Cell Tissue Res (2001) 304:439–454 DOI 10.1007/s004410100377
REGULAR ARTICLE
Holly S. Cate · Charles D. Derby
Morphology and distribution of setae on the antennules of the Caribbean spiny lobster Panulirus argus reveal new types of bimodal chemo-mechanosensilla Received: 27 November 2000 / Accepted: 9 February 2001 / Published online: 19 April 2001 © Springer-Verlag 2001
Abstract This study describes the morphology and distribution of setae on the lateral and medial flagella of the antennules of the spiny lobster Panulirus argus in an effort to identify antennular chemoreceptors in addition to the well-studied aesthetasc chemosensilla. Setae were examined using light and electron microscopy, and their distribution on flagellar annuli was analyzed. We identified ten setal types based on external morphology: hooded, plumose, short setuled, long simple, medium simple, short simple, aesthetasc, guard, companion, and asymmetric setae, with the last four types being unique to the “tuft” located on the distal half of the lateral flagellum. The three setal types whose ultrastructure was examined – hooded, long simple, and medium simple setae – had characteristics of bimodal (chemo-mechanoreceptive) sensilla. The antennules have four distinct annular types based on their setal complement, as shown by cluster analysis. This basic distribution of non-tuft setal types is similar for both lateral and medial flagella. Annuli in the tuft region have tuft setal types superimposed on a basic organization of non-tuft setal types. These results show that the antennules possess a diverse set of setae, that these setae have a highly ordered arrangement on the antennules, that at least four (and probably many more) of these setal types are chemosensilla, and suggest that most antennular chemosensilla are bimodally sensitive. Keywords Sensilla · Chemoreceptor · Mechanoreceptor · Olfaction · Panulirus argus (Crustacea, Decapoda)
This work was supported by the National Science Foundation Grant IBN 0077474, National Institutes on Deafness and Other Communication Disorders Grant DC00312, the Georgia Research Alliance, and the Georgia State University Research Program Enhancement Fund H.S. Cate (✉) · C.D. Derby Department of Biology and Center for Behavioral Neuroscience, Georgia State University, PO Box 4010, Atlanta, GA 30302-4010, USA e-mail:
[email protected] Tel.: +1-404-6511646, Fax: +1-404-6512509
Introduction Crustaceans detect chemicals through cuticular sensilla, which are distributed over the entire body and are particularly abundant on appendages (Laverack 1988; Derby 1989). The antennules have a large complement of chemosensilla and are highly specialized for olfaction (Reeder and Ache 1980; Devine and Atema 1982). To date, most studies on olfaction in crustaceans have focused on aesthetasc sensilla (e.g., Carr et al. 1990; Hallberg et al. 1992; Fadool and Ache 1994; Ache and Zhainazarov 1995; Olson and Derby 1995; Gleeson et al. 1996; Derby et al. 1997; McClintock et al. 1997; Xu et al. 1998). However, several lines of evidence suggest that other (non-aesthetasc) chemosensilla on the antennules play an important role in the detection of waterborne chemicals. In this light, the aim of the present study was to identify candidate non-aesthetasc chemosensilla on the antennules of the spiny lobster, Panulirus argus. Numerous types of chemosensilla have been described in decapod crustaceans. Many are bimodal, being innervated by both mechano- and chemoreceptive cells. Examples include hair peg sensilla (Schmidt 1989) and funnel canal organs (Gnatzy et al. 1984; Schmidt and Gnatzy 1984) on the dactyls, and hedgehog sensilla (also called fringed or squat setae) (Shelton and Laverack 1970; Derby 1982; Altner et al. 1983) on the chelae. To date, aesthetasc sensilla, which are unimodal chemosensilla, are the only described chemosensilla on the antennules of decapods (ultrastructure: Spencer and Linberg 1986; Grünert and Ache 1988; physiology: Spencer 1986; Michel et al. 1991; Ache and Zhainazarov 1995). Aesthetasc sensilla are innervated by olfactory receptor neurons that project to the olfactory lobe (Sandeman and Luff 1973). There is evidence for non-aesthetasc chemosensilla on the antennules of spiny lobsters, despite the fact that these sensilla have not yet been identified. The antennules are bifurcated appendages with lateral and medial flagella. Aesthetascs are located on annuli (segments) in
440
the distal portion of each lateral flagellum (Laverack 1964; Gleeson et al. 1993) in an area referred to as the aesthetasc tuft (Laverack 1964). Fuzessery (1978) provided the first direct evidence for non-aesthetasc chemosensilla by recording action potentials in the antennular nerve following odor-stimulation of the medial flagellum. More recently, studies on the lobster’s capacity to perform odor-mediated tasks following selective sensillar ablation have revealed that non-aesthetasc antennular chemosensilla are sufficient for many odor-mediated behaviors (Horner et al. 2000; Steullet et al. 2000b; Derby et al. 2001). The aim of this study was to identify and describe the distribution of candidate non-aesthetasc antennular chemosensilla. We examined the morphology, abundance, and distribution of setae on lateral and medial flagella of the Caribbean spiny lobster Panulirus argus using light and electron microscopy. Ten setal types were identified based on their external morphology. Three of these types have morphological characteristics consistent with bimodal (chemo-mechanoreceptive) sensilla and are thus candidate non-aesthetasc chemosensilla. These setae are distributed over the entire antennule. Furthermore, based on a quantitative analysis of the distribution of all setal types, we distinguished four classes of antennular annuli. We discuss possible functions for non-aesthetasc antennular chemosensilla and patterns of their distribution.
