AOBPreview originally published online on January 30, 2006
Annals of Botany 2006 97(4):601-610; doi:10.1093/aob/mcl009
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A Comparative Study on the Anatomy and Development of Different Shapes of Domatia in Cinnamomum camphora (Lauraceae)
1 The Nagoya University Museum, Furo-cho, Chikusa-ku, Nagoya, 464-8601, Japan, 2 National Institute for Basic Biology, Okazaki Institute for Integrated Bioscience, Myodaiji-cho, Okazaki, 444-8585, Japan and 3 The Kyoto University Museum, Kyoto, 606-8501, Japan
* For correspondence. E-mail nishida{at}num.nagoya-u.ac.jp
Received: 2 August 2005 Returned for revision: 13 October 2005 Accepted: 6 December 2005 Published electronically: 30 January 2006
 |
ABSTRACT |
|---|
 TOPFOOTNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
• Background and Aims Domatia are small organs usually found in the axils of major veins on the underside of leaves and, although they have received wide attention from ecologists, few detailed reports exist on their anatomy or development. This study is focused on the domatia of Cinnamomum camphora (Lauraceae) and is the first comparative study on the anatomy and development of the different shapes of domatia within a single plant.
• Methods Four types of domatia in C. camphora leaves were observed on paraffin sections under a microscope.
• Key Results The domatia consisted of six histological parts: the upper epidermis, the upper mesophyll tissue, spongy tissue, the lower mesophyll tissue, the tissue filling the rim opening, and the lower epidermis. They differed from the non-domatial lamina mainly in the cell structure of the upper and lower mesophyll tissue and the rim tissue. Differences in domatium shapes were mainly associated with differences in the structure of the upper mesophyll and in the number and size of the rim tissue cells. Differences in the development of domatium types were observed in terms of initiation timing, differentiation of the upper mesophyll cells and degree of rim tissue development.
• Conclusions In domatia, active anticlinal division in the lower mesophyll cells, as compared with the upper mesophyll cells, was coordinated with dynamic growth of rim tissue cells and resulted in cavity formation. The anatomical or developmental differences among the four types of domatia were related to the positions of the domatia within a leaf. In terms of the ecological implications, the major anatomical difference between the domatia used by herbivorous and carnivorous mites was in the development of the rim tissue.
Key words: Anatomy, cavity formation, Cinnamomum camphora, domatia, development, Lauraceae
|
INTRODUCTION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Domatia are small plant organs usually found in the axils of major veins on the underside of leaves, and occur in 28 % of the approx. 290 woody dicotyledonous plant families (O'Dowd and Willson, 1989
). Among ecologists, they have received wide attention following the postulation of their function in the interaction between plants and mites. Their hypothesized function is as a shelter or nursery for carnivorous or fungivorous mites; plants may provide these mites with domatia as refuge from their predators or as cells to hatch their eggs, and in turn may benefit indirectly from reductions in herbivory or pathogens (O'Dowd and Willson, 1989; Pemberton and Turner, 1989). Many ecological studies have supported this hypothesis (e.g. Walter and O'Dowd, 1992; Grostal and O'Dowd, 1994; Agrawal, 1997; Norton et al., 2001), although there have been some contradictory reports of domatia being associated with herbivorous mites (Kasai et al., 2002a, b; Nishida et al., 2005).
