Gas exchange characteristics of Typha latifolia L. from nine sites across North America

botany ELSEVIER

Aquatic Botany49 (1995) 203-215

Gas exchange characteristics of Typha latifolia L. from nine sites across North America Alan K. Knapp a,. Joseph B. Yavitt b a Division of Biology, Ackert Hall, Kansas State University, Manhattan, KS 66506, USA h Department of Natural Resources, Fernow Hall, Cornell University, Ithaca, NY 14853, USA Accepted 6 October 1994

Abstract Gas exchange characteristics were measured in the field for nine populations of Typha latifolia L. from Florida to Minnesota in North America. These populations spanned a substantial gradient in growing season length and environmental conditions. The purpose of this study was to assess geographic variability in stomatal conductance (gst) in T. latifolia populations, as well as to identify key environmental and plant factors that may affect gst and, potentially, trace gas emissions through stomata. Midday rates of net photosynthesis (A) and gst were measured under full sunlight conditions at each site, and more intensive measurements of diurnal and seasonal variability were made at selected sites. In general, A varied less than gs, ( 1.5 vs. 3 fold) among sites with maximum A (27.4 /zmol m-2 s - i ) and gst (1076.7 mmol m-2 s-~) measured at the southern sites. We found that A and gst in T. latifolia increased significantly with increasing temperature and light level, and varied diurnally and seasonally. Moreover, stomata closed completely at night, in contrast to some other wetland plants. Both A and gst also increased significantlyalong the length of leaves from the base of plants to the upper canopy. Finally, although gst was quite variable, A did not appear to be limited by gst under typical field conditions. It was concluded that generalizations that stomata in wetland plants are relatively unresponsive to environmentalfactors are not consistent with field responses measured in T. latifolia. As a result, because gst may influence trace gas flux in this species, spatial and temporal variations in gst need to be considered when emissions are estimated.

1. Introduction Typha latifolia L. or broadleaf cattail is widely distributed across North America in a variety of wetland habitats (Hotchkiss, 1972). This emergent aquatic macrophyte plays an important role in wetland succession (Van der Valk, 1981; Weisner, 1993), waterfowl * Correspondingauthor. 0304-3770/95/$09.50 © 1995 ElsevierScienceB.V. All rights reserved SSD10304-3770(94)00433-1

204

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

habitat quality (Murkin et al., 1982; Mallik and Wein, 1986), and the emission of sedimentderived trace gasses to the atmosphere (Sebacher et al., 1985; Knapp and Yavitt, 1992; Chanton et al., 1993). Although limited genetic variation among southeast US populations of this species was detected by Mashburn et al. (1978), McNaughton (1966) conducted extensive common-garden studies establishing that significant ecotypic variation was present in T. latifolia populations from Texas to Canada. He concluded that there were marked differences among geographically distinct populations in phenological characteristics and responses to temperature. In contrast, photosynthetic rates varied much less than expected among latitudinally disparate populations when T. latifolia was compared with other widespread plant species. Moreover, McNaughton (1973) reported that although photosynthetic rates decreased significantly with leaf age, stomatal conductance did not vary substantially. More recent studies of wetland plants have shown that maximum photosynthetic rates (A) can be quite high (Drake, 1984) relative to non-wetland species and, as in terrestrial plants, correlated with stomatal conductance (gst). However, responses in gst to environmental factors such as light and humidity are more variable and may be distinct from terrestrial plants (Jones, 1987; Jones, 1988; Lafluer, 1988). In particular, some wetland species have been noted for a lack of stomatal responsiveness to light, such that gst remains relatively constant throughout the day (Koch and Rawlik, 1993) and in some species, stomata are reported to remain open continuously, even in darkness or when A is zero (Jones, 1988; Schutz et al., 1991; Torn and Chapin, 1993). Reductions in diffusional limitations to A in aquatic plants might be expected as productivity and not water conservation should provide the greater selective advantage relative to plants in more waterlimited terrestrial systems. As part of a study to document geographic variation in trace gas emissions from T. latifolia populations (Yavitt and Knapp, 1995), gas exchange characteristics were measured in nine T. latifolia stands in North American wetlands ranging in location from Minnesota and Wisconsin in the north to New Mexico and Florida in the south. Because sedimentderived trace gasses such as CH4 may diffuse into these plants and then escape to the atmosphere through stomatal openings (in T. latifolia, as well as in other wetland species (Sebacher et al., 1985; Schutz et al., 1991; but see also Nouchi et al., 1990) ), it is particularly important to ascertain, (1) the inter-population variation in stomatal conductance in this widespread species, and (2) the environmental factors that may influence gst. This field survey of variability in gas exchange characteristics of T. latifolia coupled with additional, more intensive measurements made on selected populations allowed us to address the following specific questions: 1. Is there significant variation between widespread populations in A and gst in T. latifolia when measured at mid-season under field conditions? 2. Does gst vary diurnally and seasonally in T. latifolia? In particular, does gst vary significantly with light and do stomata close at night? Also, across broad environmental gradients, does gst vary predictably with temperature, humidity or other climatic factors? 3. Do A and gst vary significantly along the length of T. latifolia leaves? Or are measurements made at the top of the canopy representative of gas exchange along the entire (up to 1.5 m) length of leaves? 4. Finally, is A strongly correlated with gst in T. latifolia (as in terrestrial plants), and does gst potentially limit A under typical field conditions?