Materials and methods Animals All experiments were performed on young adult Caribbean spiny lobsters (Panulirus argus) of 60–80 mm carapace length. Animals were collected in the Florida Keys and held at Georgia State University. Lobsters were housed in 800-l aquaria at 20–25°C containing recirculating, filtered artificial seawater (Instant Ocean) and fed shrimp and squid 3 times/week. Scanning electron microscopy Lateral and medial flagella of the first antennae were each cut into six pieces. Tissue was fixed overnight in 70% ethanol, dehydrated in a graded alcohol series, transferred to acetone, and then submersed in dihydroxypropane for final drying. Stubs were coated with Norland optical adhesive 81 (Norland Products Inc.) that was hardened by UV light until it was tacky. Tissue was mounted onto this partially cured glue so that the tissue would not charge but could be removed and turned to observe all sides of the tissue. Stubs were sputter-coated with gold/palladium (Desk II sputter coater, Denton) and examined using a Leica S420 scanning electron microscope at 5–7.5 kV. Transmission electron microscopy Lateral and medial flagella of the first antennae were removed from lobsters 4–12 h postmolt. Tissue was cut into small pieces, fixed overnight in 2.5% glutaraldehyde/1% paraformaldehyde in 0.2 M phosphate buffer (pH 7.4), postfixed for 2 h in 1% osmium tetroxide in 0.2 M phosphate buffer, dehydrated in a graded ethanol series, transferred to acetone, and infiltrated with Eponate 12 resin (Pelco). Semithin (0.5–1 µm) and thin (i.e., gold and silver)
sections were taken using an ultramicrotome (Reichert) and a diamond knife (Microstar). Semithin sections were collected on 12-well slides (Erie Scientific), stained with toluidine blue, and analyzed using a compound microscope (Zeiss Axioskop). Thin sections were collected on uncoated nickel-mesh grids, stained with uranyl acetate and lead citrate, and viewed on a JEOL JEM100 CX II transmission electron microscope operating at an accelerating voltage of 80 kV. Image analysis Scanning electron micrographs covering the entire surface area of each flagellum were digitally captured at either 150 or 200 times magnification and montaged to reconstruct the flagella using Adobe Illustrator software (Adobe Systems Inc.). Numbers and types of setae on each annulus of the flagella were counted from these micrographs; these and other quantitative data are expressed as means ± standard error of the mean. When a seta could not be clearly identified by the image, the preparation was reexamined using scanning electron microscopy under higher magnification. Data analysis We used a hierarchical cluster analysis (Bieber and Smith 1986; Everitt 1993) to determine whether annuli fell into distinct groups based upon their complement of setal types and, if so, whether the groups were correlated with a qualitative classification of the annuli into four annulus “types.” First, we counted the number of each of the different setal types located on each annulus of the lateral and medial flagella. Second, we qualitatively classified these annuli into four main “types” (hooded, simple, hooded-tuft, and simple-tuft) based upon their complement of setal types. Third, we compared our qualitative classification with the results of a hierarchical cluster analysis. The cluster analysis (Statistica Software, StatSoft) was derived from a distance matrix of Pearson correlation coefficient values between any two annuli, as determined by the number of each of the ten setal types on each annulus. Use of the Pearson correlation coefficient as the distance metric emphasized the relative rather than the absolute similarity in the complements of setae, thus placing less emphasis on the differences in total numbers of setae and focusing on the relative numbers of setal types. Ward’s method of linkage was used as the clustering algorithm in the cluster analysis. A hierarchical cluster analysis was also used in classification of non-tuft simple setae. We used this analysis to determine whether simple setae fell into distinct groups based upon their length and base width measurements and, if so, whether the groups were correlated with our qualitative classification of simple setae into three size classes. In this case, the cluster analysis was derived from a distance matrix of euclidean distances (to compare absolute, rather than relative, measures). Ward’s method of linkage was used as the clustering algorithm in the cluster analysis. Terminology We use the term “seta” (plural = setae) for an articulated cuticular outgrowth that has not been shown to be innervated by sensory cells, “sensillum” (plural = sensilla) for an articulated cuticular outgrowth that has been shown to be innervated, and “setal type” to refer to the grouping of setae or sensilla based upon morphological or functional characteristics, or both. We use the terminology of Watling (1989) to describe the basic morphological features of setal types. In addition, we use descriptive names that are established in the literature even though they are not part of Watling’s nomenclature; many of these are “simple” setae, such as aesthetasc, guard, companion, and asymmetric setae.
