Do Hydrothermal Shrimp Smell Vents?

Abstract

Deep-sea species endemic to hydrothermal vents face the critical challenge of detecting active sites in a vast environment devoid of sunlight. This certainly requires specific sensory abilities, among which olfaction could be a relevant sensory modality, since chemical compounds in hydrothermal fluids or food odors could potentially serve as orientation cues. The temperature of the vent fluid might also be used for locating vent sites. The objective of this study is to observe the following key behaviors of olfaction in hydrothermal shrimp, which could provide an insight into their olfactory capacities: (1) grooming behavior; (2) attraction to environmental cues (food odors and fluid markers). We designed experiments at both deep-sea and atmospheric pressure to assess the behavior of the vent shrimp Rimicaris exoculata and Mirocaris fortunata, as well as of the coastal species Palaemon elegans and Palaemon serratus for comparison. Here, we show that hydrothermal shrimp groom their sensory appendages similarly to other crustaceans, but this does not clean the dense bacterial biofilm that covers the olfactory structures. These shrimp have previously been shown to possess functional sensory structures, and to detect the environmental olfactory signals tested, but we do not observe significant attraction behavior here. Only temperature, as a signature of vent fluids, clearly attracts vent shrimp and thus is confirmed to be a relevant signal for orientation in their environment.

Keywords: antennules; behavior; chemosensory perception; grooming; hydrothermal shrimp; olfaction; thermal detection.

Figure 1

State of the art on hydrothermal shrimp olfaction. (a) Previous studies conducted by our team on olfaction in hydrothermal shrimp are summarized in this figure. Panel 1 shows the anatomy and ultrastructure of the olfactory organs of M. fortunata. The lower image shows the antennule, which consists of a median flagellum (left) and a lateral flagellum with aesthetasc sensilla (right). The upper image shows a cross section of an aesthetasc with the cuticle (c), bacteria (b) on the outer surface and dendritic segments (d) of olfactory neurons inside. Panel 2 shows a scheme of an aesthetasc sensilla from a marine crustacean decapod. Detection of various ecologically relevant chemical stimuli (at short and long range) by the antennal appendages was measured by electroantennography. Several ionotropic receptors, including the IR25a co-receptor putatively involved in olfaction, were identified and shown to be mainly expressed in the lateral flagellum of the antennules bearing the aesthetascs. Panel 3 shows a schematic representation of the organization of crustacean brain in dorsal view. Olfactory neuropils (black arrow) receive the sensory input from olfactory neurons that innervate the aesthetasc sensilla. (b) Key results from studies on olfactory organs anatomy and stimuli detection [23,24,25], and on brain anatomy [26,27].

Figure 2

Species collection and experimental condition for behavioral experiments at in situ pressure. * Type of test applied either to a batch of shrimp or to single specimens; ** duration and condition of maintenance in pressurized aquaria before the experiment; *** number of shrimp for each experiment (some experiments were performed several times, see figure legends for more information).

Figure 3

Species collection and experimental condition for behavioral experiments at atmospheric pressure. * Type of test applied either to a batch of shrimp or to single specimens; ** duration and condition of maintenance at atmospheric pressure before the experiment; *** number of shrimp for each experiment (some experiments were performed several times, see figure legends for more information); # experiments with sulfide and temperature were carried out for comparison with hydrothermal species.

