Consistent and variable leaf anatomical characters ... - Oxford Journals

2 downloads 0 Views 2MB Size Report
seven quantitative leaf anatomy characters from transverse leaf sections to test their .... freezing microtome. .... Numbers refer to list of characters in Table 2.
bs_bs_banner

Botanical Journal of the Linnean Society, 2013, 172, 371–384. With 2 figures

Consistent and variable leaf anatomical characters in Carex (Cyperaceae) CLARE BUGG1, COLIN SMITH1, NIGEL BLACKSTOCK2, DAVID SIMPSON3 and PAUL A. ASHTON1* 1

NGAS, Edge Hill University, St Helens Road, Ormskirk, Lancs., L39 4QP, UK Liverpool Community College, Bankfield Road, Liverpool, Merseyside, L13 0BQ, UK 3 Royal Botanic Gardens Kew, Richmond, Surrey, TW9 3AB, UK 2

Received 16 May 2012; revised 28 September 2012; accepted for publication 5 January 2013

The anatomy and morphology of leaves in Carex have the potential to be taxonomically useful. However, studies on the variability of leaf characteristics in the genus are sparse. Researchers therefore risk using leaf anatomical characters without the knowledge of whether they are consistent in a species. We examined 22 qualitative and seven quantitative leaf anatomy characters from transverse leaf sections to test their consistency across 11 Carex spp. The characters were clearly described and primarily microscopic. Some characters were found to exhibit high levels of intraspecific variation, whereas other characters exhibited high levels of consistency in a species, including the shape of the leaf section, the density of papillae and the size of epidermal cells. Caution must be applied when choosing leaf anatomy to delimit taxa because of the intraspecific variability found in some characters, but sufficient numbers of invariant characters exist to provide useful taxonomic separation. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384.

ADDITIONAL KEYWORDS: morphology – phenotypic plasticity – sedges – taxonomy – transverse section – variation.

INTRODUCTION Cyperaceae, with 106 genera and c. 5400 species (Govaerts et al., 2007), especially the largest genus, Carex L. with c. 1800 species (Govaerts et al., 2007), have historically been considered taxonomically difficult. This perception is a product of the great diversity, coupled with the characteristics of reduced reproductive organs, lack of distinctive coloration and ability to hybridize. Consequently, non-floral characters, such as the morphology and anatomy of the rhizomes, roots, culms and leaves, have been utilized as an aid in taxon delimitation. This accounts for the frequent inclusion of some of these characters in identification field guides and handbooks (e.g. Jermy et al., 2007; Poland & Clement, 2009; Stace, 2010). The use of macromorphological characters, i.e. those observable without the aid of a microscope *Corresponding author. E-mail: [email protected]

(Naczi, 2009), for identification in Cyperaceae can be problematic as they may exhibit phenotypic plasticity related to habitat type (Szczepanik-Janyszek & Klimko, 1999; Stenstrom, Jonsdottir & Augner, 2002). Gross morphological features are not always sufficiently reliable to provide clear boundaries between taxa, with the result that anatomical features, which tend to be less strongly environmentally modified, have been examined for reliable taxonomic separation (e.g. Metcalfe, 1971; Toivonen & Timonen, 1976; Menapace, Wujek & Reznicek, 1986). Several studies have demonstrated the utility of leaf anatomy in separating closely related species and in determining hybrids. For example, Starr & Ford (2001) utilized the distribution of surface hairs and papillae along with silica body characteristics in recognizing eight species in Carex section Phyllostachys (J.Carey) L.H.Bailey. Similarly, the closely related C. otrubae Podp. and C. vulpina L. have been found to differ in various anatomical characteristics (Porley,

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

371

372

C. BUGG ET AL.

1999; Smith & Ashton, 2006). At the sectional level, epidermal characters were used to distinguish species of Carex section Phacocystis Dumort. (Standley, 1987, 1990) and anatomical characters helped define major clades in three sections of Carex (Naczi, 2009). In contrast, Szczepanik-Janyszek & Klimko (1999) noted that leaf anatomical characters were ineffective in the taxonomic classification of Carex section Muehlenbergianae (L.H.Bailey) Kük. in Poland, but only four anatomical characters were included in their analysis. Starr & Ford (2001), reviewing previous studies, questioned whether anatomical characters are inherently unreliable or whether previous negative conclusions regarding their phylogenetic utility in Carex could be explained by factors such as personal bias in selection of characters, poor circumscription of characters or limited sampling. Intraspecific anatomical variation is sparsely documented and little understood. Molina, Acedo & Llamas (2006) demonstrated that some anatomical variation in C. hirta L. was related to water availability, but some characters were variable even under constant conditions. This work was undertaken using a hydroponics approach rather than studying the variation present naturally in a species and therefore may not represent the range of environmental variation present. This study addresses the extent of intraspecific variation in leaf characters from transverse leaf sections in Carex. To avoid the pitfalls of bias, circumscription and small sample size identified by Starr & Ford (2001), a range of clearly circumscribed species of contrasting ecology, geography and taxonomy are considered, with a large number of leaf characters recorded without a priori expectations across a representative sample size of each species.