Compared with the attention given in the field of ecology, anatomical or developmental studies on domatia have been relatively scarce. Domatia show a wide variation in morphology (Hamilton, 1897; Jacobs, 1966; O'Dowd and Willson, 1989), and they are usually categorized into four basic types: the pit, pouch, pocket and tuft types (O'Dowd and Willson, 1989). However, domatia with intermediate or mixed characters have also been reported (Nishida, 2004). There have been some studies that refer to the anatomy or development of these diversified domatia, but most have only roughly described their structure or growth (Hamilton, 1897; Tô Ngoc Anh, 1966; Schnell et al., 1968; Wilkinson, 1979; O'Dowd and Willson, 1989; Martinez-Solis et al., 1993; Cerana and Ariza Espinar, 1995). Hamilton (1897) and O'Dowd and Willson (1989) provided reviews on domatia and observed the anatomical characters of these organs in 11 and 27 plant species, respectively, but their main focus was ecological. Tô Ngoc Anh (1966) and Schnell et al. (1968) compared domatium development among eight species of Rubiaceae and 14 species from various families, respectively, and both concluded that cell differentiation was delayed in domatia; however, neither study detailed the differences among the plant species. Wilkinson (1979) reviewed previous reports and noted the lack of anatomical studies. Martinez-Solis et al. (1993) and Cerana and Ariza Espinar (1995) described different domatia in Rhamnus (Rhamnaceae) and Mikania (Asteraceae) species, respectively, but did not examine their development. More detailed studies on both the anatomy and development of different domatia are needed to understand the factors involved in the diversity of these organs.
Cinnamomum camphora is an evergreen tree in the Lauraceae family indigenous to China, Taiwan and Japan; it is currently used as timber and was previously harvested for camphor (Mabberley, 1997). This species was recently reported to have different shapes of domatia within a single leaf, which were inhabited by different types of mites (Nishida et al., 2005). Because of these variations, we chose C. camphora to analyse the morphological diversity of domatia.
Domatia of C. camphora have an opening on the lower side and an expansion toward the upper side of the leaf (Fig. 1). According to Nishida et al. (2005), C. camphora domatia can be categorized into four types based on the width and pubescence of the opening: the pouch-type domatia, with a narrow (0·06 ± 0·04 mm) and pubescent opening; the pubescent pit-type domatia, with a wide (0·21 ± 0·08 mm) and pubescent opening; the glabrous pit-type domatia, with a wide (0·20 ± 0·07 mm) and glabrous opening; and the dish-type domatia, with a very wide (0·26 ± 0·09 mm) and glabrous opening. The type of domatium is roughly correlated with the position on the leaf where it occurs. This relationship between the types and positions allowed us to track and compare the development of the different shapes of domatia. In addition, a comparative study on the different shapes of domatia within a single plant excluded any between-plant differences and enabled us to determine the genuine differences among the shapes.
|
In this study, we describe the anatomy and development of the four domatium types in C. camphora. Based on the data, we present the differences in morphology and development among the types of domatia, and discuss developmental factors that could influence the variation in domatium types.
|
MATERIALS AND METHODS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Cinnamomum camphora (L.) J. Presl was collected on the campus of Nagoya University in west-central Japan from March 18 to June 14, 2002. We randomly selected terminal buds or branches of trees that were approx. 8 m tall and had already reached the regenerative stage, and used leaves in the middle position inside the buds or on the branches. We avoided using the terminal or lowest leaf on the branch, because it sometimes had abnormal domatia. For serial microtome sections, the materials were fixed with FAA (formalin alcohol acetic acid) overnight (or longer), dehydrated in an ethanol series and embedded in paraplast. The 8–10 µm thick transverse sections of leaf blades were stained with Mayer's haematoxylin, safranin and fast green (Takaso et al., 1997
).
To describe differences in the morphology of domatia, we compared the number and size of the cells in the opening rims. Samples were collected from three different leaves of an individual tree for each domatium type (six samples for each type, since both the right and left sides of the rim sections were surveyed). Samples were fixed with FAA overnight (or longer), then either sectioned by microtome following the same procedure as described above or placed in 70 % ethanol and sectioned by hand. The number of the rim tissue cells was counted in those areas surrounded by a vertical line from the end of the vascular tissue and a horizontal line from the rim projection (see broken lines in Fig. 2A, B, C and D for the circumscription), and the dimensions of the 10 largest cells in each tissue sample were measured in a transection. All statistical analyses were conducted using JMP version 4·0 (SAS Institute Inc., 2000).
|
|
RESULTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Morphology
Differences in anatomy between domatia and non-domatial lamina were observed mainly in the structure of the upper and lower mesophyll tissues, and by the presence or absence of tissue around the opening and trichomes. Differences among the domatium types were found in the cell type of the upper mesophyll tissue and in the number and size of the tissue around the opening. First, we examined the morphological features of pouch domatia, the most commonly observed type on camphor leaves. We then looked at the pubescent pit, glabrous pit and dish domatia, focusing on the differences from pouch domatia.