A.K. Knapp, J.B. Yavitt / Aquatic Botany 49 (1995) 203-215

205

2. Methods

2.1. Study sites Study sites spanned a latitudinal gradient from 25 ° 50'N in Florida to 47 ° 10'N in Minnesota (Table 1). This gradient corresponded to reductions in growing season length and decreasing temperatures from lower to higher latitudes. Wetland types varied from relatively small seeps and sloughs to river bank and extensive marsh systems. All sites were dominated by T. latifolia although in the Florida site, Typha domingensis Pers. may interade with T. latifolia (Smith, 1967). At this site, care was taken to select populations with morphological and reproductive characteristics most similar to T. latifolia (i.e. leaves greater than 1.5 cm in width; Smith (1967); Hotchkiss (1972)). Measurements were made from spring 1992 through November 1993 in populations that included at least 50 individuals.

2.2. Gas exchange - - midday comparisons At each of the sites (Table 1 ), measurements of A and gst were made at ambient temperature and humidity with levels of sunlight > 1500/xmol m -2 s-~ (photon flux density; PFD = 0.4-0.7/xm). Cloudcover at Whiteface River, MN and Clear Lake, IA precluded high PFD measurements at these sites. Three to nine plants were measured at each site and three replicate measures of A and gst were made on 6-10 cm 2 portions of an individual leaf for each plant. T. latifolia leaves are quite long (up to 1.5 m) and as many as six pairs emerge from the stem-leaf base near the water level (Constable et al., 1992). We consistently sampled the upper 1/3 of the canopy and selected mature leaves that showed little evidence of senescence (brown edges or tips). Gas exchange was measured with a closedflow gas exchange system (LICOR 6200, LICOR, Lincoln, NE.) and a 1/4 1 plexiglass cuvette. Leaves were sealed into the cuvette and the rate of net CO2 uptake (typically 5-10 /zl 1-~ in less than 30 s) was determined. Humidity was held constant ( + 2%) during measurements by routing a portion of the flow through a desiccant. Leaf and air temperatures in the cuvette remained within 1.5°C of air temperature and the cuvette was opened and flushed with ambient air between measurements. Repeated measurements were made to Table 1 Location and description of T. latifolia dominated wetlands sampled in 1992 and 1993. Air temperature (T~r) on the date of mid-season sampling is also shown Site

Latitude

Wetland type

T~r (°C)

Everglades, FL Sevilleta, NM Cimmeron River, KS Konza Prairie, KS Ithaca, NY Clear Lake, IA Cherokee Marsh, WI Cedar Creek, MN Whiteface River, MN

25 ° 50'N 34 ° 20'N 37 ° 20'N 39 ° 20'N 42 ° 30'N 43 ° 15'N 43 ° 20'N 45 ° 20'N 47 ° 10'N

Slough Marsh River bank Seep Marsh Lake edge Marsh Marsh Marsh

32.7 34.5 31.0 28.0 21.5 19.3 26.0 23.5 19.5

206

A.K. Knapp, J.B. Yavin/Aquatic Botany 49 (1995) 203-215

insure that steady-state responses were achieved. Ambient humidity, leaf and air temperature and PFD were recorded at each site.