MATERIAL AND METHODS Mature vegetative leaves were collected from a range of locations, between June 2001 and June 2010, primarily in the UK. Leaves from healthy, typical individuals were collected and preserved in the field by wrapping in wet tissue paper and placing in a plastic bag, with transfer to a standard refrigerator within 24 h. Within 2 weeks of collection, transverse sections from the central section of the leaf were cut with a freezing microtome. Once embedded in a fixing compound, sections between 20 and 50 mm thick were cut, depending on the characteristics of the material. Cut sections were double stained with Safranin and Light Green and then mounted in Canada balsam. Five species were chosen for a large sample set (C. flacca Schreber, C. filiformis L, C. otrubae, C. remota L and C. vulpina). These represent species from subgenus Carex (C. filiformis and C. flacca) and

subgenus Vignea (C. remota, C. otrubae and C. vulpina). They also include widespread UK species with a broad ecological amplitude (C. otrubae, C. remota and C. flacca) and species with precise habitat requirements and restricted distributions (C. filiformis and C. vulpina). Abundance is approximated by the number of UK hectads each species was present in from 2000 to 2009 ((Botanical Society of Britain and Ireland, 2011). Collection covered both the ecological and geographical ranges of each species in the UK. Additionally, where possible, material from outside Britain was also included in the study. Species details are presented in Table 1. Sample sites are available on request from the corresponding author. Voucher specimens are deposited at LIV. Twenty-five leaf sections for each of the five species were examined using a Leitz Diaplan microscope. Each section was taken from a different individual from a different sample site. For each section, the presence or absence of 22 qualitative anatomical characteristics was recorded and seven quantitative measurements were taken using a Delta Pix Invenio 5D 5M pixel CCD camera with DpxViewPro software. Characters for analysis were chosen based upon previous studies (Metcalfe, 1969; Standley, 1987; Porley, 1999; Starr & Ford, 2001; Molina et al., 2006; Smith, 2007). A number of these were refined to provide quantitative data or to reduce ambiguity. One novel character (the angle of leaf margin) was added following experience with the material; the angle was taken at each margin of the leaf section and an average calculated. The characters measured are shown in Table 2 and illustrated in Figures 1 and 2. The presence of each qualitative character was expressed as a percentage of the total number of leaf sections in that species. Characters present or absent in ⱖ 90% of leaf sections for a species were considered to be consistent. Mean and standard deviation were calculated for all quantitative characters. To allow further comparison of the variability of characters, a coefficient of variation (CV) was calculated. To test the reliability of the findings, a further sample set was examined. Rolling means from the first sample set suggested that ten samples would be sufficient to elicit consistency of a character in a species and therefore ten samples of a further six species were studied: C. limosa L., C. nigra (L.) Reichard, C. salina Wahlenb. and C. saxatilis L. of subgenus Carex and C. leporina L and C. paniculata L. of subgenus Vignea. These include common species with broad habitat requirements (C. nigra, C. leporina and C. paniculata) and species that are rarer and more habitat specific (C. saxatilis, C. limosa and C. salina). Collection was made from a variety of sites largely in the UK, but also from elsewhere in Europe.

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

LEAF ANATOMY IN CAREX (CYPERACEAE)

373

Table 1. Species, sample size, subgenus, geographical spread and habitat of the 11 species used in the study Species

Sample size

Subgenus

Geographical spread

Habitat

C. paniculata

10*

Vignea

Fens and wet woodland

C. vulpina

25*

Vignea

C. otrubae

25*

Vignea

C. remota

25*

Vignea

C. leporina

10

Vignea

C. saxatilis

10

Carex

C. flacca

25*

Carex

C. filiformis

25*

Carex

C. limosa

10

Carex

C. salina

10*

Carex

C. nigra

10*

Carex

Widespread 734 Restricted 13 Widespread 1028 Widespread 1349 Widespread 1527 Restricted 19 Widespread 2069 Restricted 11 Restricted 151 Restricted 1 Widespread 1719

Damp grassland Damp grassland and coastal Woodlands Damp grassland. Montane bogs Grassland Species-rich mesotrophic grassland Bogs Saltmarsh Damp acid grassland and bogs

*Denotes sample included non-UK material. The value given in the geographical spread column represents the number of 10 ¥ 10 km squares in which the species is recorded in Britain and Ireland.

Qualitative character consistency between species from the two subgenera and between restricted and widespread species was compared (Table 4). Similar comparisons were also made for quantitative characters (Tables 5 and 6). The potential relationship between variation and extent of distribution was investigated using Spearman’s rank correlation coefficient. The same statistic was used to test the strength of the relationship between numbers of consistent quantitative characters and mean CV in a species.