Figure 4

Setups for experiments in the pressure vessel IPOCAMP. (a) The pressure vessel IPOCAMP (internal diameter 20 cm, height 60 cm [36]); (b) experimental setup for grooming observations. Arrows indicate the inlet and outlet of circulating seawater. This device also used for chemical stimuli experiments, see (c), is a recently upgraded version of the lid of the vessel IPOCAMP. The new lid comprises a large viewport and consequently allows direct observation as well as video recording with a high definition camera (ca, AG-HCK10G HD camera head, AG-HMR10 portable recorder, Panasonic). Three shrimp were placed in a PVC cage closed at the top with a transparent polyethylene lid. The behavior of the shrimp was recorded throughout the experiment. Pictures (b1–b3) are views of the following animals in the IPOCAMP aquarium during grooming observations: (b1) R. exoculata females and male (marked with a black line); (b2) R. exoculata juveniles; (b3) Palaemon serratus juveniles. Body length of the observed shrimp is about 4–5 cm for R. exoculata and P. serratus juveniles, and about 2 cm for R. exoculata juveniles. (c) Experimental setup for chemical stimuli experiments. Arrows indicate the inlet and outlet of circulating seawater. The seawater inlet pipe passes through a thermostatically controlled bath. The lid is equipped with an isobaric line (i) that allows the introduction of small elements (e.g., food, stimulus) without disrupting the pressure inside the aquarium. During the experiment, two control gels (grey bars) and one stimulus gel (black bar) were introduced into the tank through the isobaric line at 45 -min intervals. Picture (c1) shows Mirocaris fortunata in the IPOCAMP aquarium during this experiment. Specimens lie on the plastic bottom of the tank with holes for water entrance, two stainless steel tubes that contain gels are visible. (d) Experimental setup for temperature and sulfide pulse experiments. Three sapphire viewports in the pressure vessel lid allow the insertion of an endoscope (e) (Fort, Dourdan, France) and two optical -fiber light -guides for the behavioral observation. The experiments are recorded by a CCD camera (JVC, TK-C1380) and a DVD recorder (DVO-1000MD, Sony). A diffuser system, consisting of a plastic cap with holes, is placed over the seawater inlet hole. Warm -temperature pulses were obtained by immersing a heat exchanger (HE) located on the water inlet line in a temperature -controlled water bath that had been preset to the desired pulse temperature. Two Pt-100 autonomous temperature loggers (S2T6000D, NKE Instruments) are positioned in the upstream water flow (T1) and on the plastic plate at the bottom of the tank several centimeters (>5 cm) from the hot water diffuser (T2). For the sulfide pulse experiments, the HE2 and temperature loggers were removed, and the sulfide solutions were injected into the seawater inlet pipe. Picture (d1) shows Rimicaris exoculata in IPOCAMP during a temperature pulse experiment at 25 °C. Specimens gathered around the warm water diffuser (marked with a dark circle).

Figure 5

Experimental setups for experiments in aquaria at atmospheric pressure. Perspective (a) and birds eye (b) views of the experimental setup for two-choice experiments on single P. elegans and M. fortunata. The shrimp was placed in the aquarium (P. elegans, room temperature; M. fortunata, 10 °C) and left to explore for 5 min. Two gauze bags (one control (C), one stimulus (S)) were introduced on each side of the tank. The shrimp were observed for 5 min. Perspective (c) and birds eye (d) views of the experimental setup for multiple-choice experiments on single P. elegans and M. fortunata. The shrimp was placed in the aquarium (P. elegans, room temperature; M. fortunata, 10 °C) after the introduction of four tubes containing agarose gels (three controls (C), one stimulus (S)) in the corners of the tank. The behavior of the shrimp was recorded for 30 min using a camera placed above the aquarium. Perspective (e) and birds eye (f) views of the experimental setup for two-choice experiments on a batch of M. fortunata. The experiments were conducted on several individuals in rearing tanks at 9 °C containing a heating thermostat (H) set at 25 °C. Two gels (one control (C), one stimulus (S)) were introduced on each side of the tank. The shrimp were observed for 30 min. Perspective (g) and birds eye (h) views of the experimental setup for experiments of choice between ON and OFF thermostats on a batch of P. elegans and M. fortunata. The experiments were conducted in rearing tanks at 9 °C containing several individuals of P. elegans or M. fortunata. Two temperature thermostats (one ON, the other OFF) were placed on the lateral sides of the tank, close to the surface. The number of shrimp positions on each thermostat was counted for 180 min, and then after one night. Arrows indicate the inlet and outlet of circulating seawater (SW).

