Xanthocnemis zealandica Larval Behavior

The study centers on Xanthocnemis zealandica, a widespread dragonfly species endemic to New Zealand. Its larvae are commonly found in the littoral zones of lakes and ponds, streams, and rivers. While usually semivoltine (taking two years to mature), they can mature in one or three years. The study utilized a population whose seasonal regulation was detailed in Deacon (1979), with general species biology outlined in Rowe (in press). Larvae were collected primarily from a pond at the Cass Biological Field Station (Remus pond, 43°02'S 171°46'E) and supplementary stocks from L. Sarah (43°03'S 171°47'E). Additional larvae were examined from several other sites across New Zealand, including Kaiiwi Ls (35°48'S 173°39'E), Bethel's swamp (36°54'S 174°27'E), L. Pupuke (36°47'S 174°46'E), Dillmanstown (42°39'S 171°12'E), Belfast (43°25'S 172°39'E), R. Avon (43°31'S 172°35'E), and Roxburgh (45°29'S 169°19'E). Importantly, X. zealandica was the only species found at these locations. The parentage of all ova used was meticulously determined by observing and examining copulating males, ensuring controlled experimental conditions. The researchers' detailed approach to data collection highlights the rigorous methodology employed in this study of New Zealand's endemic dragonfly populations.

Xanthocnemis zealandica larvae exhibit a remarkably extensive repertoire of 25 agonistic displays, all linked to site defense. These displays are well-developed by instar 5, indicating an early onset of territorial behavior. A sedentary behavior pattern is adopted from the earliest free-living stage. Late instar larvae demonstrate a clear preference for specific sites on stems, favoring those with diameters between 4-7 mm, as observed in experimental settings. This preference for particular stem diameters suggests a selective pressure driving habitat choice and territorial behavior. The observation of these displays in various locations across New Zealand with no significant differences highlights the consistency of the species' territorial behavior patterns. The comprehensive cataloging of the displays provides a rich dataset for understanding the communication strategies involved in X. zealandica's territorial defense and inter-individual competition. The identification and description of these displays represent a key contribution to the knowledge of this species' behavioral ecology.

The larval behavior of X. zealandica contrasts sharply with that of other Xanthocnemis species. While X. zealandica larvae exhibit a sedentary, territorial nature, other species are described as cryptic, thigmotactic, and inactive. The recent radiation within the Xanthocnemis genus, likely originating in the late Pleistocene, offers an evolutionary perspective on the development of this distinct behavioral strategy. The paper discusses other Xanthocnemis species; X. sobrina (a cold stenothermic species), X. tuanuii (found on Chatham Island), and X. sinclairi (a cold-adapted high subalpine species). The contrasting behaviors between X. zealandica and its sibling species suggest that environmental pressures have played a crucial role in shaping their respective behavioral and ecological characteristics. The observed differences offer insights into the adaptive diversification of this genus within the New Zealand landscape, highlighting the impact of environmental factors and evolutionary history on the development of species-specific behaviors.

The researchers employed a less intrusive method for initiating conflicts between X. zealandica larvae, avoiding the potential biases associated with directly manipulating the animals. A more natural observation approach was used, which involved allowing a variable, but generally short interval to elapse before introducing a second larva (the invader) into the observation chamber. This methodology aims to observe the natural responses of the larvae in a more realistic context, rather than eliciting responses as simple automatons to externally imposed stimuli. The rationale behind this approach diverges from previous studies where immediate interaction was triggered (Jansson & Vuoristo 1979, Baker 1981a, Jackson & Pollard 1982). By minimizing manipulation, this study seeks to avoid the ‘shock effect’ and to obtain more reliable data on the natural behavioral patterns displayed during interactions between conspecifics.

The study examined the ontogeny of agonistic displays in Xanthocnemis zealandica larvae, focusing on the relationship between visual development and display effectiveness. In very young instars, vision is limited, and optical signals are likely unimportant. Second instar larvae possess only seven ommatidia per compound eye, severely restricting form discrimination. However, ommatidia numbers increase rapidly; third instar larvae have 12, fourth instar larvae have 28, fifth instar larvae around 88, and sixth instar larvae more than 200. This rapid increase in visual acuity is significant. Fourth instar larvae respond to prey 1.3mm away (one body length), and sixth instar larvae exhibit a dorsal pseudopupil covering nine ommatidia, indicating specialized visual capabilities (Horridge, 1978). The appearance of a dark spot badge on the caudal lamellae coincides with the eye's development to a point where detection of this badge by potential opponents is feasible. This visual signal becomes increasingly important in later instars as the larvae's visual acuity improves.