RESULTS OVERVIEW Only two qualitative characters reveal no intraspecific variation across the species studied: abaxial surface densely papillose and stomata present on abaxial leaf surface. The first character is variable between species, the second was uniform in the species examined here (Tables 3 and 4). In addition, four characters (4, 5, 8 and 22) are consistent in all but one species. There are a similar number of consistent qualitative characters in each of the two subgenera, although some characters are consistent in one subgenus and not the other. In addition to the characters listed above (10 and 12) that are consistent in both subgenera, subgenus Vignea exhibits intraspecific consist-

ency for four other characters (6, 8, 14 and 20), whereas subgenus Carex exhibits intraspecific consistency for characters 4, 7, 18, 19 and 22. Similarly, when quantitative characters are compared, members of subgenus Vignea have similar levels of mean CV to those of subgenus Carex. There is no significant difference between the number of intraspecifically consistent characters found in restricted and widespread species for qualitative characters (Table 4) and no relationship between UK abundance and mean CV (r = -0.24, NS). Similarly, there is no relationship between qualitative variation (measured by number of variable characters per species) and mean CV (r = 0.02, NS). Two groups of characters are revealed as alternatives for the same feature. Air cavities are either predominantly quadrate (18) or transversely elongate (19). No other shape was recorded in this survey. Likewise, leaf section shape always fell into one of the three categories (dorsiventrally flattened with a prominent midrib on the adaxial surface = strongly keeled, V-shaped or thinly crescentiform; 20, 21 and 22).

SPECIES Carex leporina, C. remota, C. flacca, C. nigra and C. salina were the most consistent in qualitative

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

374

C. BUGG ET AL.

Table 2. Qualitative and quantitative anatomical characters used in the study Qualitative characters (yes/no)

Reference*

15 16 17

> 50% of vascular bundles round in shape > 50% of vascular bundles positioned closer to abaxial than adaxial surface > 75% of vascular bundles a similar size > 50% of vascular bundles forming girders that span width of lamina Vascular bundles at midpoint of lamina more prominent than others Vertical length of adaxial epidermal cells less than or equal to vertical length of abaxial epidermal cells Papillae present on adaxial surface Adaxial surface densely papillose Papillae present on abaxial surface Abaxial surface densely papillose Stomata present on abaxial leaf surface Stomata present on adaxial leaf surface Bulliform cells in more than one layer Bulliform cell vertical length more than adaxial epidermal cell vertical length (average) Silica bodies present on either surface Silica body at leaf margin on least one margin of leaf section Surface hairs present on either epidermis

18 19 20 21 22

> 50% > 50% Shape Shape Shape

1 2 3 4 5 6 7 8 9 10 11 12 13 14

air cavities ± quadrate air cavities ± transversely elongate of leaf section strongly keeled of leaf section V-shaped of leaf section thinly crescentiform

Smith (2007) Smith (2007) Metcalfe (1969) Porley (1999) Smith (2007) Starr & Ford (2001), Standley (1987) Standley (1987) Smith (2007) Standley (1987) Starr & Ford (2001), Smith (2007) Smith (2007) Starr & Ford (2001), Smith (2007) Porley (1999), Metcalfe (1969) Starr & Ford (2001) Smith (2007) Smith (2007) Starr & Ford (2001), Molina et al. (2006) Porley (1999), Metcalfe (1969) Porley (1999) Smith (2007) Smith (2007) Smith (2007)

Quantitative characters 23 24 25 26 27 28 29

Total number of vascular bundles Total number of bulliform cells Width of adaxial epidermal cell Length of bulliform cells Angle of leaf margin (average) Depth of keel Angle of keel

Standley (1987) Porley (1999) Porley (1999) Porley (1999) Novel Smith (2007) Porley (1999), Starr & Ford (2001)

*Where the reference is presented in bold, the character was modified from the original study to reduce ambiguity or provide quantitative data.

characters (18 or 19 consistent characters out of 22) and C. vulpina and C. otrubae the most variable. Carex leporina and C. otrubae were the species with the most consistent quantitative characters. The most variable characters were seen in C. limosa, C. paniculata, C. nigra and C. salina (see Tables 5 and 7).

QUALITATIVE

CHARACTERS

The consistency of each qualitative character in each species is presented in Table 3. This is summarized across all species in Table 4. Vascular bundles Vascular bundle characters were generally found to be variable in a species. The shape, position and size

were the least consistent characters, with six or seven of the 11 species showing variability. The vascular bundle at the midpoint of the lamina was consistently prominent in five species, a similar size to other vascular bundles in one species (C. paniculata) and variable in four species. The presence of girders of sclerenchyma with embedded vascular bundles that spanned the width of the lamina was the only valuable vascular bundle character, with only C. otrubae showing variation with girders present in 48% of samples. Epidermal cells The length of the adaxial epidermal cells in relation to the abaxial epidermal cells was a highly consistent

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

LEAF ANATOMY IN CAREX (CYPERACEAE)

375

A

B

C

Figure 1. Qualitative characters. Numbers refer to list of characters in Table 2. A, leaf section of C. paniculata showing thinly crescentiform shape (22) approximately 100 mm depth at the keel. B, leaf section of C. remota showing V-shape (21) approximately 100 mm depth at the keel. C, leaf section of C. vulpina showing strongly keeled leaf shape (20) approximately 100 mm depth at the keel. D, microscope image of C. salina, showing: oval-shaped vascular bundle (1), forming girder (4), positioned closer to adaxial surface (2); papillae on abaxial surface (7); quadrate air cavity (18). © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