For comparison, the non-domatial lamina is shown in Fig. 2E. It contained a single upper epidermal layer, one or two palisade layers (except for compact cells sometimes observed next to the bundle sheath extension), about six spongy tissue layers and one lower epidermal layer.
Pouch-type domatia
Pouch domatia consisted of six histological parts: the upper epidermis, the upper mesophyll tissue, spongy tissue, the lower mesophyll tissue, the tissue filling the rim opening, and the lower epidermis (Fig. 2A, F, G). Because the upper and lower mesophyll tissues and the tissue filling the rim opening were differentiated in a way characteristic of domatia, we call them ‘upper domatial tissue’, ‘lower domatial tissue’ and ‘rim tissue’ for convenience (these terms are used for the other domatium types as well).
The upper epidermal cells were smaller than the upper epidermal cells of non-domatial lamina (compare Fig. 2E and F). The upper domatial tissue consisted of variously sized collenchymatous cells (Fig. 2F), which were seen only in the domatium region or above the major veins. The spongy cells were similar to those of non-domatial lamina, but the intercellular space of the domatium spongy tissue was smaller than in the non-domatial lamina (compare Fig. 2E and F). The lower domatial tissue consisted of two layers of compact rectangular cells (Fig. 2F). The rim tissue contained cells that were similar to those of the lower mesophyllous tissue in the major vein area, i.e. collenchymatous cells (Fig. 2G). The number and size of the rim tissue cells were large in the pouch domatia (Table 1); approx. 124 cells were observed on one side of the rim arm in the transection, and cells varied in size but sometimes (especially in the central part of the tissue) elongated toward the centre of the domatia and reached approx. 50 µm in length. The lower epidermal cells along the cavity were as large as those in the non-domatial lamina, and the cells around the rim were small and continuous with those in the major veins (Fig. 2A; compare Fig. 2E with F and G).
|
Trichomes were abundant around the rim of the domatia, whereas they were not observed in the non-domatial lamina or in the major veins of fully grown leaves (Fig. 2A). Stomata or other special organs (e.g. glands) were absent.
Pit-type domatia
There were two types of pit domatia in camphor leaves: the pubescent pit type (Fig. 2B, H and I) and the glabrous pit type (Fig. 2C and J). The main difference between the pubescent and glabrous pit domatia was the presence or absence of trichomes. The following features, except for the trichomes, were common in the two pit types, but distinct from the pouch or dish domatia.
Like pouch domatia, pit domatia consisted of six parts (Fig. 2B and C). The upper domatial tissue consisted of rectangular, compact parenchymatous cells or, rarely, palisade cells, except for the collenchymatous cells above small veins (Fig. 2H and J). The rim tissue was filled with collenchymatous tissue similar to that in the pouch type, but the number and size were less than those of the pouch type (Fig. 2I and J). There were 27–69 cells on one side of the rim arm in the transection. Cells in the rim tissue were less often elongated than those of pouch domatia, and the large cells were approx. 20 and 28 µm long, on average, in the pubescent type and glabrous type, respectively (Table 1). There were a few trichomes around the rim tissue in the pubescent pit domatia, but the number was low compared with the pouch type (approx. 10 per domatium).
Dish-type domatia
The histology of the dish-type domatia was similar to that of the other types (Fig. 2D), but the structures of the upper domatial tissue and the rim tissue were different (Fig. 2K). The upper domatial tissue consisted of one or two palisade layers. The rim was only slightly developed; there were around 12 cells on one side of the rim arm in the transection, and the cells were not elongated (approx. 15 µm long on average; Table 1). Trichomes were not observed.
