each of the dates, we conducted a 31-h campaign to determine the diel variation of stem CO₂ efflux among six trees. The first measurement period was 15:00 HR on the first date, and measurements were repeated at approximately 2-h intervals through the entire nocturnal period and the photoperiod of the second date. We then continued the measurements until 22:00 HR of the second date. These methods encompassed the final 3.5 h of the first photoperiod, an entire nocturnal period, an entire photoperiod, then the first 4 h of the following nocturnal period. The six trees in each location were comprised of two representative species for each of three stem growth forms.

Our three stem growth forms were cycads, arborescent monocots, and lignophyte trees with secondary vascular cambium. The cycad species included Cycas semota K.D. Hill and Encephalartos senticosus Vorster in Thailand, Cycas angulata R.Br. and Zamia furfuracea L.f. in Guam, and Cycas micronesica K.D. Hill and Dioon spinulosum Dyer ex Eichler in Philippines. The arborescent monocot species included Phoenix sylvestris (L.) Roxb. and Washingtonia robusta H. Wendl in Thailand, Areca catechu L. and Pandanus tectorius Parkinson ex Zucc. in Guam, and Beaucarnea recurvata Lem. and Ravenala madagascariensis Sonn. in Philippines. The lignophyte trees included Albizia saman (Jacq.) F. Muell. and Ficus benjamina L. in Thailand, Elaeocarpus joga Merr. and Tabernaemontana rotensis (Kaneh.) P.T. Li. in Guam, and Acacia auriculiformis A.Cunn. ex Benth. and Araucaria heterophylla (Salisb.) Franco in Philippines.

We selected trees that appeared vigorously growing and healthy with no obvious signs of physical trauma on the stems. We restricted our choice to individuals with a minimal increase in stem diameter at the root collar and no buttress roots. For the P. tectorius individual on Guam, we selected a tree with no stilt roots.

A CIRAS EGM-4 analyzer fitted with an SRC-1 close system chamber (PP Systems, Amesbury, MA. U.S.A.) was used to quantify the efflux of CO₂ from stem surfaces as previously described.²⁵ We used a fixed stem height of 30–40 cm above the root collar. The EGM-4 recorded air temperature and the chamber’s increase in CO₂ concentration above ambient for a 2-min period. The change in CO₂ concentration was used to calculate the flux by dividing by area and time. We conducted three periods of efflux at different radial locations for each tree for each measurement period.

The stem surface temperature was measured with an infrared thermometer (Milwaukee Model 2267–20, Milwaukee Tool, Brookfield, WI, USA) for every tree at every measurement period. Relative humidity was determined with a sling psychrometer at the beginning and end of each measurement period. Stem diameter at the 30–40 cm height of measurements and total stem height were recorded one time for each tree. Each tree required 8–9 min to complete, and each round of measurements took ~60 min in the Thailand location and 70–75 min in the other two locations.

The influences of time of day and species on E_s were assessed for each of the stem growth forms using repeated-measures analysis of variance with three replications. Time of day was designated as the repeated measure and the analysis used an autoregressive covariance structure (PROC MIXED; SAS Institute, Cary, NC, USA). Means separation for significant factors was conducted by Tukey’s HSD test.

Results

The environmental traits of our three study locations were similar (Table 1). The Thailand nocturnal maximum in relative humidity was less than that for the other two locations, but the diurnal minimum in relative humidity was similar among the locations. Stem temperature tracked air temperature during the nocturnal measurements and was 1.0–1.5°C less than air temperature for most of the measurements during the diurnal measurements. The minimum stem temperatures generally occurred prior to sunrise, and the maximum temperatures generally occurred mid-afternoon.

Table 1.

Environmental characteristics of three study locations used to characterize diel patterns of CO₂ efflux from stem surfaces of three tree stem growth forms.