376

C. BUGG ET AL.

D

Figure 1. Continued

character across the eleven species studied. Only C. saxatilis showed any variability (with 80% of samples showing the character). Eight species had longer adaxial epidermal cells than abaxial epidermal cells, with one, C. nigra, having consistently longer epidermal cells on the abaxial than adaxial surfaces. Papillae Presence or absence of papillae on the adaxial surface was consistent in eight of the species examined (present in three, absent in five) and variable in three species. When adaxial papillae density was considered, ten species were consistent, with only one (C. nigra) being variable. Presence of papillae on the abaxial surface was consistent in nine species. Where present, they were uniformly dense, except in C. paniculata and C. nigra, in which they were only sparsely papillose. Stomata Stomata were universally present on the abaxial leaf surface of all leaf sections of all species examined. Presence of stomata on the adaxial leaf surface was a

variable character, with seven of the 11 species having some individuals with this character. Four species showed consistent presence of stomata on the adaxial surface (C. leporina, C. nigra, C. limosa and C. salina). Bulliform cells Three species consistently had bulliform cells in more than one layer (C. paniculata, C. nigra and C. salina) and four species consistently showed a single layer of bulliform cells. Four species had variable numbers of bulliform cell layers. The average vertical length of bulliform cells in relation to the average vertical length of adaxial epidermal cells was consistently longer in all but two species. Leaf surface characters In four species (C. leporina, C. paniculata, C. flacca and C. nigra), silica bodies on the leaf surface were uniformly absent. No species always exhibited this character, its frequency of occurrence varying from 4% (C. flacca) to 80% (C. paniculata). The presence of a silica body at either of the leaf margins was even

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

LEAF ANATOMY IN CAREX (CYPERACEAE)

377

A

B

Figure 2. Quantitative characters. Numbers refer to list of characters in Table 2. A, microscope image of C. salina showing measurement of bulliform cell length (14). Measurements were taken as an average of five cells. B, microscope image of C. otrubae showing measurement of width of epidermal cell (25). Measurements were taken as an average of ten cells. C, microscope image of C. nigra showing measurement of keel angle (29). D, microscope image of C. nigra showing measurement of leaf margin angle. Measurement was an average of both leaf margin angles. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

378

C. BUGG ET AL.

C

D

Figure 2. Continued

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

96 56 88 100 28 100

16 0 0 0 84 100 8 100 36 16 56 52 48 100 0 0 12

100 40 100 0 100

30 0 30 0 90 100 100 100 0 80 0 100 0 0 20 80 15

C. vulpina

70

C. paniculata

20 80 96 4 0 12

0

64 68

40 4 0 0 28 100 16 100

100

100

48 48

20

96

C. otrubae

96 4 0 100 0 17

0

20 28

0 0 0 0 72 100 0 100

100

100

100 0

28

100

C. remota

100 0 0 100 0 19

0

0 50

0 0 0 0 100 100 10 100

100

50

100 0

60

100

C. leporina

*Consistent characters (i.e. present in ⱖ 90% or ⱕ 10% samples) are marked in bold. †Species in subgenus Vignea are grouped to the left and those in subgenus Carex to the right.

1

> 50% of vascular bundles round in shape 2 > 50% of vascular bundles positioned closer to abaxial than adaxial surface 3 > 75% of vascular bundles similar size 4 > 50% of vascular bundles forming girders that span width of lamina 5 Vascular bundles at median point more prominent than others 6 Length of adaxial epidermal cells greater than or equal to length of abaxial epidermal cells 7 Papillae present on adaxial surface 8 Adaxial surface densely papillose 9 Papillae present on abaxial surface 10 Abaxial surface densely papillose 11 Stomata present on adaxial leaf surface 12 Stomata present on abaxial leaf surface 13 Bulliform cells in more than one layer 14 Bulliform cell length more than upper epidermal cell length (average) 15 Silica bodies present on either surface 16 Silica body at leaf margin on least one side of leaf section 17 Surface hairs present on either epidermis 18 Air cavities ± quadrate 19 Air cavities ± elongate 20 Leaf section strongly keeled 21 Leaf section V-shaped 22 Leaf section curved Total number of consistent characters (/22)

Qualitative character

% of samples with character present†

Table 3. Consistency of qualitative characters in each species*

100 0 0 100 0 16

0

10 30

0 0 0 0 30 100 0 80

80

50

100 100

100

60

C. saxatilis

100 0 0 100 0 19

0

4 0

0 0 100 100 72 100 36 100

100

92

84 100

96

8

C. flacca

0 100 100 0 0 15

56

36 16

0 0 100 100 44 100 0 100

100

96

84 100

24

72

C. filiformis

100 0 30 70 0 15

0

20 10

90 90 90 90 100 100 0 80

0

90

50 90

80

30

C. limosa

100 0 10 90 0 18

0

0 30

100 70 30 0 100 100 100 100

0

100

100 100

70

0

C. nigra

100 0 40 60 0 19

0

0 0

100 100 10 90 100 100 100 100

100

100

30 100

100

10

C. salina

LEAF ANATOMY IN CAREX (CYPERACEAE)

379

> 50% of vascular bundles round in shape > 50% of vascular bundles positioned closer to abaxial than adaxial surface > 75% of vascular bundles a similar size > 50% of vascular bundles forming girders that span width of lamina Vascular bundles at median point of lamina more prominent than others Vertical length of adaxial epidermal cells greater than or equal to vertical length of abaxial epidermal cells Papillae present on adaxial surface Adaxial surface densely papillose Papillae present on abaxial surface Abaxial surface densely papillose Stomata present on adaxial leaf surface Stomata present on abaxial leaf surface Bulliform cells in more than one layer Bulliform cell length greater than adaxial epidermal cell length (average) Silica bodies present on either surface Silica body at leaf margin on least one side of leaf section Surface hairs present on either epidermis > 50% air cavities ± quadrate > 50% air cavities ± transversely elongate Shape of leaf section strongly keeled Shape of leaf section V-shaped Shape of leaf section thinly crescentiform Total number consistent characters

*Bold indicates consistency within the category.