441
Results Morphology of antennular flagella The first antennae, or antennules, of the lobster are composed of four segments. The most distal segment bifurcates into a pair of flagella (Fig. 1). The lateral flagellum has two distinct regions: (1) a “tuft,” which is in the distal half of the flagellum and is distinguished by a much higher density of setae; and (2) a proximal, non-tuft region (Fig. 1). The medial flagellum does not have a tuft region. The entire lengths of both lateral and medial flagella are composed of smaller segments, called annuli. The
Fig. 1 Anterior aspect of the spiny lobster Panulirus argus. The paired first antennae, or antennules, have four segments, the most distal of which bifurcates into two flagella (lateral and medial). The lateral and medial flagella are composed of smaller segments, called annuli. Setae are located along the entire lengths of the lateral and medial flagella and are particularly dense in the “tuft” on the distal region of the lateral flagellum. (Drawing modified from Grünert and Ache 1988)
Fig. 2 Diameters and lengths of annuli of the lateral and medial flagella. These measures were taken from 36 annuli (indicated on the abscissa axis) equally spaced along the length of each flagellum. Drawings of the lateral and medial flagella are shown for orientation to the relative regions from which measures were taken. The medial flagellum used for these data is actually 13 mm longer and contains 16 more annuli than the lateral flagellum. Lines (A, B, and C) through the flagella indicate the locations of sections shown in Fig. 3
length of the flagellum and number of annuli in a flagellum depend on the size of the animal (Harrison et al. 2001b), but medial flagella are typically longer, more narrow, and have more annuli than lateral flagella. For example, the antennule that is featured in this paper is from a lobster of 70 mm carapace length that had a lateral flagellum 54 mm long with 153 annuli and a medial flagellum 67 mm long with 169 annuli. A total of ten antennules, consisting of six male and four female lateral and medial flagella, were examined for this study. Right and left antennules and male and female antennules had no qualitative differences in setal types or distributions; therefore, quantitative analysis of all setae on the antennular flagella was performed on only one antennule. This representative antennule was used for the data shown in Figs. 2, 5, 6, 7, 8, and 10. The diameters and lengths of individual annuli vary along the axis of the flagella (Fig. 2). In general, the diameters and lengths of the lateral flagellar annuli and the diameters of the medial flagellar annuli decrease from the proximal to the distal end. There is little change, however, in the lengths of lateral flagellar annuli in the tuft region and in the lengths of the medial flagellar annuli along the entire flagellar axis. The shape of the annuli is different in the tuft region and the non-tuft region. In the non-tuft regions of the medial and lateral flagella, the annuli are circular in cross-section (Fig. 3A, B). In the tuft region of the lateral flagellum, the annuli are flattened on the ventral surface where aesthetascs and other tuft setae (see below) are present (Fig. 3C).