Study Location	Sunrise	Sunset	Max/Min relative humidity (%)	Max/Min temperature (°C)
Thailand	05:55 HR	18:47 HR	80/62	32/27
Guam	06:27 HR	17:51 HR	88/63	30/25
Philippines	06:15 HR	17:30 HR	86/59	32/22

Cycads

We included three Cycadaceae species and three Zamiaceae species in this case study (Table 2). The plants ranged in height from 1.7 to 2.4 m and stem diameter from 19 to 34 cm. Cycad E_s was not influenced by time of day (P = .2595) or the interaction of time of day with species (P = .8884). In contrast, the species main effect was highly significant (P < .0001).

Table 2.

Characteristics of 18 tree species used to determine the diel pattern of carbon dioxide efflux from the base of stems.

Species	Family	Study location	Plant height (m)	Stem diameter (cm)
	Cycads
Cycas angulata	Cycadaceae	Guam	2.1	31
Cycas micronesica	Cycadaceae	Philippines	1.8	19
Cycas semota	Cycadaceae	Thailand	1.7	21
Dioon spinulosum	Zamiaceae	Philippines	2.3	25
Encephalartos senticosus	Zamiaceae	Thailand	2.4	34
Zamia furfuracea	Zamiaceae	Guam	1.9	23
	Monocots
Areca catechu	Arecaceae	Guam	7.6	18
Beaucarnea recurvata	Asparagaceae	Philippines	4.4	24
Pandanus tectorius	Pandanaceae	Guam	8.6	19
Phoenix sylvestris	Arecaceae	Thailand	9.1	41
Ravenala madagascariensis	Strelitziaceae	Philippines	8.8	26
	Lignophytes
Acacia auriculiformis	Fabaceae	Philippines	13.2	42
Albizia saman	Fabaceae	Thailand	20.5	92
Araucaria heterophylla	Araucariaceae	Philippines	14.6	31
Elaeocarpus joga	Elaeocarpaceae	Guam	6.2	36
Ficus benjamina	Moraceae	Thailand	16.6	48
Tabernaemontana rotensis	Apocynaceae	Guam	7.8	29

The E_s of our six cycad species were separated into three overlapping groups within a range of 1.5–4.0 µmol·m⁻²·s⁻¹ (Figure 1). The Cycadaceae species exhibited a range in E_s that was similar to that of the Zamiaceae species.

Figure 1.

Carbon dioxide efflux from the base of stems of six cycad species as influenced by time of day. Open triangle = Cycas micronesica, closed triangle = Dioon spinulosum, open square = Zamia furfuracea, closed square = Cycas angulata, open circle = Encephalartos senticosus, closed circle = Cycas semota. Dashed lines and triangles from Angeles City, Philippines. Dotted lines and squares from Mangilao, Guam. Solid lines and circles from Chonburi, Thailand. Bar on the x-axis depicts nocturnal hours. Species with the same letter on the right vertical axis are not different according to Tukey’s HSD test.

Arborescent monocots

The monocot species that we selected were representatives of four families including the palm family Arecaceae (Table 2). The plants ranged in height from 4.4 to 31.2 m and stem diameter ranged from 18 to 49 cm at the height of measurements. Monocot tree E_s was not influenced by time of day (P = .9415) or the interaction of time of day with species (P = .9747). In contrast, the species main effect was highly significant (P < .0001).

The E_s of our six monocot species were separated into two distinct groups (Figure 2). The pandanus and palm species exhibited E_s (2.8–3.5 µmol·m⁻²·s⁻¹) that was significantly greater than that of the Asparagaceae and Strelitziaceae species (1.1–1.3 µmol·m⁻²·s⁻¹).

Figure 2.

Carbon dioxide efflux from the base of stems of six arborescent monocot species as influenced by time of day. Open triangle = Ravenala madagascariensis, closed triangle = Beaucarnea recurvata, open square = Pandanus tectorius, closed square = Areca catechu, open circle = Washingtonia robusta, closed circle = Phoenix sylvestris. Dashed lines and triangles from Angeles City, Philippines. Dotted lines and squares from Mangilao, Guam. Solid lines and circles from Chonburi, Thailand. Bar on x-axis depicts nocturnal hours. Species with the same letter on the right vertical axis are not different according to Tukey’s HSD test.