15 16 17 18 19 20 21 22

7 8 9 10 11 12 13 14

6

5

3 4

1 2

Qualitative character measured

0 0 0 8 1 3 5 0 2

3 2 4 4 5 11 6 9

9

7

4 8

4 5

5 3 9 1 8 6 3 10 3

5 8 5 7 0 0 3 0

1

1

0 2

3 0

6 8 2 2 2 2 3 1 2

3 1 2 0 6 0 2 2

1

3

7 1

5 6

2 0 4 3 3 5 4 4 5

2 5 4 5 2 5 4 5

5

3

2 4

4 1

4 3 5 6 6 4 4 6 7

6 5 5 6 3 6 5 4

5

5

2 6

3 4

Carex /6

Vignea /5

Variable

Always present

Always absent

Consistency in subgenera

Number of species (/11)

2 2 3 4 4 3 3 5 7

4 5 5 5 2 5 5 3

4

3

1 5

2 3

Rare /5

4 1 6 5 5 6 5 5 6

4 5 4 6 3 6 4 6

6

5

3 5

5 2

Common /6

Consistency and distribution

Table 4. Consistency of qualitative characters across the species studied (columns 2–4); columns 6 and 7 show characters consistent in the two subgenera and characters 9 and 10 show characters consistent in rare and common species*

380 C. BUGG ET AL.

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

13 22.9 25 36.4 19.9 19.6 23.7 27.2 32.1 31.7 17.2

Air cavities Air cavities were found to be either consistently quadrate in shape or consistently transversely elongate in shape in all species except C. vulpina and C. otrubae. Shape of leaf section In most species studied, the shape of the transverse leaf section proved to be a highly consistent character. Only three species showed marked intraspecific variability (C. paniculata, C. limosa and C. salina).

QUANTITATIVE

CHARACTERS

27.2 ± 3.1 23.5 ± 2.9 26 ± 4.6 33.1 ± 4.9 19.4 ± 2.1 21.9 ± 3 20.6 ± 3.1 18.8 ± 3.5 16.8 ± 2.5 11.3 ± 1.7 13 ± 1.5

The variability of each character in each species can be seen in Table 5. Table 6 shows the variation of each character across the species sampled. Table 7 shows overall variation present in each species.

SD, standard deviation; CV, coefficient of variation.

40.2 15.2 13.6 16.7 19.3 10.3 26.5 13.3 69.4 33 44.8 19 ± 7.63 8.5 ± 1.28 8.8 ± 1.56 6.5 ± 1.08 7.4 ± 1.43 11.3 ± 1.16 13.2 ± 3.52 10.2 ± 1.36 8.2 ± 5.7 18.9 ± 6.24 19.1 ± 8.56 14.1 15.5 12.9 7.4 15.7 10.7 14.6 13.5 18.7 5.3 7.4 19.6 ± 2.76 22 ± 3.4 22 ± 2.85 10.8 ± 0.80 14.7 ± 2.31 13.3 ± 1.42 17.6 ± 2.57 13.9 ± 1.88 11.1 ± 2.1 14.8 ± 0.79 14.9 ± 1.10 C. paniculata (10) C. vulpina (25) C. otrubae (25) C. remota (25) C. leporina (10) C. saxatilis (10) C. flacca (25) C. filiformis (25) C. limosa (9) C. nigra (10) C. salina (10)

381

more variable, with only C. flacca and C. salina showing complete absence. However, these results should be viewed with a degree of caution as the density of such bodies, although quite variable across Carex spp., is often so low that any one section may miss the bodies. The thickness of the section may also influence the ability to detect silica bodies. Surface hairs were consistently absent in nine of the 11 species examined. Where they were present, in C. vulpina and in C. filiformis, they were not universally present.

60.1 ± 7.8 37.0 ± 8.5 47.8 ± 11.9 76.2 ± 27.8 57.2 ± 11.4 84.4 ± 16.6 126.8 ± 30.1 96.6 ± 26.3 73.3 ± 23.5 93.6 ± 29.6 61.1 ± 10.5 23.2 26.9 15.3 19.2 16.9 22.9 23.5 18 16.8 32.2 20.4 125.9 ± 29.3 45.2 ± 12.1 55.9 ± 8.6 88.6 ± 16.9 69.5 ± 12.9 101.9 ± 25.6 85.9 ± 20.2 85.9 ± 15.5 86 ± 14.5 93.9 ± 33.3 98.8 ± 20.2 39.8 19.1 17.4 16.4 6.6 19.2 18 13.8 19.4 16.6 28.5 293 ± 116.5 333.5 ± 63.9 332.7 ± 57.8 241.4 ± 39.5 193.8 ± 12.8 241.5 ± 46.3 218.6 ± 39.3 170.9 ± 23.5 206 ± 40 169.5 ± 28.2 285 ± 81.2 24.9 20.3 16.5 16.8 13.7 32.4 14.8 22.3 43.1 30 26.6 46.2 ± 11.5 56.6 ± 11.5 64.7 ± 10.7 86.4 ± 14.5 53.3 ± 7.3 49.9 ± 16.2 56.9 ± 8.4 47.1 ± 10.5 30 ± 12.9 21.7 ± 6.5 36.1 ± 9.6 11.5 12.6 17.5 14.8 10.6 13.8 14.9 18.4 14.9 15 11.8