2.3. Responses to light Gas exchange responses to PFD were assessed in three ways. In a T. latifolia population in Kansas, a portion of a leaf was secured in the chamber and A and g~t were measured as described above. Light levels were then decreased with neutral density screens to ca. 1000, 600, 475 ~mol m - 2 s - 1, etc. and after a new steady-state response was achieved, A and gst were again measured. The minimum PFD leaves were exposed to was ca. 50/zmol m -2 s - ~. These measurements were replicated on separate plants. To determine the response of A and gst to pre- and post-dawn changes in PFD, plants from a population in New Mexico were measured. The low humidity and warm temperatures at this site allowed us to make gas exchange measurements early in the morning without problems associated with dew formation on leaves. For a 1 h period that began just prior to sunrise when PFD was < 30/~mol m - 2 s - ~until PFD increased to ca. 190/xmol m - 2 s - t, three replicate measures of A and gst were made in separate individuals located within 2 m of each other. A total of four plants were measured. Finally, gst was measured at night (2300 CDT) in August, 1992 in a Kansas population. A diffusion porometer (model Mk3, DeltaT Devices, Cambridge, UK) was used to measure abaxial and adaxial gst on leaves from five plants. Because gst was expected to be low at night, special care was taken to repeatedly calibrate the porometer before and after individual plant measurements to insure that the instrument was operating correctly.

2.4. Diurnal variation Diurnal variability in gas exchange was examined in the Florida population. Diurnal measures of gas exchange were made by sampling three individual plants (three replicates per plant) every 2 h in July and November 1993. Midday measurements of A and gst were also made in March and combined with the July and November data to provide a seasonal perspective at this site.

2.5. Measurements along the length of leaves Because T. latifolia leaves may be > 1.5 m in length, and both age and the light environment may vary significantly along the length of leaves, we quantified differences in gas exchange at three locations on leaves. We measured A and gst on leaves at the upper 1/ 3 of their length, the middle 1/3 and the lower 1/3 where leaves were just emerging from their bases. Measurements were made on three individuals at each leaf location and were repeated in Kansas in July, Wisconsin in August and Florida in November 1993. All leaves were exposed to full sunlight until a steady-state response was achieved.

2.6. Responses to C02 The response of A to intercellular CO2 (Ci) was determined for plants in Kansas and Florida in July 1993. Leaves were sealed in the gas exchange cuvette with oil-based putty

A.K. Knapp, J.B. Yavitt /Aquatic Botany 49 (1995) 203-215

207

1/1

I

30.0

1200

Typha latifolia

3

o

mt

25.0

7

E "5 E

20.0

lO.O

~

5.0

~.

0.0

5"

40O 3 o 3

~' .c

Q.

P~

15.0

I

24

28

R.

40

36

32 NM

KS

KS

44NY Wl

MN

0

Latitude (o)

Fig. 1. Midday rates of net photosynthesis and stomatal conductance in T. latifolia populations ordered by latitude. Values are means of three to nine plants sampled when light levels were > 1500 p,mol m -2 s - t photon flux density• Error bars represent the standard error of the mean. The solid and dashed horizontal lines indicate the overall (all sites combined) means for net photosynthesis and stomatal conductance, respectively. Two additional sites were sampled but data are not shown because cloud cover led to reduced light levels. Letters along the xaxis indicate the state in which populations were located (Table 1 ).

to supplement the foam gasket in the cuvette. Prior to measurement of A, external C O 2 w a s varied by exhaling into the cuvette or by routing air flow through soda lime (McDermitt et al., 1989). In this manner, various external CO2 concentrations were achieved and A was measured periodically as the CO2 concentration decreased in the cuvette. When leaf and air temperatures increased to > 3°C above ambient, the cuvette was opened and flushed with ambient air. At high and low CO2 concentrations in the cuvette, the steep diffusion gradient between inside and outside the cuvette may result in significant errors due to CO2 leakage. These were small, but quantified by inserting a physiologically inert leaf replica into the cuvette and quantifying the leakage into or out of the cuvette. A and Ci were calculated according to the equations in McDermitt et al. (1989) and corrected for leaks at the various levels of CO2.