Lignophyte eudicot and gymnosperm trees

We included five families of lignophyte tree species, including the legume family Fabaceae (Table 2). The plants ranged in height from 6.2 to 20.5 m and stem diameter from 29 to 92 cm. The time of day exerted a significant influence on E_s of these woody tree species (P = .0049). Similarly, the species main effect was highly significant (P < .0001). In contrast, the interaction of time of day with species was not significant (P = .9221), indicating the significant changes in E_s throughout the day were similar among the species.

The six lignophyte species that we included exhibited a range in E_s of about 1.5–4.5 µmol·m⁻²·s⁻¹ (Figure 3). The E_s of these lignophyte trees formed three distinct groups. The two Fabaceae species exhibited the greatest E_s, and the two native tree species from Guam exhibited the least E_s. The middle to end of the photoperiod exhibited maximum E_s for all six species and most of the nocturnal period exhibited a relatively stable minimum E_s.

Figure 3.

Carbon dioxide efflux from the base of stems of six lignophyte tree species. Open triangle = Araucaria heterophylla, closed triangle = Acacia auriculiformis, open square = Elaeocarpus joga, closed square = Tabernaemontana rotensis, open circle = Ficus benjamina, closed circle = Albizia saman. Dashed lines and triangles from Angeles City, Philippines. Dotted lines and squares from Mangilao, Guam. Solid lines and circles from Chonburi, Thailand. Bar on x-axis depicts nocturnal hours. Species with the same letter on the right vertical axis are not different according to Tukey’s HSD test.

Discussion

The range in E_s was similar among the cycad and monocot tree species, and the E_s of these trees was on average below that of the lignophyte growth form. This outcome was not consistent with our first prediction. The results indicated that the absolute total volume of respiring tissue within a tree stem does not predict E_s. One explanation may be that the metabolic activity of the peripheral tissues of a tree stem controls E_s regardless of the respiratory activity of the tissues in the center of the stems. Therefore, the highly active secondary vascular cambium and growing cells of the young phloem and xylem of the lignophyte stems respire CO₂ that easily conducts to the stem surface despite the fact that the center of the lignophyte stem is not metabolically active. In contrast, the less active peripheral ground tissue of cycad and monocot tree stems respires relatively less CO₂ that can easily conduct toward the stem surface even though most of the center of the stem contains living tissue. A second explanation may be that the respiratory activity of the copious volume of respiring tissue of cycads and arborescent monocot stems may be less than that of the limited volume of respiring tissue within lignophyte stems because respiratory potential may be inversely correlated with the volume of parenchyma tissue.^28-30 A third explanation may involve the radial position of the most active xylem, since much of the variation in E_s can be explained by the influence of sap flow on the internal CO₂ movement.¹⁶ The contribution of root-respired CO₂ to stem CO₂ efflux may be substantial as a result of sap flow.^13,15,17 The active xylem for lignophyte stems is located close to the stem surface. In contrast, the active xylem for the cycad and monocot stems is located inside a wide peripheral layer of ground tissue. Therefore, the CO₂ that is translocated from roots through xylem may readily conduct radially toward the stem surface for lignophyte trees, but the root-derived CO₂ may not easily conduct toward the stem surface of cycad or monocot stems because of the copious peripheral ground tissue. Indeed, relative radial CO₂ conductance may differ among tree species and has been proposed as one factor that explains differences in E_s.³¹ More studies that include a greater number of species from each growth form are needed to confirm if our results are canonical or idiosyncratic for the six species we selected for each growth form.