CV Mean ± SD CV Mean ± SD CV Mean ± SD CV Mean ± SD CV Mean ± SD Mean ± SD Species (no. of samples)

CV

CV

Mean ± SD

Mean angle of leaf margins (°) Mean angle of keel (°) Mean depth of keel (mm) Mean length of bulliform cell (mm) Mean adaxial epidermal cell width (mm) Total number of bulliform cells Total number of vascular bundles

Table 5. Descriptive analysis of the quantitative characters in each species

LEAF ANATOMY IN CAREX (CYPERACEAE)

Number of vascular bundles The mean number of vascular bundles in a species varied from 10.8 (C. remota) to 22.0 (C. vulpina and C. otrubae). Within a species it was a highly consistent character, having the lowest mean CV (12.3%) of all characters measured. Carex remota and C. nigra exhibited little variation for this character, whereas C. vulpina and C. otrubae show a high degree of variability; however, in both cases this can be attributed to the appearance of an extreme value in a single sample. Bulliform cells Total numbers of bulliform cells in a sample ranged from five (C. remota) to 32 (C. paniculata). Species exhibiting variation (C. salina, C. paniculata, C. nigra and C. flacca) were primarily those that had bulliform cells in more than one layer. All species with a single layer showed low levels of variation, except C. limosa. As a character, across the genus, the mean CV (27.5) was the highest, mainly attributable to the high variation present in the species with two layers. The mean length of bulliform cells showed a similar level of variation (CV 23.8), although this was independent of the species with two layers. The mean

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

382

C. BUGG ET AL.

Table 6. Variation in quantitative characters across the species studied

Mean SD CV

Total number of vascular bundles

Total number of bulliform cells

Mean adaxial epidermal cell width (mm)

Mean length of bulliform cell (mm)

Mean depth of keel (mm)

Mean angle of keel (°)

Mean angle of leaf margins (°)

15.9 2.0 12.3

11.9 3.6 27.5

21.1 3.0 14.4

49.9 10.9 23.8

244 49.9 19.5

86.5 21.7 21.3

74.0 18.5 24.4

Table 7. Coefficient of variation (CV) for all characters by species Species

CV

C. paniculata C. vulpina C. otrubae C. remota C. leporina C. saxatilis C. flacca C. filiformis C. limosa C. nigra C. salina

23.8 18.9 16.9 18.2 14.7 18.4 19.4 18.1 29.1 23.4 22.4

average length of bulliform cells ranged from 21.7 mm (C. nigra) to 86.4 mm (C. remota). Adaxial epidermal cell width The mean width of adaxial epidermal cells was one of the most consistent quantitative characters, with similar standard deviations (SD 1.7–4.9) and a low CV (14.4) across the genus. The narrowest cells were recorded in C. nigra (11.3 mm) and C. salina (13 mm) and the widest in C. remota (33.1 mm). Keel characters Keel depth varies in its consistency across the species considered. Carex leporina showed low levels of intraspecific variation (CV 6.6), whereas C. paniculata showed by far the largest variation (CV 39.8), with keel depths varying from 184.2 to 494.5 mm. Carex salina also had a high SD, although this can be attributed to the presence of one high value. Keel angle was slightly more variable across the genus, with broadly comparable levels of variation between species ranging from C. limosa (CV 18) to C. nigra (CV 32.2). The values for keel angle varied from the acute-angled C. vulpina (mean 45.2°) to the obtuseangled C. paniculata (mean 125.9°). Angle of leaf margins Leaf margin angle was the most variable of all characters measured, with the second highest mean CV

(24.4%). In particular, large SDs were recorded for C. remota (CV 36.4), C. limosa (CV 32.1) and C. nigra (CV 31.7), whereas C. paniculata was the least variable (CV 13). Carex vulpina had the most acute leaf margins (mean 37°) and C. flacca had the most obtuse (mean 126.8°).

DISCUSSION A number of the leaf anatomy characters examined in this study show a low degree of variability in species and thus may be considered useful taxonomic characters. These were: 1. 2.

The shape of the transverse leaf section. The relative size of adaxial to abaxial epidermal cells. 3. The presence of papillae in high densities. 4. The presence of girders. 5. Prominent mid-lamina vascular bundles. 6. The number of layers of bulliform cells. 7. The presence of stomata on the abaxial surface. 8. The shape of air cavities. 9. The size relationship between bulliform and epidermal cells. 10. The number of vascular bundles. 11. The width of adaxial epidermal cells. 12. The depth of the keel. Two of these characters (relative length of bulliform cells to epidermal cells and presence of abaxial stomata) were uniform across the species studied. However, this does not necessarily mean that these characters are uniform across the genus; further extensive sampling would be needed to reveal this. Relative length of bulliform and epidermal cells is previously undocumented, whereas variation in stomatal presence/absence on either or both surfaces has been shown to be a useful character in separating species in Carex section Phacocystis (Dean & Ashton, 2008). Stomatal size and density have been shown in a number of studies to be related to environment (e.g. Stenstrom et al., 2002), although work on C. hirta found no relationship between stomatal density and water availability, suggesting it was a genetically fixed character and of use taxonomically (Molina et al., 2006).