Figure 6

Frequency of grooming behavior of olfactory appendages in Rimicaris exoculata and Palaemon serratus. Frequencies are expressed as the number of grooming events per min (mean ± S.D.) for each batch of individuals for adult specimens of R. exoculata (n = 3) and for the 3 batches of individuals for juveniles of P. serratus (n = 9). The observations were carried out over a period of 53 min for each batch, and observations for the 3 batches of P. serratus juveniles were pooled, therefore corresponding to a total observation time of 159 min. For R. exoculata juveniles (n = 8), 9 observations were carried out over a period of approximately 17 min on randomly chosen individuals, corresponding to a total observation time of 153 min. The experimental conditions are as follows: (1) R. exoculata adult batch 1: direct observation, 23 MPa, 10 °C; (2) R. exoculata adult batch 2: 7 h maintenance before observation, 30 MPa, 20 °C; (3) R. exoculata adult batch 3: 48 h maintenance before observation, 30 MPa, 10 °C; (4) R. exoculata juveniles: 96 h maintenance before observation, 30 MPa, 20 °C; (5) P. serratus juveniles: 3 weeks to 1 month maintenance before observation, atmospheric pressure, 10 °C. * Significant difference in the frequency of grooming (n = 9, Student’s paired sample test, p = 0.012).

Figure 7

Scanning electron micrographs of Palaemon serratus (juveniles) before (Ps1, Ps2) and after (Ps10, Ps11) grooming experiments. (a,c,e) Antennules of Ps10 (a), Ps1 (c) and Ps2 (e) specimens showing gradation from absent (a), to moderate (c) and intense (e) bacterial fouling on the antennal segments. (a) Antennules of Ps10 specimen, showing the two ramus and the aesthetascs completely devoid of bacterial fouling; (b) close-up on the aesthetascs of Ps11 specimen completely devoid of bacterial fouling; (c,d) antennae of Ps1 specimen with a light fouling of bacteria. Frame in (c) is enlarged in d showing the bacterial morphological variety, with thick and thin filamentous bacteria, some rods and cocci; (e,f) antennules of Ps2 specimen with very dense fouling covering the entire surface of the segments. Frame in (e) is enlarged in (f), showing bacterial density and morphological diversity. As: aesthetascs, R1 and R2: the two rami of the lateral antennular flagellum. Scale bars: (a–c,e) = 100 µm; (d,f) = 10 µm.

Figure 8

Scanning electron micrographs of Rimicaris exoculata before (Rex1, Rex5) and after (Rex38) grooming experiments. (a,c,e) Antennules of Rex1 (a), and Rex5 (c,e) specimen showing gradation from no, light and high bacterial fouling on the antennal segments; (b,d,f) Antennule of Rex38 specimen showing dense bacterial fouling on the aesthetascs (b) and beaked setae (d,f) (terminology used in [23]) including their terminal pore. In (d) arrow shows the location of the pore covered by rods). Frame in (e) is enlarged in (f), showing the bacterial (rods) mat covering the surface of the segments, the filamentous bacteria (fb) in the inter-segmental areas and the base of a short thin setae, as well as and an intermediate beaked setae covered by rod-shaped bacteria. As: aesthetascs, Bs: beaked seta, fb: filamentous bacteria, rb: rod-shaped bacteria. Scale bars: (a,c,e) = 100 µm; (b,f) = 10 µm; (d) = 2 µm.

Figure 9

Responses to sulfide pulse stimuli at in situ pressure on a batch of Mirocaris fortunata and Rimicaris exoculata. (a) Mirocaris fortunata. Mean % of shrimp over an area of 6-cm² surface around the seawater inlet hole. Three injections were carried out with increasing concentrations of sulfide solutions (25, 50 and 100 µM), and these sulfide pulses are indicated by a black bar along the time scale. One-hour interval separates each sulfide pulse. The experiment was conducted twice on the same batch (n = 20) of shrimp (i.e., each point represents the mean of two replicates). The experimental setup is depicted in Figure 4c, with the following modifications: removal of the HE2, the two temperature loggers and the diffuser plastic cap; (b) Rimicaris exoculata. % of shrimp in contact with the diffuser. Three injections were carried out once on the same batch of shrimp (n = 20), with increasing concentrations of sulfide solutions (10, 75 and 300 µM). The number of shrimp close to the pulse entrance was 0 for the experiment with 10 µM Na2S, and thus the result line is merged with the x-axis. A black bar along the time scale indicates these sulfide pulses. One-hour interval separates each sulfide pulse. The experimental setup is depicted in Figure 4c, with the following modifications: removal of the HE2 and the two temperature loggers.