The diel patterns of E_s for our lignophyte eudicot and gymnosperm tree species revealed an increase in E_s during the diurnal period in conformity with previous reports.^4,7-9 The highest E_s in our study occurred during the late afternoon at a time when the transfer of dissolved CO₂ from root respiration would enter stems via sap flow. Moreover, the species with the greatest mean E_s were legume trees, and this same result was reported for a range of tree growth forms in Costa Rica.²⁵

Our results were consistent with our second prediction, in that all six species within each growth form exhibited similar patterns with respect to time of day, but the three growth forms did not exhibit similar diel patterns. In contrast to the lignophyte species, the cycad and arborescent monocot tree species did not reveal any observable pattern with respect to the diel cycle of E_s. Most published studies were conducted in colder climates, and some of the diel patterns of E_s may have been controlled by the influence of temperature on respiration.^4,7 More research in colder climates is needed with cycad and monocot tree species to determine if a greater diel variation in temperature would cause a change in diel patterns of E_s of these pachycaulous stems.

Radial stem growth of some arborescent monocot species does involve a secondary cambium.^22,23 However, this monocot cambium is not a vascular cambium and most of the new tissue construction is ground tissue with bundles of vascular tissue interspersed. In all cycads and arborescent monocots, the presence of a relatively thick zone of ground tissue comprised of parenchyma or sclerenchyma is peripheral to the vascular tissue. For example, we have reported a cortex width of 5–6 cm for some Cycas species.³² Our results suggest that this peripheral ground tissue may buffer the radial conductance of xylem CO₂ such that the temporal variations of sap flow do not overtly influence the temporal patterns of E_s. Alternatively, this ground tissue may temporarily store or consume dissolved CO₂ such that the ultimate diffusion to the open air occurs as a relatively stable flux throughout the diel cycle. Biological consumption of stem CO₂ may occur through stem photosynthesis or other processes.³³ Substantial CO₂ storage has been reported in parenchyma cells, even for lignophyte stems.³⁴ The temporal and spatial dynamics of carbohydrate movement and the location of storage pools are not well understood for pachycaulous growth forms, and this continues to limit our understanding of the relationship between carbohydrate relations and E_s.

Several of our tree species are threatened with extinction. The cycad species Cycas micronesica, Dioon spinulosum, and Zamia furfuracea are Red-listed as Endangered.^35-37 The Guam trees Cycas micronesica and Tabernaemontana rotensis are listed as Threatened on the United States Endangered Species Act.³⁸ Moreover, cycad species comprise the most threatened plant group internationally.³⁹ Conservation biologists include the stakeholder group with the role of determining the causal mechanisms behind the abiotic and biotic threats that generate declines in populations of threatened species. Conservation physiology is a critical sub-discipline of conservation biology because an understanding of the physiological responses of organisms to their ever-increasing threats is required for success.^40–42 We have studied various aspects of cycad stem behavior,^{26,27,32,43-48} and a continued focus on stem physiology and all other aspects of plant behavior would greatly improve conservation programs of cycads and other threatened tree species. The continued study of stem respiration is an ideal conservation physiology endeavor because the measurements are nondestructive and can be applied without damage to individuals of the most threatened plant taxa.

In summary, the considerable literature on E_s remains biased toward lignophyte eudicot and gymnosperm tree species that produce dense xylem from bifacial secondary vascular cambium. We have shown that the use of this literature for predicting and modeling ecosystem-level CO₂ relations will lead to equivocal results for habitats with cycad or arborescent monocot species. Based on our species, the range and greatest E_s among cycad and monocot trees appear to be less than the range and greatest E_s among lignophyte trees. Inaccuracies in upscaling the contributions of E_s to the habitat or ecosystem-level carbon cycle are highly likely if only one time of day is used for E_s studies on lignophyte tree species. However, our results indicate that similar inaccuracies are less likely when only one time of day is used for E_s studies on cycad or arborescent monocot species.

Funding Statement

This research was funded by the United States Forest Service Cooperative Agreement number 17-DG-11052021-217. Support and access to germplasm provide by Mr. Kampon Tansacha, Director of Nong Nooch Tropical Botanical Garden.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.