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

LEAF ANATOMY IN CAREX (CYPERACEAE) Other authors have previously described other characters as consistent and hence taxonomically valuable. The transverse leaf shape is widely included in field guides as a vegetative discriminant (e.g. Jermy et al., 2007; Poland & Clement, 2009) and Standley (1987) identified the relative length of the adaxial and abaxial epidermal cells as the most taxonomically useful cell dimension character in section Phacocystis. In addition, Molina et al. (2006) identified epidermal projections as the most consistent anatomical characters in C. hirta. While our study confirmed this finding for papillae at high density, it is not true for other epidermal characters (e.g. silica bodies and surface hairs). Although plastic changes in the gross morphology of sedges attributable to environmental differences is well documented (e.g. Naczi, Reznicek & Ford, 1998; Stenstrom et al., 2002), the effects on anatomical features are less well known. Environmental effects on the number of vascular bundles and cell dimensions in C. emoryi Dewey were identified by Standley (1987, 1990). Several studies in other plant groups have drawn similar conclusions, for instance in Festuca L. (Namaganda, Krekling & Lye, 2009) and Aegilops L. (Kharazian, 2007). A number of characters with high variability suggesting a major environmental influence were recorded in this study. These included the shape, position and size of vascular bundles, presence of stomata on the adaxial surface, presence of silica bodies, presence of surface hairs, length and number of bulliform cells and the angles of the keel and leaf margins. Because of the high level of variability in these characters, they should be excluded from taxonomic and phylogenetic studies. The shape of air cavities was a reliable character in all species except the closely related C. vulpina and C. otrubae. Molina et al. (2006) noted that increased water availability caused an increase in the size of air cavities. However, the general consistency across the species in this study suggests the cavity shape remains fixed. The variability of average length and number of bulliform cells has already been reported in C. hirta by Molina et al. (2006) and is confirmed in other species here. The highest levels of variation in size and numbers were largely seen in species that have more than one layer of bulliform cells. The function of bulliform cells is to reduce transpiration in times of water stress by causing the leaves to roll up (e.g. Balsamo et al., 2006). Variability in their size and number may therefore be related to the environment. During this study, samples were collected from as wide a range of typical habitats for each species as possible, perhaps explaining the degree of variability seen in these characters.

383

Two species had surface hairs, but with high intraspecific variability. Similar high levels of variation in the density of hairs on the leaves of C. hirta even within the same ramet were recorded by Molina et al. (2006). Jermy et al. (2007) noted that when C. hirta grows in damp, shady situations it loses its pubescence. Our study shows that this is also an inconsistent character in other Carex spp. Likewise, high variation was found in the presence of silica bodies, although this may be attributable to sampling issues, as mentioned earlier. This is a character that has been used in previous taxonomic studies (Oh, 1987; Starr & Ford, 2001; Oda & Nagamasu, 2008), but the use of silica bodies is complicated by limited understanding of how environmental factors and development affects their structure (Starr & Ford, 2001). The number of layers of bulliform cells has been used previously to distinguish between C. vulpina and C. otrubae, with C. vulpina described as having at least three layers of bulliform cells, in contrast to C. otrubae with only a single layer (Porley, 1999). These findings were not supported here. Multiple layers were recorded in only 8% of samples of C. vulpina and in 16% of samples of C. otrubae. Although the character was consistent in seven of the 11 species examined, our results suggest caution in using it for taxonomic studies and highlight the necessity of determining intraspecific variation in a species before taxonomic distinctions are made. Within a species, the amount of genotypic differentiation commonly increases with time of isolation, so, in species with wide distributions, a greater degree of genotypic differentiation would be expected among different populations (Gitzendanner & Soltis, 2000; Stenstrom et al., 2002). This has been recorded for leaf anatomy in Aegilops (Kharazian, 2007) and Standley (1989) found two distinct anatomical forms of C. stricta Lam. related to distribution. However, no such relationship was recorded here. The highest levels of quantitative variation were found in C. limosa, a restricted species, and the least variation was seen in C. leporina, a common species found in a broad range of habitats and locations. Possibly the leaf characters are under sufficient selection pressure to offset any geographical differentiation. Intraspecific variation, which occurred to varying degrees in the species and characters in this study, is a product of developmental differences, environmental variation influencing phenotypic plasticity and genotypic differences. By restricting this study to mature leaves, developmental differences should have been minimized, hence any differences are environmental or genotypic. Variable characters certainly exist and, until further work reveals the relationships between leaf anatomy and environmental variables,

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384

384

C. BUGG ET AL.

these will remain of little taxonomic use. Characters that are uniform in a species, of which this study has revealed several, show little influence of the environment and must reflect genetic uniformity. Although the pitfalls of bias, circumscription of characters and small sample size identified by Starr & Ford (2001) are real, they can be overcome to reveal reliable anatomical characters that are potentially useful in taxonomic and phylogenetic contexts.