2. 7. Data analysis Correlation analysis was used to examine relationships between A vs. gst, and gst vs. atmospheric humidity, leaf temperature and the leaf to air vapor pressure difference across all sites combined. ANOVA was used to detect statistically significant (P < 0.05) differences in A and gst according to time of year or location on the leaf (Statistix, 4.0, Analytical Software, St. Paul, MN). Multiple regression and stepwise techniques were used to assess the relative influence that environmental variables such as temperature and humidity had on gas exchange parameters. For comparisons of two samples, t-tests were used. 3. R e s u l t s

3.1. Variation among populations At mid-season,A and gst varied significantly among the sites sampled (Fig. 1 ). Maximum A measured was 27.4 + 0.4 (SE) /zmol m - 2 s - 1 in the New Mexico site or 53% greater

208

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

than the m i n i m u m A o f 17.9 _ 0.8 /zmol m - 2 s - I at Cedar Creek, MN. These values represent mean values from each site. Maximum A measured in an individual T. latifolia leaf was 29.8 _ 0 . 2 / ~ m o l m - 2 s -1 at the Cimmeron River, Kansas site. Across all sites, mean A was 22.9 _ 1.2/xmol m - 2 s - ~ and ambient CO2 concentrations during measurements were consistently between 350 and 4 0 0 / z l 1-1. Stomatal conductance also varied significantly among sites (Fig. 1 ). Maximum gst was measured in Florida ( 1076.7 + 63.9 mmol m - 2 s - l ) and was three-fold higher than gst at Cedar Creek, M N ( 349.9 + 33.4 mmol m - 2 s - ~). Maximum gst measured in any individual plant was 1298.3 ___ 14.1 mmol m - 2 s - 1 in Florida. Across all sites, mean gst was 716.6 + 89.1 mmol m - 2 s - 1. Typha latlfolia--AII Sites r = 0.62

1200

900

600

300

5-03

0 1.0

1.5 2.0 2.5 3.0 Atmospheric Vapor Pressure (kPo)

I

3.5

E

I

B.

© 1200

E E v 0 C 0

Typha latifolia--AII Sites • r = 0.71

•

900

600

300

"0 C © (D

0 25.0

30.0

-6

35.0

Leaf Temperature

(°C)

0

E ©

1200

C.

o

U3 900

600 •

300

Typho Iotifolio--AII Sites r = 0.67

0 1.0

1.5 2.0 2.5 3.0 Leaf - Air V a p o r P r e s s u r e D e f i c i t ( k P a )

T. latifolia and ( A ) atmospheric v a p o r pressure, ( B ) leaf temperature and (C) the l e a f to air vapor pressure deficit. Data are from all seven sites sampled at mid-season at high PFD and from early and late season measurements in Florida. Fig. 2. Relationship between stomatal conductance in

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

30.0

209

1000

Typho Iotifolio - -

Konzo Proirie

20.0 Im

~.~ ~"

•

o

600

o

~o.o 5.o

20o

0.0 500

1000

"~" 5 _~ ";

0 2000

1500

Photosynthetic Photon Flux Density O.~rno; m - 2 s - 1 )

Fig. 3. Responses in net photosynthesis and stomatal conductance to photosynthetic photon flux density in T. latifolia from a Konza Prairie, Kansas population. Data are from two plants measured on separate days in August 1992. Neutral density screens were used to reduce light levels incident on entire leaves.

2000

o _

.

30 T o

. . . . .

,**.-i

i

il

26 22

},o 6 2 -2

Fig. 4. Diurnal course of net photosynthesis and stomatal conductance in T. latifolia from a population adjacent to Everglades National Park, FL. Photon flux density (PFD) is also shown for these mid-season and late-season days. Vertical bars indicate standard error of the means (n = three plants or PFD measurements). Net photosynthesis a n d gst were significantly and positively correlated with each other w h e n data from all sites were c o m b i n e d ( r = 0.66). This analysis included data from Fig. 1 plus gas e x c h a n g e data collected under other e n v i r o n m e n t a l conditions encountered in the field. In addition, gst was significantly and positively correlated with atmospheric vapor pressure, leaf temperature and the leaf-to-air vapor pressure deficit (Fig. 2). These latter analyses i n c l u d e d the full sunlight, m i d d a y measurements from Fig. 1 as well as early and late season data from the Florida site. 3.2. Gas exchange responses to PFD In the K o n z a Prairie, K a n s a s population, both A and gst achieved 90% of their m a x i m u m full sunlight levels at ca. 1 2 0 0 / z m o l m - 2 s - 1 (Fig. 3). However, most of the reduction in