Development
Compared with non-domatial lamina, the development of domatia was distinct in the invagination of the lower leaf surface, the differentiation of the upper or lower domatial tissue, the growth of the rim tissue and the timing of trichome growth. Development of the non-domatial lamina and the four types of domatia is summarized in Fig. 3, and diagrams illustrating ontogenetic patterns of non-domatial lamina and domatia are shown in Fig. 4.
|
|
Non-domatial lamina (Fig. 3A–D)
For comparison, development of the non-domatial lamina was as follows. The non-domatial lamina had already formed about six layers (one upper epidermal, about four mesophyll and one lower epidermal layers) before the domatia initiated (Fig. 3A). The lower leaf surface was already pubescent at this stage. By the time the blade grew to approx. 7 mm in length, the uppermost layer of the mesophyll cells had elongated; the lowest layer of cells divided periclinally once and became two layers of short rectangular cells; the middle layers of mesophyll cells remained rounded and divided anticlinally and periclinally (Fig. 3B). By the time the blade was approx. 20 mm long, the uppermost mesophyll cells had become as long as the palisade cells in the fully grown leaves (Fig. 3C). Cells of the middle mesophyll layers continued to divide around that time. Cells of the middle mesophyll layers grew in size when the blade reached approx. 40 mm in length, but differentiation to spongy tissue had not yet occurred (Fig. 3D). Cells of the lower mesophyll layers remained compact and rectangular at that time (indicated by the arrow in Fig. 3D). The differentiation of these cells and the lowest mesophyll cells into spongy tissue started when the blade reached approx. 80 mm in length. The blade was fully grown when it became approx. 90 mm long (Fig. 2E).
Pouch-type domatia (Fig. 3E–H)
Development of the pouch domatia initiated with an invagination on the lower part when the blade was approx. 3 mm long and still in a bud (Fig. 3E). While the mesophyll and epidermal cells of the non-domatial lamina divided evenly in all layers and retained parallel arrangement at this stage (Fig. 3A), those of the domatium region showed a difference in the anticlinal cell division rate: cells in the lower layers of the domatium region divided more actively than cells in the upper or middle layers. This difference was coordinated with increased cell numbers in the lower domatium region near the neighbouring veins, which created the invagination. Trichomes were not observed along the edge of the invagination, whereas non-domatial lamina around the domatia were pubescent.
By the time the blade grew to approx. 7 mm long, but was still in the bud, invagination of the lower leaf surface became conspicuous (Fig. 3F). Cells in the uppermost mesophyll layer elongated longitudinally to some extent. Cells in the lowest layer of the mesophyll divided periclinally once, and became two layers, and were more or less rectangular in shape. The lower mesophyll cells and the lower epidermal cells continued to divide actively anticlinally, in contrast to the upper epidermal cells or the upper and middle mesophyll cells. Cells in the lower part near the neighbouring veins divided actively, which caused the rim tissue to become conspicuous. Trichomes began to grow around the rim.
By the time the leaf blade was 20 mm long, the domatium cavity started to expand, and the rim became thicker (Fig. 3G). The cells in the uppermost mesophyll layer stopped elongating and divided periclinally into 3–5 layers. The rectangular cells in the two lowest mesophyll layers and the lower epidermal cells continued active anticlinal division, which caused the cavity to grow. The rim tissue cells continued to divide in random directions, which caused the rim to protrude. Domatial trichomes increased, while those of the non-domatial lamina started to fall off.
By the time the blade was approx. 40 mm long, the cavity became large as the neighbouring veins grew thick, and this led the upper leaf surface of the domatium region to protrude outward (Fig. 3H). Cells of the upper mesophyll layers became rounded and similar to those above the veins. The lowest two layers of mesophyll cells and the lowest epidermal cells remained as small as before and continued active anticlinal division. The rim tissue cells increased in size, causing the rim to become thicker.
The domatium completed its growth as the blade reached approx. 90 mm in length (Fig. 2A). By this time, the upper domatial tissue cells and rim tissue cells had become collenchymatous (Fig. 2F and G). The middle layers of the mesophyll differentiated to spongy tissue, whereas cells of the two lowest mesophyll layers remained rectangular and compact (Fig. 2F).
Pit-type domatia (Fig. 3I–O)
Pit domatia underwent almost the same developmental process in both the pubescent and glabrous types. The only difference was the absence of trichome growth in the glabrous type.