ACKNOWLEDGEMENTS The authors would like to thank Dr Mary Dean for her contribution of samples to the study and to Tony Hunter for assisting with data collection. The work was funded by an Edge Hill University internal grant plus field work grants for PA from the British Ecological Society and the Systematics Association.

REFERENCES Balsamo RA, Willigen CV, Bauer AM, Farrant J. 2006. Drought tolerance of selected Eragrostis species correlates with leaf tensile properties. Annals of Botany 97: 985–991. Botanical Society of Britain and Ireland. 2011. BSBI Maps Scheme BSBI Available at: http://www.bsbi.org.uk/ maps_scheme.html [accessed April 2012]. Dean M, Ashton PA. 2008. Leaf surfaces as a taxonomic tool: the case of Carex section Phacocystis (Cyperaceae) in the British Isles. Plant Systematics and Evolution 273: 97–105. Gitzendanner M, Soltis PS. 2000. Patterns of genetic variation in rare and widespread plant congeners. American Journal of Botany 87: 783–792. Govaerts R, Simpson DA, Bruhl J, Eforova T, Goetghebeur G, Wilson K. 2007. World checklist of Cyperaceae. Sedges. Kew: Kew Publishing. Jermy AC, Simpson DA, Foley MJY, Porter MS. 2007. Sedges of the British Isles, 3rd edn. London: BSBI. Kharazian N. 2007. The taxonomy and variation of leaf anatomical characters in the genus Aegilops L. (Poaceae) in Iran. Turkish Journal of Botany 31: 1–9. Menapace FJ, Wujek DE, Reznicek AA. 1986. The systematic significance of achene micromorphology in Carex retrorsa (Cyperaceae). Brittonia 39: 278–283. Metcalfe CR. 1969. Anatomy as an aid to classifying the Cyperaceae. American Journal of Botany 56: 782–790. Metcalfe CR. 1971. Anatomy of monocotyledons. Part V. Cyperaceae. Oxford: Oxford University Press. Molina A, Acedo C, Llamas F. 2006. The relationship between water availability and anatomical characters in Carex hirta. Aquatic Botany 85: 257–262. Naczi RFC. 2009. Insights on using morphologic data for phylogenetic analysis in sedges (Cyperaceae). Botanical Review 75: 67–95.

Naczi RFC, Reznicek AA, Ford BA. 1998. Morphological, geographical and ecological differentiation in the Carex willdenowii complex (Cyperaceae). American Journal of Botany 85: 434–447. Namaganda M, Krekling T, Lye KA. 2009. Leaf anatomical characteristics of Ugandan species of Festuca L. (Poaceae). South African Journal of Botany 75: 52–59. Oda J, Nagamasu H. 2008. Two new species of Carex sect. Capitellatae (Cyperaceae) from Japan. Acta Phytotaxonomica et Geobotanica 59: 55–66. Oh YC. 1987. The SEM and LM epidermal characters of the leaf blades in Korean sedge taxa I. Carex. Korean Journal of Plant Taxonomy 17: 235–258. Poland J, Clement E. 2009. The vegetative key to the British Flora. Southampton: John Poland, (in association with the BSBI). Porley RD. 1999. Separation of Carex vulpina L. and Carex otrubae Podp. (Cyperaceae) using transverse leaf sections. Watsonia 22: 431–432. Smith C. 2007. Inter- and intraspecific variation in British members of the genus Carex L. (Cyperaceae), with particular focus on the species Carex filiformis L. Unpublished PhD Thesis University of Lancaster. Smith C, Ashton PA. 2006. Distinction between the sedges Carex vulpina L. and C. otrubae Podp. and the potential for identification of hybrids. Watsonia 26: 15–25. Stace C. 2010. New Flora of the British Isles, 3rd edn. Cambridge: Cambridge University Press. Standley LA. 1987. Anatomical and chromosomal studies of Carex section Phacocystis in eastern North America. Botanical Gazette 148: 507–518. Standley LA. 1989. Taxonomic revision of the Carex stricta complex in eastern North America. Canadian Journal of Botany 67: 1–14. Standley LA. 1990. Anatomical aspects of the taxonomy of sedges (Carex, Cyperaceae). Canadian Journal of Botany 68: 1449–1456. Starr JR, Ford BA. 2001. The taxonomic and phylogenetic utility of vegetative anatomy and fruit epidermal silica bodies in Carex section Phyllostachys (Cyperaceae). Canadian Journal of Botany 79: 362–379. Stenstrom A, Jonsdottir IS, Augner M. 2002. Genetic and environmental effects on morphology in clonal sedges in the Eurasian Arctic. American Journal of Botany 89: 1410– 1421. Szczepanik-Janyszek M, Klimko M. 1999. Application of anatomical methods in the taxonomy of sedges (Carex L.) from the section Muehlenbergianae (L.H. Bailey) Kuk. occurring in Poland. Roczniki Akademii Rolniczejw Poznaniu–CCCXVI 2: 97–107. Toivonen H, Timonen T. 1976. Perigynium and achene epidermis in some species of Carex, subg. Vignea (Cyperaceae), studied by scanning electron microscopy. Annales Botanici Fennici 13: 49–59.

© 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 371–384