A.K. Knapp, J.B. Yavitt /Aquatic Botany 49 (1995) 203-215

210

A occurred as PFD was decreased below 600/zmol m - 2 s - 1, whereas most of the reduction in gst occurred below 400/xmol m - 2 s - 1. Diurnal measurement of A and gst in Florida at mid-season and in November, 1993 confirmed that significant variability in A and g,t occurred as PFD varied during the day (Fig. 4). Maximum g,t in July and November occurred at mid-morning when atmospheric humidity was high and PFD was greater than 400/~mol m - 2 s - 1. In contrast, maximum A occurred when PFD was greatest (Fig. 4). Significant reductions in A and gst were measured at dawn and at sunset on both days, consistent with data in Fig. 3. To characterize better the early morning dynamics of A and g, plants from the New Mexico site were measured at 10--15 min intervals beginning prior to sunrise (Fig. 5). These data indicate that gst was ca. 200 mmol m - 2 s - 1 in T. latifolia early in the morning prior to sunrise (PFD = 26.4 ~mol m -2 s - 1) and that gst increased by three-fold during the hour after sunrise when PFD increased to 190 ~mol m -2 s - 1. ¢. 10,0

'

'~ ~E

/

'

'

1000 o3

T / ~

Typho latifolla - New Mexico

/

8.0

6.ot

o

8o0 o 600

~~

,oo ~ ~ ~ n~

2.0

z

o.o/o

3 200 .................

6:15

,

,

,

6:25

6:35

6:45

o

6:55

Time of Doy (MDT)

G

/ =1

0 / o

Fig. 5. Early morning course of increase in photon flux density (PFD), and net photosynthesis and stomatal conductance in T. latifolia from a New Mexico population at mid-season. Measurements were made prior to and after direct PFD was incident on leaves. Vertical bars indicate standard errors of the mean (n = three plants).

Typho Iotifolio -

3O

~, E -5 E

r~

25

Florido

1500 Photosynthesis Stomoto[ C o n d u c t o n c e

1200

5" -

o~

.=

2O 900 15

c

g

a:

~

~"

5 0

600

~" 3 o__

300

3t

0 Morch

July

November

1993 Fig. 6. Seasonal variation in midday values of net photosynthesis and stomatal conductance in T. latifolia from a population located adjacent to Everglades National Park, FL. At each date, gas exchange was measured at PFD > 1500 p,mol m -2 s - 1 and at ambient temperature and humidity. Error bars indicate standard error of the means (n = three plants). ANOVA indicated that A decreased significantly (P < 0.05) during the growing season and that g= in July was significantly higher compared with other sampling days.

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

211

Typha Iotifolio

3O

on Leaf Leafbose = Middle;;33 [Z~ = Upper

Location =

c~

25

E -6 E

20 15 ,0

g.

5 0

Wisc

I¢o

Karts

1500 Location on Leaf = Leaf base EE;S = Middle 1 / 3 r~l Upper 1 / 3

? E 1200 -5 E 900 o

Mean

FIo

600

o

-6 "6 E 0

3OO

o Wisc

Kans

glo

Mean

Fig. 7. Variation in net photosynthesis and stomatal conductance along the length of T. latifolia leaves. Three populations were sampled and mean values for the three populations combined are also shown (with standard error bars; n -- 3). All measurements were made at PFD > 1500/,Lmol m -2 s -1 and ambient temperature and humidity. Lower portions of leaves (near leaf bases) were allowed to acclimate to high PFD prior to measurements. ANOVA for all sites combined indicated that both A and g~t decreased significantly (P < 0.05) from the upper (older) portions of leaves to the leaf bases.

4

0

-

Typha latifolia

I~n I

E

Kansas & Florida Sites 30

E

20 I

10 ©

oo.~_ •

x c i / c a = 0.78 (AI

es)

I o~

1O0

200

300

400

500

600

Intercellular CO 2 (/~1 I - 1 )

Fig. 8. Response of net photosynthesis to intercellular CO2 concentration on T. latifolia. Data are from plants (n = 3) located at the Konza Prairie, KS and the Everglades, FL sites. Also shown is the mean ratio of intercellular CO~ (C~) to ambient CO2 (C,) for all plants from the seven populations sampled at mid-season in full sunlight.