442
Fig. 3A–C Light micrographs of 0.5-µm-thick transverse sections through the tuft and non-tuft regions of the antennular flagella showing the shape of the circumference of the annuli. Positions on the flagella for the sections are shown in Fig. 2. A Section from the medial flagellum. B Section from the non-tuft region of the lateral flagellum, three annuli proximal to the tuft region. C Sec-
Table 1 Morphometrics and abundance of antennular setal types of P. argus
Setal type
Lengtha (µm) (n=25b)
Diameter of basea (µm) (n=25b)
Total no. on both flagella
54±1 98±5 42±3 402±26 140±9 59±4
14±1 10±1 24±0 24±1 13±1 9±1
1167 107 377 30 302 293
37% 20% 33% 30% 35% 34%
595±31 1228±56 231±18 223±40
22±1 51±2 17±1 12±1
1268 157 196 78c
100% 100% 100% 100%
Non-tuft 1. Hooded 2. Plumose 3. Short setuled 4. Long simple 5. Medium simple 6. Short simple a Values are means ± SEM b n=3 for asymmetric setae c Counts of asymmetric setae
are based on one seta per annulus in the aesthetasc tuft (Gleeson et al. 1993; Steullet et al. 2000a)
Tuft 7. Aesthetasc 8. Guard 9. Companion 10. Asymmetric
tion from the tuft region of the lateral flagellum. The inner and outer dendrites (arrows) and cell bodies (arrowheads) of olfactory receptor neurons innervating the aesthetasc sensilla are seen just dorsal to the ventral surface of the flagellum (dashed lines in B and C highlight the outer edge of the cuticle) (d dorsal; v ventral)
Identification and distribution of antennular setal types The morphology, abundance, and distribution of the setal types on the lateral and medial flagella were examined. Based on these features, ten setal types were identified: (1) hooded, (2) plumose, (3) short setuled, (4) long simple, (5) medium simple, (6) short simple, (7) aesthetasc, (8) guard, (9) companion, and (10) asymmetric (Fig. 4). A summary of morphological characteristics and absolute abundance of each setal type is shown in Table 1. Six setal types (Table 1 : 1–6) are present on both lateral and medial flagella, being distributed around two-thirds of the circumference of tuft annuli and around the entire circumference of non-tuft annuli of lateral and medial flagella. The other four setal types (Table 1 : 7–10) are only present in the distal portion of lateral flagella, and together they define the tuft (Fig. 1). The relative abundance and distributions of the individual setal types are shown in Figs. 5 and 6, respectively. The aesthetasc sensilla and the hooded setae are the two most abundant se-
% of total on lateral flagella
tal types on the antennular flagella and together they make up 61% of the setae on the flagella (Fig. 5). Hooded setae are the most widely distributed setal type on both medial and lateral flagella (Fig. 6). Detailed descriptions of the distribution and morphology of each setal type are given later in “Results.” One of the aims of this study was to test the hypothesis that the main difference between the flagella is the presence of the tuft setal types on lateral flagella. To address this aim, we examined whether there are differences in the density of setae along the lengths of flagella and between flagella. We found that the distribution and density of the six non-tuft setal types are similar in tuft and non-tuft annuli. The main difference in these two regions, tuft and non-tuft, is that tuft setal types occupy one-third of the circumference of tuft annuli. Non-tuft setal types are present in similar distributions and densities on the lateral and dorsal surfaces of tuft annuli. In fact, when the surface area occupied by tuft setae and the numbers of tuft setae are subtracted from tuft annuli
443 Fig. 4A–H Scanning electron micrographs of the ten setal types located on the antennular flagella of P. argus. Six setal types (A–F) are located on annuli of both the lateral and medial flagella, and four setal types (G, H) are located on annuli of the tuft region of the lateral flagellum. A Hooded seta; B plumose seta; C short setuled seta; D long simple seta; E medium simple seta; F short simple seta; G aesthetasc (a), guard (gs), and asymmetric setae (as) in the medial tuft region of the lateral flagellum; H aesthetasc (a), guard (gs), asymmetric (as), and companion setae (cs) in the proximal tuft region of the lateral flagellum
(Fig. 7, dashed line), the number of setae per unit surface area is similar in the distal half of both antennular flagella (Fig. 7). Different distributions of setal types create four types of annuli
Fig. 5 Relative abundance of setal types on the lateral and medial flagella of the antennule of P. argus
Setae are located on every annulus of both the lateral and medial flagella; however, the types and patterns of distribution of the setae differ across the annuli, creating four “types” of annuli. We first determined the four annular “types” qualitatively by comparing the number of each setal type on each annulus of lateral and medial flagella (see “Materials and methods”). Based on these