Figure 10

Response to food and sulfide stimuli during experiments at in situ pressure on a batch of Rimicaris exoculata and Mirocaris fortunata. The shrimp were placed in the IPOCAMP aquarium for a recovery period of 2 h at 30 MPa and 10 °C (see Figure 4c for the setup description). Three gels (2 controls, 1 stimulus) were then introduced consecutively through an isobaric line with an interval of 45 min. The number of times of contact of the shrimp with the newly introduced gel was quantified over a period of 45 min (a) Rimicaris exoculata. Mean number of times of contact (±S.E.M.) with the stimulus (S, agarose gel loaded with a 2mM sulfide solution at pH11) and the control gels (C1 and C2, agarose gels). Three batches of 10 shrimp were tested once. The mean number of times of contact with the stimulus gel was compared to those of control gels with a two-tailed t-test for correlated samples (same batch of shrimp under different test conditions) (df = 2), and was not significantly different (NS) (S/C1, p = 0.068; S/C2, p = 0.098). (b) Rimicaris exoculata. Mean number of contacts (±S.E.M.) with the stimulus (S, agarose gel loaded with a 2mM sulfide solution at pH 4, black bar) and the control gels (C1, agarose gel, white bar; C2, pH 4 agarose gel, light grey bar). Three batches of 10 shrimps were tested once. The mean number of times of contact with the stimulus gel was compared to that of control gels with a two-tailed t-test for correlated samples (df = 2), and was not significantly different (NS) (S/C1, p = 0.901; S/C2, p = 0.965). (c) Mirocaris fortunata. Mean number of times of contact (±S.E.M.) with the stimulus (S, agarose gel loaded with mussel extract, black bar) and the control gels (C1 and C2, agarose gels, white bars). Two batches of 6 and 5 shrimp were tested once, and no statistical analysis was performed (n = 2).

Figure 11

Responses to food and sulfide stimuli during multiple choice experiments on single individuals of Mirocaris fortunata and Palaemon elegans. (a) Mirocaris fortunata. Upper graph (bar chart): mean time (% ± S.E.M.) spent in the tube containing the stimulus odor (S, agarose gel loaded with mussel extract, black bar) and in the 3 control tubes (C, agarose gels, grey bar). The means were compared with a two-tailed t-test (not significantly different, n = 10). (b–d) Palaemon elegans. Upper graphs (bar charts): mean time (% ± S.E.M.) spent in the tube containing the stimulus odor (S, agarose gel loaded with mussel extract ((b), n = 23), mussel extract + sulfide 2 mM ((c), n = 17), sulfide 2 mM ((d), n = 12), black bar) and in the 3 control tubes (C, agarose gels, grey bar). The means were compared with a two-tailed t-test ((b), p = 0.008, ** p < 0.01, significantly different; (c), p = 0.001, *** p < 0.005, significantly different; (d), not significantly different). Lower graphs (pie charts) present the distribution of the shrimp according to their first entrance in the stimulus tube (S, black), a control tube (C, grey) or none (white).

Figure 12

Responses to warm-temperature pulses at in situ pressure on a batch of Rimicaris exoculata. (a) Profile of warm temperature pulses. The baseline temperature was set at 4 °C, the first two pulses were set at 25 °C, the next two pulses at 10 °C and the last two pulses at 5 °C, with one hour between each pulse; (b) percentage of shrimp (n = 20) in contact with the diffuser for each pulse. The temperature profile is plotted for each 30 min observation period (symbolized by a grey bar on the temperature profile in (a)). For the last temperature pulse, observations after 15 min are missing due to the poor quality of the video (camera field of view obstructed by shrimp and low light).

Figure 13

Two-choice experiment (on vs off heating thermostats) on Mirocaris fortunata and Palaemon elegans. Distribution of M. fortunata (a) and P. elegans (b) on the on and off thermostats over time. Two batches of 28 and 19 M. fortunata were tested 6 times each (n = 12 replicas per point, except overnight, n = 4 replicas). Two batches of 20 P. elegans were tested 4 times each (n = 8 replicas per point, except overnight, n = 4 replicas). The thermostats either on (set to 25 °C) or off were introduced on each side of the rearing tank (9 °C) in the upper region. The shrimp were observed for 30 min. The on and off thermostats were inverted between two consecutive trials. The distribution is presented as mean % individuals (± S.E.M.) on each thermostat.