212

A.K. Knapp,J.B. YavittlAquatic Botany49 (1995)203-215

Finally, measurements of gst at night in August in a Kansas T. latifolia population indicated that both abaxial and adaxial gst was zero. 3.3. Variation throughout a year

In the Florida T. latifolia population, midday A and gst were sampled in March, July and November (Fig. 6). Net photosynthesis was greatest early in the season and decreased significantly (P < 0.05) throughout the year. In contrast, gst was significantly highest in July when temperature and humidity were highest (Fig. 6). For both A and gst, minimum values were measured in November. 3.4. Variation along the length of leaves

Variation in gas exchange owing to the measurement location on leaves was determined at mid-season in the Wisconsin and Konza Prairie, Kansas populations and in November in the Florida population (Fig. 7). At all sites, significantly higher A was measured on the upper 1/3 of T. latifolia leaves and minimum A usually occurred near the leaf bases. When data were combined for all three sites, there were significant (P < 0.05) reductions in A as measurement sites varied from the top to the bottom of the canopy (Fig. 7). A similar pattern was noted for gst with highest values always in the upper portions of the canopy and minimum gst near the leaf base. 3.5. Responses to C02

In general, A increased in response to increasing intercellular CO2 (Ci) until Ci reached ca. 300/zl 1-1 (Fig. 8). At Ci above 300/.d 1-1, A did not increase. Across all sites, the average ratio of Ci to atmospheric CO2 (Ca) was 0.78 with an average Ci of ca. 280/zl lunder midday, full sunlight conditions. Thus, little stomatal limitation to A appeared to be present in these populations.

4. Discussion

There was significant variation among populations in net photosynthesis (A) and stomatal conductance (g~t) in T. latifolia at mid-season under full sunlight conditions in the field (Fig. 1). This variation ( 1.5 fold inA, 3 fold in g~t) may be due to the ecotypic variability documented in this species by McNaughton (1966), coupled with environmental differences among sites in air temperature and atmospheric humidity encountered in the field (Table 1, Fig. 2). Across the environmental gradient, gst increased with increasing temperature and humidity (Fig. 2). Combining temperature and humidity in a multiple regression model did not significantly increase the variance explained by temperature alone. Net photosynthesis was less strongly correlated with temperature and humidity (data not shown), but was strongly related to PFD (see below). In most terrestrial plants, gst decreases markedly as evaporative demand (atmospheric vapor pressure deficit or the leaf to air vapor pressure deficit (LAVPD)) increases (Losch and Tenhunen, 1981). In wetland species,

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

213

stomatal responses to increases in evaporative demand are often similar, but responses may also be less dramatic or none may be evident (Jones and Muthuri, 1984; Jones, 1987; Lafluer, 1988; Koch and Rawlik, 1993). For example, gst in T. domingensis actually increased slightly as atmospheric humidity decreased, and gst was constant as VPD varied in Cladiumjamaicense Crantz (Koch and Rawlik, 1993). In T. latifolia, when data were combined for all sites, gst increased as LAVPD increased (Fig. 3). This response to LAVPD is initially surprising. Indeed, if responses to LAVPD were assessed within a population, it is unlikely that gst would increase as LAVPD increases. Instead, these data suggest that the variation in temperature across sites more strongly influenced gst than the variability in LAVPD. Significant diurnal and seasonal variability in A and gst were evident in T. latifolia in the Florida population (Figs. 4 and 6). In a study of Florida populations of T. domingensis and Cladiumjamaicense, Koch and Rawlik (1993) reported only 15-19% changes in gst diurnally and seasonally. However, Jones (1987) reported that significant diurnal variation in gst (similar to that in Fig. 4) occurred in Cyperuspapyrus L. Most of the diurnal variability in A and gst could be explained by variability in PFD (Figs. 3 and 5), although highest gst was measured at midmorning when humidity was highest. However, a stepwise multiple regression model fit to these diurnal data did not include atmospheric vapor pressure as a significant variable with PFD. Diurnal patterns in A were most strongly related to PFD ( P = 0.90), with gst less strongly related to PFD (r z = 0.61). Plant water status was not measured concurrently, but it is possible that substantial diurnal variation occurred in leaf water potential even though plant bases were submerged at all sampling dates and locations. Such variability has been measured in other emergent aquatic plants (Jones and Muthuri, 1984; Drake, 1984). Complete stomatal closure does occur in T. latifolia at night (Chanton et al., 1993), but substantial stomatal opening occurs at very low PFD levels early in the morning (Fig. 5), potentially in response to low levels of blue light (Zeiger et al., 1981 ). Stomatal closure has also been reported in wetland Carex spp. in response to darkness (Mordssey et al., 1993), however, Jones (1988) reported that stomata of T. domingensis remained open when A was zero. Reduction in gst to zero at night in T. latifolia has important implications for estimates of CH4 emissions from wetlands dominated by this species. Indeed, stomatal closure at night is consistent with the reported accumulation of CH 4 within plants that occurs at night, as well as the subsequent morning peak in CH4 emission (Chanton et al., 1993; Yavitt and Knapp, 1994). Gas exchange characteristics in T. latifolia also varied significantly (two fold in A, almost three fold in gst; Fig. 7) depending on the portion of the leaf measured. Because T. latifolia leaves emerge from the base of the plant, these results suggest that A and gst increased in older portions of leaves. This result is most likely explained by the light environment experienced by different sections of these long leaves and not their age per se. Indeed, McNaughton (1973) reported that A decreased as leaves aged in T. latifolia and this trend was noted in the seasonal measurements from the Florida population (Fig. 6). Finally, although gs~ in T. latifolia varied with many environmental factors, gst does not appear to be limiting to A under typical field conditions. Although A and gst were significantly correlated (r = 0.66) in T. latifolia, the strength of this correlation was much lower than is often reported for terrestrial plants (for example, r = 0.85-0.95; Knapp, 1985; Yoshie,