The early development of pit domatia was similar to that of pouch domatia, but generally lagged behind. Pit domatia began development with an invagination of the lower epidermis, but the timing was later than in pouch domatia, not beginning when the blade had been approx. 3 mm long (Fig. 3I), but beginning only after the blade had grown to approx. 7 mm in length (Fig. 3J and M). While domatia initiated, cells of the upper mesophyll layers elongated (Fig. 3 J and M). Trichomes began to grow almost simultaneously as invagination initiated in the pubescent type (Fig. 3 J). No trichomes grew from the opening rim in the glabrous type, although non-domatial lamina outside of the opening was sparsely pubescent (Fig. 3 M).
When the blade reached approx. 20 mm in length, cells of the upper mesophyll layers remained long and differentiated partly into palisade cells and partly into compact cells (Fig. 3K and N). The other mesophyll cells developed in a similar way to those in pouch domatia, although the timing was always later and the end result was not as pronounced. The rim tissue cells did not develop to become as large, in number or in size, as those in pouch domatia even when the blade was approx. 40 mm long (Fig. 3L and O).
Dish-type domatia (Fig. 3P–R)
Dish domatia started their development much later than the other domatium types. Even the neighbouring veins did not appear when the other domatia initiated. The veins developed when the blade grew to approx. 7 mm long (Fig. 3P), but the domatia themselves started to grow only after the blade became approx. 20 mm long (Fig. 3Q). The rim tissue cells did not grow larger, and the opening rim barely developed (Fig. 3R). Cells of the upper mesophyll layers elongated before the domatium initiation (Fig. 3P) and became palisade cells when development was completed (Fig. 2D and K).
|
DISCUSSION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Differences between domatia and non-domatial lamina
The major differences in the structure of C. camphora domatia in relation to the non-domatial lamina were an invagination in the lower part of the leaf, compact cells in the lower domatial tissue, trichomes at the fully grown stage (except for glabrous pit and dish domatia), collenchymatous cells in the upper domatial tissue (except for pit and dish domatia) and the rim tissue. The first three differences are discussed here, and the last two are discussed in the following section.
Invagination of the domatia was initiated by the difference in the growth rate between the upper and lower part of the lamina, i.e. the lower mesophyll and lower epidermal cells divided more actively, as compared with the upper mesophyll or epidermal cells. This difference in growth rate alone may have resulted in the uneven lamina thickness above the cavity, but the accompanying dynamic growth of the rim tissue stretched the cavity surface and reduced the unevenness. Tô Ngoc Anh (1966) attributed the reduced development in the domatium zone to the peripheral tissues around the zone overarching the depression, which explains the cavity initiation but not the cavity expansion. To our knowledge, our report is the first detailed description of domatial cavity formation.
The compact cells in the lower domatial tissue were the common feature in the four domatium types, and were distinct from non-domatial lamina. Similar compact cells were observed in non-domatial lamina of the young leaves (arrows in Fig. 3D), but they later differentiated into spongy cells (Fig. 2E). In contrast, the lower domatial tissue remained rectangular and compact, even after they had completed their development (Fig. 2F, H, J and K). Tô Ngoc Anh (1966) observed that cells in the pit-type domatia of Coffea were more or less elongated and narrow with a large nucleus, and interpreted this as a conserving of immature characters. Although we did not observe a large nucleus relative to the other tissues, the morphology of the cells in the domatial tissue appears to be common in both C. camphora and Coffea species. Further observations of domatia in other plant taxa are needed to determine whether this feature is universal.
Regarding domatial trichomes, Tô Ngoc Anh (1966) referred to their formation in Coffea and suggested that the trichomes could be explained by a slowing of growth on the leaf surface relative to the level during the early growing stage. We cannot agree with this interpretation, as the initiation of timing in trichomes was quite different between domatial and non-domatial lamina.
Differences among the different domatium types
Although O'Dowd and Willson (1989) roughly classified all C. camphora domatia into the pit type, Nishida et al. (2005) recognized four (pouch, pubescent pit, glabrous pit, and dish) types and distinguished among them by the opening width and pubescence. Furthermore, we found that the four types of C. camphora domatia were different in their anatomy and developmental properties. The two major differences among the four types in terms of anatomy were the cell structure of the upper domatial tissue and the number and size of cells in the rim tissue.