444 Fig. 6A, B Distribution of the ten setal types on the antennular flagella of P. argus. The solid bars indicate the distribution of the respective setal types on annuli along the length of both the medial (A) and lateral (B) flagella
Fig. 7 The density of setae on annuli of the lateral and medial flagella. These measures were taken from 36 annuli (indicated on the abscissa axis) equally spaced along the length of each flagellum, as described for Fig. 2
numbers, the distributions of the setal types create four main “types” of annuli: “hooded annuli,” “simple annuli,” “hooded-tuft annuli,” and “simple-tuft annuli” (Fig. 8). Annuli of the distal tips of lateral and medial flagella are not included within these four types. These “distal tip annuli” and their setae are in a terminal developmental stage (Steullet et al. 2000a; Harrison et al. 2001b); on adult lobsters, these annuli typically break off after the animal molts (see Fig. 5 of Steullet et al. 2000a). Aside from distal tip annuli, it is rare to find annuli that do not fit within one of the four annular “types.” However, intermediate types that have similar numbers of hooded and simple setae are present in low numbers (e.g., 18 out of 169 annuli in the flagellum in Fig. 10A); we call these “intermediate annuli.” Additionally, there are usually one or two non-tuft annuli per flagellum that do not have any setae and one or two tuft annuli that do not have setae other than tuft setae; we call these “bare annuli.” Hooded annuli and simple annuli are restricted to medial flagella and to the region of lateral flagella proximal
to the tuft region. Setae project from the distal end of each annulus around the entire circumference. A circumferential ring of hooded setae distinguishes hooded annuli (Fig. 8A, B, D), but these annuli also can have plumose, short-setuled, short simple, medium simple, or long simple setae. Simple annuli, on the other hand, are characterized by an encirclement of short and medium simple setae (Fig. 8A, B, D). Simple annuli also can have plumose, short setuled, or hooded setae and usually have one long simple seta. Hooded annuli are much more numerous than simple annuli. Simple annuli are typically separated by 3–15 hooded annuli, on both lateral and medial flagella. Simple annuli and hooded annuli have similar numbers of setae, but the relative amounts of simple and hooded setae differ between annular types. On hooded annuli, greater than 40% of the setae are hooded setae. These annuli have 5.4±0.2 (mean ± SEM) hooded setae and 0.5 ± 0.1 simple setae per annulus (n=158). In contrast, on simple annuli, greater than 40% of the setae are medium and short simple setae. These annuli have
445
Fig. 8A–D Scanning electron micrographs of the antennular flagella showing the four types of annuli: hooded-tuft annuli, simpletuft annuli, hooded annuli, and simple annuli. A The proximal region of the lateral flagellum. This region is composed of simple (s) and hooded (h) annuli. Medium simple setae (ms) are the most numerous setal type on simple annuli. Hooded setae (hs) are the main setal type on hooded annuli. B The proximal region of the medial flagellum. This region is composed of hooded (h) and simple (s) annuli. Medium simple setae (ms) are the most numerous setal type on simple annuli. A single long simple seta (ls) is often present on simple annuli. Hooded setae (hs) are the main setal type found on hooded annuli. C The distal (tuft) region of the lateral flagellum. This region is composed of simple-tuft (s-t) and hooded-tuft (h-t) annuli. Medium simple setae (ms) are the most numerous non-tuft setal type on simple-tuft annuli. Hooded setae (hs) are the main non-tuft setal type on hooded-tuft annuli. A single long simple seta (ls) is often present near companion setae on the lateral face of simple-tuft annuli. D The distal region of the medial flagellum. This region is composed of hooded (h) and simple (s) annuli. Medium simple setae (ms) are the most numerous setal type on simple annuli
5.4±1.4 simple and 1.3±1.2 hooded setae per annulus (n=46). Simple-tuft annuli and hooded-tuft annuli are found only in the distal region of lateral flagella, in the “tuft” (Fig. 8C). Both of these annular types have aesthetasc, guard, companion, and asymmetric setae on their ventral face, organized in a characteristic pattern (Laverack 1964; Gleeson et al. 1993). In addition, these two types of tuft annuli can also have hooded, plumose, short setuled, and simple setae, which are located on the dorsal, lateral, and mesial faces of these annuli. In rare cases, some hooded, plumose, short setuled, or simple setae are