214

A.K. Knapp, J.B. Yavitt /Aquatic Botany 49 (1995) 203-215

1986; Collatz et al., 1991). This suggests thatA and gst are not as tightly coupled in T.

latifolia as in plants in potentially more water-limited habitats. Moreover, the response of A to increasing Ci in T. latifolia (Fig. 8) is somewhat atypical of a C3 plant in that increasing Ci above levels typically measured in the field did not result in significant increases in A. This may be partially explained by the high concentration of CO2 in leaf aerenchyma (Constable et al., 1992). The A:Ci relationships developed in this study were done during the morning hours, when aerenchyma CO2 concentrations may be 18 times ambient (Constable et al., 1992). The effect that this potential CO2 source has on leaf-level A and Ci is unknown. Regardless, it appears that gst in T. latifolia is maintained at levels that reduce gas-phase limitations to A (Fig. 8). In summary, both A and particularly gst varied significantly in T. latifolia when compared among widely dispersed populations. Temperature was the environmental variable most strongly related to variations in A and gst under full sun conditions. Moreover, A and gst varied diurnally, seasonally and according to the position on the leaf measured. On a diurnal basis, light was the variable that explained most of the variance in A and gst. If stomatal opening influences trace gas flux from wetland plants, this variability in gst needs to be considered when regional and global estimates of emissions are made. Equally important, generalizations about wetland plants that suggest that gst is relatively unresponsive to environmental factors are not supported by the results of this study.

Acknowledgements Logistical assistance provided by the Sevilleta and Cedar Creek LTER sites is gratefully acknowledged. Research supported by funding from the National Geographic Society (4610-91 ), the National Science Foundation (DEB 90-11662) and the Kansas Agricultural Experiment Station (95-242-J).

References Chanton, J.P., Whiting, G.J., Happell, J.D. and Gerard, G., 1993. Contrasting rates and diurnal patterns of methane emission from emergent aquatic macrophytes. Aquat. Bot., 46:111-128. Collatz, G.J., Ball, J.T., Grivet, C. and Berry, J.A., 1991. Physiological and environmental regulation of stomatal conductance, photosynthesis and transpiration: a model that includes a laminar boundary layer. Agrie. For. Meteorol., 54: 107-136. Constable, J.V.H., Grace, J.B. and Longstreth, D.J., 1992. High carbon dioxide concentrations in aerenchyma of Typha latifolia. Am. J. Bot., 79: 415-418. Drake, B.G., 1984. Light response characteristics of net CO2 exchange in brackish wetland plant communities. Oecologia, 63: 263-270. Hotchkiss, N., 1972. Common Marsh, Underwater and Floating-leaved Plants. Dover Publications, NY, 123 pp. Jones, M.B., 1987. The photosynthetic characteristics of papyrus in a tropical swamp. Oecologia, 71: 355-359. Jones, M.B., 1988. Photosynthetic responses of C3 and C4 wetland species in a tropical swamp. J. Ecol., 76: 253262. Jones, M.B. and Muthuri, F.M., 1984. The diurnal course of plant water potential, stomatal conductance and transpiration in a papyrus (Cyperus papyrus L.) canopy. Oecologia, 63: 252-255. Knapp, A.K., 1985. Effect of tire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairie. Ecology, 66:1309-1320.