In terms of cell structure, the cells of the upper domatial tissue were collenchymatous in pouch domatia (Fig. 2F), partly palisade and partly compact in (glabrous and pubescent) pit domatia (Fig. 2H and J), and palisade in dish domatia (Fig. 2K). During development of the four domatium types, cells of the upper domatial tissue of all types elongated once anticlinally (Fig. 3F, J, M and P), suggesting that the upper domatial tissue underwent a similar process during this early development. However, the tissue of pouch domatia soon divided periclinally (Fig. 3G) and became collenchymatous (Fig. 3H), whereas the tissue of pit and dish domatia did not divide but became compact or palisade (Fig. 3L, O and R). The collenchymatous cells in pouch domatia were similar to the cells above the veins, and the palisade cells in pit or dish domatia were the same as in the non-domatial lamina. Compact cells were sometimes observed next to the bundle sheath extension.
According to Nishida et al. (2005), different domatium types are seen in different positions on the camphor leaf: pouch domatia tend to occur near the base of the leaf (between the midrib and the basal secondary veins), pit domatia mostly occur near the central position of the leaf (between the midrib and medial secondary veins), and dish domatia tend to occur near the edge of the leaf (between the midrib and distal secondary veins or between secondary and tertiary veins). In other words, pouch domatia tend to occur between more developed veins, whereas pit and dish domatia do so between less developed veins. The similarity in cell structure between pouch domatia and the vein part, and the association between the domatium types and the degree of vein development, suggest that differentiation of the upper domatial tissue may be affected by the degree of vein development. This suggestion is supported by the fact that the compact cells of the pit domatia were similar to the cells near the bundle sheath extensions (CP in Fig. 2E), and by the fact that the cells were collenchymatous above the veins, even in pit domatia (e.g. CL in Fig. 2H). We postulate that diffusive substances from the veins influence the tissue differentiation; more developed veins may secrete more substances and cause cells of the upper domatial tissue to be closer to the collenchymatous cells above the veins. Alternatively, the amount of substance may be the same among the veins, but may be effective only during the early stage of domatium development. Since the domatium formation initiates earlier near the basal region than in the marginal regions (see below for details), the domatia affected by the substances may be limited to the basal region. Elucidation of the substances should clarify these postulates.
The number and size of the rim tissue cells were largest in pouch domatia, smaller in pit domatia and smallest in dish domatia. We assume that these differences in the rim tissue cells may be partly due to differences in development time. The timing of the initiation of development was one of the important differences in the pattern of development among the different domatium types. Earlier initiation of the domatia may provide a longer period of growth for the rim tissue, which could lead to the development of tissue having more and larger cells. This difference in initiation timing also corresponds to differences in the positions of domatia between veins. Vascular development in a dicotyledon leaf begins with the differentiation of procambium in the future midvein, and the larger lateral veins are initiated earlier than the smaller ones (Esau, 1960). Although we are not sure if the time lag of vascular development affects the timing of initiation in the different types of domatia (since vascular bundles were already mature when the domatia initiated), a similar relationship was observed between major vein development and domatium initiation. We also found that development of rim tissue cells was more active in pouch domatia than in pit or dish domatia, perhaps because the pouch domatia occurred near the basal region of the leaf. Although there has been no study on the developmental pattern of C. camphora leaves, the cell division rates are more active in the base than the tip for the ovate leaves of dicots in general, as is known for sunflowers (Granier et al., 2000), tobacco (Poethig and Sussex, 1985) and Arabidopsis (Donnelly et al., 1999), all of which have simple leaves with shapes similar to those of C. camphora. If C. camphora leaves have the same tendency with respect to cell division rates as the leaves of these other plants, the well developed domatia with a large opening rim near the leaf base could be partly explained by this tendency. We intend to investigate this possibility with a more detailed study on the development of C. camphora leaves.