located within the tuft; examples of a hooded seta and a plumose seta within the tuft at the usual site of companion setae are shown in Fig. 9. Simple-tuft annuli and hooded-tuft annuli are distinguished from each other by the distributions of non-tuft setae on the annuli (Fig. 8C). On hooded-tuft annuli, greater than 70% of the non-tuft setae are hooded setae. These annuli have 2.9±0.9 hooded and no simple setae per annulus (n=34). In contrast, simple-tuft annuli are distinguished by having greater than 55% simple setae. These annuli have 4.1±0.6 simple setae and 0.4±0.2 hooded setae per annulus (n=8). Hooded-tuft annuli are more numerous than simple-tuft annuli. Simple-tuft annuli are typically separated by 3–15 hoodedtuft annuli. We performed a cluster analysis to determine whether these annular “types” are indeed distinct types as defined by statistical analysis (Fig. 10). The numbers of each of the ten setal types present on each annulus of a lateral and a medial flagellum were used to produce a distance matrix for the cluster analysis (see “Materials and methods”). Cluster analysis then compared the relative similarity of annuli based upon their complement of setal types and linked annuli into clusters. The shortest linkage distance between clusters identifies the most similar annuli; the longer distances indicate a greater degree of dissimilarity between annuli. After the cluster analysis was completed, we compared our classification of each annulus, based upon its location with respect to the tuft region and the minimum percentage criteria listed above, with the cluster analysis.
446
Fig. 9A, B Scanning electron micrographs showing examples of non-tuft setae that are occasionally located within the tuft region of the lateral flagellum. A A hooded seta (hs) is located in the usual site of a companion seta (cs) on the mesial side of the flagellum. B A plumose seta (ps) is located in the usual site of a companion seta (cs) on the lateral side of the flagellum. Also shown in these micrographs are guard setae (gs) and a long simple seta (ls)
This cluster analysis was performed separately for the medial flagellum (Fig. 10A), the lateral flagellum (Fig. 10B), and the lateral flagellum with the tuft setal types removed from the data set (Fig. 10C). We classified each annulus as a hooded annulus, simple annulus, hooded-tuft annulus, simple-tuft annulus, distal tip annulus, bare annulus, or intermediate annulus. All annuli were included in each analysis. For the medial flagellum, the two main clusters assigned by cluster analysis correlated extremely well with the two types of annuli identified by our subjective classification scheme, which are hooded annuli and simple annuli (Fig. 10A). For the lateral flagellum, the two main clusters assigned by cluster analysis correlated extremely well with the types of annuli in the tuft region and annuli of the non-tuft region identified by our subjective classification scheme (Fig. 10B). The hooded-tuft and simple-tuft annuli were grouped with the tuft annuli cluster of the cluster analysis, and the simple and hooded annuli were grouped with the non-tuft cluster (Fig. 10B). Once the tuft setal types were removed from the analysis of the lateral flagellum, the two main clusters as assigned by cluster analysis cor-
Fig. 10A–C Cluster analysis of annuli of lateral and medial flagella. This analysis is based on the numbers of the setal types on the annuli. Shortest linkage distances represent the most similar annuli. Longer distances indicate a greater degree of dissimilarity. A Cluster analysis of annuli of a medial flagellum. This analysis is based on the numbers of each setal type on each of the 169 annuli on the medial flagellum, which are represented on the abscissa. The locations of annular types identified by our subjective classification scheme are indicated at the top of each main cluster and below the abscissa. B Cluster analysis of annuli of a lateral flagellum. This analysis is based on the numbers of each setal type on each of the 153 annuli on the lateral flagellum, which are represented on the abscissa. The locations of annular types identified by our subjective classification scheme are indicated at the top of each main cluster and below the abscissa. C Cluster analysis of annuli of the lateral flagellum in B with tuft setae removed from the analysis. This analysis is based on the numbers of only the non-tuft setal types on each of the 153 annuli on the lateral flagellum, which are represented on the abscissa. The locations of annular types identified by our subjective classification scheme are indicated at the top of each main cluster and below the abscissa
447 Table 2 Innervation of four setal types on the antennular flagella of P. argus (N/A not applicable)
Number of inner dendrites Scolopale present Type I dendrite (mechanoreceptive)a Ciliary rootlet in scolopale Ciliary segments long (4–6 µm) Ciliary region begins more distal Type II dendrite (chemoreceptive)a Ciliary rootlet not in scolopale Ciliary segments short (