A.K. Knapp, J.B. Yavitt/Aquatic Botany 49 (1995) 203-215

215

Knapp, A.K. and Yavitt, J.B., 1992. Evaluation of a closed-chamber method for estimating methane emissions from aquatic plants. Tellus, 44B: 63-71. Koch, M.S. and Rawlik, P.S., 1993. Transpiration and stomatal conductance of two wetland macrophytes ( Cladium jamaicense and Typha domingensis) in the subtropical Everglades. Am. J. Bot., 80:1146--1154. Lafluer, P., 1988. Leaf conductance of four species growing in a subarctic marsh. Can. J. Bot., 66: 1367-1375. Losch, R. and Tenhunen, J.D., 1981. Stomatal responses to humidity-phenomenon and mechanism. In: P.G. Jarvis and T.A. Mansfield (Editors), Stomatal Physiology. Cambridge University Press, Cambridge, UK, pp. 137161. Mallik, A.U. and Wein, R.W., 1986. Response of a Typha marsh community to draining, flooding, and seasonal burning. Can. J. Bot., 64: 2136-2143. Mashburn, S.J., Sharitz, R.R. and Smith, M.H., 1978. Genetic variation among Typha populations of the southeastern United States. Evolution, 32: 681-685. McDermitt, D.K., Norman, J . i . , Davis, J.T., Ball, T.M., Arkebaner, T.J., Welles, J.M. and Roemer, S.R., 1989. CO2 response curves can be measured with a field-portable closed-loop photosynthesis system. Ann. Sci. For., 46: 416-420. McNanghton, S.J., 1966. Ecotype function in the Typha community-type. Ecol. Monogr., 36: 297-325. McNaughton, S.J., 1973. Comparative photosynthesis of Quebec and California ecotypes of Typha latifolia. Ecology, 54: 1260-1270. Morrissey, L.A., Zobel, D.B. and Livingston, G.P., 1993. Significance of stomatal control on methane release from Carex-dominated wetlands. Chemosphere, 26: 339-355. Murkin, H.R., Kaminsld, R.M. and Titman, R.D., 1982. Responses by dabbling ducks and aquatic invertebrates to an experimentally manipulated cattail marsh. Can. J. Zool., 60: 2324-2332. Nouchi, I., Mariko, S. and Aold, K., 1990. Mechanism of methane transport from the rhizospbere to the atmosphere through rice plants. Plant Physiol., 94: 59-66. Schutz, H., Schroder, P. and Rennenberg, H., 1991. Role of plants in regulating the methane flux to the atmosphere. In: T.D. Sharkey, E.A. Holland and H.A. Mooney (Editors), Trace Gas Emissions by Plants. Academic Press, NY, pp. 29-63. Sebacher, D.I., Harriss, R.C. and Bartlett, K.B., 1985. Methane emissions to the atmosphere through aquatic plants. J. Environ. Qual., 14: 40-46. Smith, S.G., 1967. Experimental and natural hybrids in north American Typha (Typhaceae). Am. Midl. Nat., 78: 257-287. Torn, M.S. and Chapin, F.S., 1993. Environmental and biotic controls over methane flux from arctic tundra. Chemosphere, 26: 357-368. Van der Valk, A.G,, 1981. Succession in wetlands: a Gleasonian approach. Ecology, 62: 688-696. Welsher, S.E.B., 1993. Long-term competitive displacement of Typha latifoliaby Typha angustifolia in a eutrophic lake. Oecologia, 94:451-456. Yavitt, J.B. and Knapp, A.K., 1995. Latitudinal and temporal trends in CI-14emission from cattail plants. Limnol. Tellus, (in review). Yoshie, F., 1986. Intercellular CO2 concentration and water-use efficiency of temperate plants with different lifeforms and from different microhabitats. Oecologia, 68: 370-374. Zeiger, E., Field, C. and Mooney, H.A., 1981. Stomatal opening at dawn: possible roles of the blue light response in nature. In: H. Smith (Editor), Plants and the Daylight Spectrum. Academic Press, NY, pp. 391-407.