Conclusions and perspectives
The mechanism of C. camphora domatia formation was found to be three-dimensional outgrowth around a specific position between veins. Active anticlinal cell division in the lower mesophyll, as compared with the upper and middle mesophyll, was coordinated with dynamic growth of the rim tissue and resulted in cavity formation. Anatomical differences among the four domatium types were mainly associated with structural differences of the upper domatium tissue and number and size of the rim tissue cells, and developmental differences among the types were observed in terms of initiation timing, differentiation of the upper domatium tissue and degree of rim tissue development. These anatomical or developmental differences may be related to the position of the domatia within a leaf.
A recent study on C. camphora domatia has reported that differently shaped domatia were associated with different mite taxa: pouch domatia were inhabited by herbivorous or fungivorous mites (Eriophyidae and Tarsonemidae), and pit or dish domatia were inhabited or used by carnivorous mites (Stigmaeidae and Phytoseiidae) (Nishida et al., 2005). Taking into consideration the co-evolutionary theories on domatia thus far postulated (e.g. O'Dowd and Willson, 1989; Pemberton and Turner, 1989), it is reasonable to suppose that plants offer domatia to carnivorous mites that could reduce attack from harmful mites, while there are relatively few explanations for the benefit of providing domatia to herbivorous mites. Lundström (1887) hypothesized that plants absorb carbon dioxide from mite respiration and nitrogen-containing materials from their excrement or dead bodies, but this hypothesis has not been supported because no enzyme activity or cytological or structural evidence has ever been found (O'Dowd and Willson, 1989). Our study also does not support this hypothesis positively, because there were few stomata and apparently no special structures inside the domatia to play these roles. However, we cannot reject this hypothesis prior to direct examinations, such as a radioisotope experiment, given that there are known examples of organs absorbing substances without a special structure, such as the cavity walls of the tubers of ant-plants (Huxley, 1978).
In our results, the major anatomical difference between the domatia used by herbivorous and carnivorous mites was the development of rim tissue. The tissue structure of the other parts of the domatia, especially near the surface of the cavity, had no discriminating characteristics corresponding to the inhabitants. The gap in the number of rim tissue cells was much greater between the pouch type (inhabited by herbivores) and the other domatium types than between the pit and dish types (both used by carnivores; Table 1). Our result suggests that further study should focus on the rim tissue to determine how domatia function for herbivorous mites.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
We thank N. Dengler, R. Imaichi and an anonymous reviewer for their help in improving the manuscript, T. Nishida for statistics, and J. Yokoyama for references. This study was supported in part by Grants-in-Aid for Scientific Study from the Japan Society for the Promotion of Science (no. 13740487).
|
FOOTNOTES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Present address: Graduate School of Science, University of Tokyo,7-3-1 Hongo, Tokyo, 113-0033 Japan 
|
LITERATURE CITED |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
-
Agrawal AA. 1997. Do leaf domatia mediate a plant–mite mutualism? An experimental test of the effects on predators and herbivores. Ecological Entomology 22: 371–376.[CrossRef]
Cerana MM, Ariza Espinar L. 1995. Sobre la presencia de domacios en Mikania (Asteraceae). Kurtziana 24: 7–18.
Donnelly PM, Bonetta D, Tsukaya H, Dengler R, Dengler NG. 1999. Cell cycling and cell enlargement in developing leaves of Arabidopsis. Developmental Biology 215: 407–419.[CrossRef][Web of Science][Medline]
Esau K. 1960. Anatomy of seed plants, 2nd edn. New York: John Wiley & Sons.
Granier C, Turc O, Tardieu F. 2000. Co-ordination of cell division and tissue expansion in sunflower, tobacco and pea leaves. Dependence or independence of both processes? Journal of Plant Growth Regulation 19: 45–54.[Medline]
Grostal P, O'Dowd DJ. 1994. Plants, mites and mutualism: leaf domatia and the abundance and reproduction of mites on Viburnum tinus (Caprifoliaceae). Oecologia 97: 308–315.
Hamilton AG. 1897. On domatia in certain Australian and other plants. Proceedings of the Linnean Society of New South Wales 21: 758–792.
