Motion-driven flow in an unusual piscine nasal region

Highlights

• Olfaction in the garpike, Belone belone, is likely assisted by swimming.

• Jaws and mouth may also assist in odorant capture.

• Nasal regions show mild to strong asymmetry.

• Variation in nasal morphology influences pattern of olfactory flow.

• Different vortex-like structures were observed, including a suspected horseshoe vortex.

Abstract

Fishes have several means of moving water to effect odorant transport to their olfactory epithelium (‘olfactory flow’). Here we show that olfactory flow in the adult garpike Belone belone (Belonidae, Teleostei), a fish with an unusual nasal region, can be generated by its motion relative to water (swimming, or an external current, or both). We also show how the unusual features of the garpike’s nasal region influence olfactory flow. These features comprise a triangular nasal cavity in which the olfactory epithelium is exposed to the external environment, a papilla situated within the nasal cavity, and an elongated ventral apex. To perform our investigation we first generated life-like plastic models of garpike heads from X-ray scans of preserved specimens. We then suspended these models in a flume and flowed water over them to simulate swimming. By directing filaments of dye at the static models, we were able to visualise flow in the nasal regions at physiologically relevant Reynolds numbers (700–2,000). We found that flow of water over the heads did cause circulation in the nasal cavity. Vortices may assist in this circulation. The pattern of olfactory flow was influenced by morphological variations and the asymmetry of the nasal region. The unusual features of the nasal region may improve odorant sampling in the garpike, by dispersing flow over the olfactory epithelium and by creating favourable conditions for odorant transport (e.g. steep velocity gradients). Unexpectedly, we found that the mouth and the base of the garpike’s jaws may assist the sampling process. Thus, despite its apparent simplicity, the garpike’s nasal region is likely to act as an effective trap for odorant molecules.

Introduction

Perception of smell in water requires transport of odorant molecules from the animal’s external environment to its olfactory epithelium. Because diffusion times in water are extremely long (Denny, 1993), aquatic animals must in the first instance rely on the bulk movement of water for odorant transport (Cox, 2008). The final stage of this process, however, involves diffusion alone (Vogel, 1994).

In fishes, four principal mechanisms may be used to generate bulk movement of water in the nasal region (‘olfactory flow’): (i) the beating of non-sensory cilia; (ii) the expansion and contraction of accessory sacs; (iii) respiration; and (iv) the movement of the fish relative to water (Cox, 2008). Mechanical agitation of the nasal region may also contribute to olfactory flow, but in a non-directional manner (Cox, 2008). Which mechanism(s) operate(s) depends on the species of fish.

Our aim here was to investigate olfactory flow in the adult garpike, Belone belone (Belonidae, Beloniformes, Teleostei). The garpike is a long, slender, marine fish with beak-like jaws (Fig. 1, Fig. 2) (Wheeler, 1969). Its elongated body, together with its prominent tail and the extreme posterior positioning of its dorsal and anal fins (Fig. 1), suggests that it is an active swimmer (Helfman et al., 2009). Indeed, according to Theisen et al. (1980, p. 169), adult garpike are known to swim ‘continuously and normally rather fast’.

The garpike has an unusual nasal region (Fig. 3). The olfactory epithelium of most non-beloniform fishes is enclosed in a nasal chamber, and is typically located on one or more thin folds, or lamellae (Hara, 1975, Zeiske et al., 1992, Cox, 2008). But in the garpike the olfactory epithelium is largely exposed to the external environment: it lines a triangular nasal cavity (Fig. 3B) and the surface of a papilla situated within the nasal cavity (Fig. 3B) (Blaue, 1884, Burne, 1909, Theisen et al., 1980, Cox, 2008). The nasal cavity widens caudally and its edges are sharp. There is a recess beneath the dorsal edge (Fig. 3C). The ventral apex may be somewhat elongated (Fig. 3B).

Olfactory flow in the garpike is likely to be generated primarily by its motion relative to water. Given that adult garpike are likely to be active swimmers, we assume that swimming is the main source of this motion, although an oncoming current could augment olfactory flow. Olfactory flow cannot be generated by the beating of non-sensory cilia, because the nasal region of the adult garpike lacks ciliated non-sensory cells (Theisen et al., 1980). Nor can it be generated by accessory sacs, as the garpike also lacks these anatomical devices. Some olfactory flow could be generated by respiration (the paired nasal regions are situated close to the mouth; Fig. 2), or mechanical agitation of the nasal region (e.g. by movement of the jaws). Mechanical agitation is likely to make only a minor contribution to olfactory flow given that other belonids typically swim with their jaws closed (B. Collette, personal communication).

In undertaking this study, we had two objectives. First, we wanted to demonstrate that olfactory flow in the garpike can be generated by swimming. Second, we wanted to determine how the unusual features of the garpike’s nasal region (nasal cavity, papilla, ventral apex) influence olfactory flow and the likely effect of these influences on odorant sampling in the garpike. As a subject for studying olfactory flow, the garpike was appealing for several, mainly practical, reasons. First, swimming could be simulated by flowing water over a static, life-like, plastic model of the garpike’s head. Moving fluid over a static model, a valid device that is also used to simulate flow over, for example, aircraft (Shapiro, 1961), is equivalent to moving the model through static water, but easier to achieve. Second, the garpike’s nasal region does not possess any moving parts that might complicate olfactory flow (e.g. the flexible lamellae typical of the nasal regions of other fishes). The rigidity of the nasal region could therefore be replicated in the plastic model. Third, the exposed nasal region allows direct observation of olfactory flow. Fourth, there is only one previous study of olfactory flow in the garpike, and this was mentioned only in passing (Zeiske and Hansen, 2005). That study found that the nasal papilla was ‘an adaptation to the hydrodynamic situation, including boundary layer effects of the longirostrate fish’, but no further details were given (Zeiske and Hansen, 2005, p. 22).

The present study forms part of a programme to understand the hydrodynamics of olfaction in fishes, an important but largely unexplored area (Cox, 2008, Abel et al., 2010, Holmes et al., 2011, Howard et al., 2013, Agbesi et al., 2016).