Huxley CR. 1978. The ant-plants Myrmecodia and Hydnophytum (Rubiaceae), and the relationships between their morphology, ant occupants, physiology and ecology. New Phytologist 80: 231–268.
Jacobs M. 1966. On domatia—the viewpoints and some facts. Academie van Wetennschappen Amsterdam 69: 275–316.
Kasai A, Yano S, Nishida T, Kadono F, Takafuji A. 2002a. Spatial distribution pattern of domatia and seasonal occurrence of the eriophyid mite in relation to the foliation phenology of Cinnamomum camphora. Japanese Journal of Applied Entomology and Zoology 46: 159–162.[CrossRef]
Kasai A, Yano S, Takafuji A. 2002b. Density of the eriophyid mites inhabiting the domatia of Cinnamomum camphora Linn. affects the density of the predatory mites, Amblyseius sojaensis Ehara (Acari: Phytoseiidae), not inhabiting the domatia. Applied Entomology and Zoology 37: 617–619.[CrossRef]
Lundström AN. 1887. Pflanzenbiologische Studien. II. Dei Anpassungen der Pflanzen an Thiere. Nova Acta Regiae Societatis Scientiarum Upsaliensis 3: 1–87.
Mabberley DJ. 1997. The plant-book, 2nd edn. Cambridge: Cambridge University Press.
Martinez-Solis I, Iranzo J, Estrelle E, Ibras M. 1993. Leaf domatia in the section Alaternus (Miller) DC. of the genus Rhamnus (Rhamunaceae). Botanical Journal of the Linnean Society 112: 311–318.[CrossRef]
Nishida S. 2004. Morphology and function of domatia. Bunrui 4: 137–150.
Nishida S, Naiki A, Nishida T. 2005. Morphological variation in leaf domatia enables coexistence of antagonistic mites in Cinnamomum camphora. Canadian Journal of Botany 83: 1–9.
Norton AP, English-Loeb G, Belden E. 2001. Host plant manipulation of natural enemies: leaf domatia protect beneficial mites from insect predators. Oecologia 126: 535–542.[CrossRef]
O'Dowd DJ, Willson MF. 1989. Leaf domatia and mites on Australasian plants: ecological and evolutionary implications. Biological Journal of the Linnean Society 37: 191–236.
Pemberton RW, Turner CE. 1989. Occurrence of predatory and fungivorous mites in leaf domatia. American Journal of Botany 76: 105–112.[CrossRef]
Poethig RS, Sussex IM. 1985. The cellular parameters of leaf development in tobacco Nicotiana tabacum: a clonal analysis. Planta 165: 170–184.[CrossRef]
SAS Institute Inc. 2000. JMP, Version 4. North Carolina: SASInstitute Inc.
Schnell R, Cusset G, Tchinaye V, Tô Ngoc Anh. 1968. Contribution à l'étude des ‘acarodomaties’ la question des aisselles de nervures. Revue Générale de Botanique 75: 5–64.
Takaso T, Tobe H, Aida M. 1997. Procedure and observation techniques of sections for optical microscope. In: Fukuda H, Nishimura M, Nakamura K, eds. Experimental protocol for observation of plant cells. Tokyo: Shujunsha, 20–33.
Tô Ngoc Anh. 1966. Sur la structure anatomique et l'ontogénèse des acarodomaties et les interprétations morphologiques qui paraissent s'en dégager. Adansonia 6: 147–151.
Walter DE, O'Dowd DJ. 1992. Leaf morphology and predators: effect of leaf domatia on the abundance of predatory mites (Acari: Phytoseiidae). Environmental Entomology 21: 478–484.
Wilkinson HP. 1979. The plant surface (mainly leaf), part IV domatia. In: Metcalfe CR, Chalk L, eds. Anatomy of the dicotyledons, 2nd edn, Vol. 1. Oxford: Clarendon Press, 132–140.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Leroy, A. Jauneau, A. Quilichini, A. Dejean, and J. Orivel Comparison between the Anatomical and Morphological Structure of Leaf Blades and Foliar Domatia in the Ant-plant Hirtella physophora (Chrysobalanaceae) Ann. Bot., March 1, 2008; 101(4): 501 - 507. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||