Between the 6th and 12th instars of X. zealandica, significant changes in both the ritualized behaviors displayed and their frequency of use were observed. Certain displays, such as semaphore, caudal lamellae open/close, and abdomen lift, were only seen in instars 6-8, possibly occurring in older larvae but at such low frequencies they went unnoticed. The 'slash' display, common in early instars, occurred less frequently in later instars, but was again common in metamorphosed larvae (pharate adults). This suggests the display's effectiveness may change according to the larva's size and sensory capabilities. Smaller larvae, with limited sensory information about their opponents, might rely more on a generally effective defense like the 'slash'. Larger larvae with more highly developed sensory apparatuses shift to more specific ploys. The shift in display frequencies and forms over ontogeny reflects the changing competitive dynamics and sensory capabilities of the larvae as they mature.

The study investigates the role of the caudal lamellae and their associated dark spot badges in displays. The size of the badge varies but is typically about 10% of the total body length (15% of the body length without caudal lamellae), covering roughly 30% of the larva's cross-sectional area. Similar badges are found in many larval Coenagrionidae (MacNeill 1960). During displays, these markings enhance the visibility of the caudal lamellae, making them conspicuous to opponents. The markings remain visible even in degraded, out-of-focus images, indicating their effectiveness as visual signals. The long setae on the dorsal and ventral edges of the caudal lamellae may or may not play a function in displays; larvae (instars 2-10) never responded to prey contacting these setae, but they did react swiftly to strike displays aimed at the caudal lamellae themselves. The study shows that the caudal lamellae, alongside their badges, are significant components of the visual communication system of X. zealandica larvae.

The study details the predatory behavior of Xanthocnemis zealandica larvae, showing a sophisticated predatory sequence. After inspecting potential prey with its antennae, the larva strikes with its labium (Richard 1960, Caillere 1965, 1974). This specialized behavior is well-suited to the chaotic sensory environment of flowing water. Observations were made using individual larvae in watchglasses under a stereomicroscope with darkfield illumination. The larvae's behavior was recorded for later transcription. Observations lasted for 30 minutes or until a predetermined number of prey were consumed. The study notes that generalizations about zygopteran larvae as predators, often based on Calopteryx data, present a narrow view, ignoring the diversity of predatory behaviors across the Zygoptera. The study challenges such generalizations, showcasing the more nuanced predatory strategies of X. zealandica.

Observations of second instar X. zealandica larvae revealed a dynamic approach to prey capture. The larvae showed a rapid response to contact with prey, employing a full attack sequence involving strike, grasp, chew, and reject (F). However, as a feeding bout progressed, they displayed partial attacks, sometimes omitting the chew (P) or both grasp and chew (pi) components. The rejection of prey wasn't simply due to inept hunting, but a learned behavior; initially successful in grasping Paramecium, the larvae later learned to reject them, often after a brief chew. Following prey rejection, the larvae thoroughly cleaned their labium and palps. Smaller prey items, like harpactacoid nauplii, were consistently ignored, even when the larvae were hungry. This selective feeding suggests prey discrimination, even with the limited sensory capabilities of the young larvae.

Older X. zealandica larvae exhibit scavenging in addition to active predation, a behavior consistent with the Odonata. This opportunistic feeding allows them to utilize larger food items, potentially explaining the presence in their diet of species they seem unable to prey on in the laboratory, and even evidence of cannibalism. The study found that while large X. zealandica larvae act as visual predators in bright light, they are also adept at locating carrion using olfaction and capturing prey in total darkness, likely using mechanoreceptors or vibration detection. Their ability to hunt agile cladocerans in darkness and respond to conspecific contact at very low light levels (Chapter 2) shows a high level of sensory discrimination. The dietary diversity includes Chironomidae (consistent with Chutter 1961 for Pseudagrion, Lawton 1970b for Pyrrhosoma nymphula, and Thompson 1978c for Ischnura elegans), but can also include gastropods like Potamopyrgus antipodarum (Stark 1981), and even other species like adult Oytiscidae (Coleoptera), larval Chironomidae (Diptera), and larval Leptoceridae (Trichoptera) which they seem to scavenge (unpubl. obs.).

Xanthocnemis zealandica larvae are typically found in bottom habitats, amongst dense vegetation beds, and in or on detritus—all environments with the potential for local hypoxia. Remarkably hardy, these larvae are among the last survivors when field collections become hypoxic, often found near the water's surface or near air bubblers. Under these conditions, they rarely display static caudal swinging (SCS) behavior. The observed behavioral differences between larvae in hypoxic buckets and experimental flasks (where they remained largely inactive on the bottom) may be due to oxygen concentration cues. Their ability to survive extended periods without substantial oxygen, along with their aestivation behavior during droughts (Rowe, in press), suggests a robust respiratory physiology. These larvae's adaptability to variable oxygen levels and their ability to aestivate are key adaptations to their diverse habitat.

The study employed experimental aquaria (25 x 40 x 15 cm deep) with controlled temperature (16°C) and a 16h light/8h dark regime. Artificial stems of varying diameters (2, 4.5, 7, 9, and 11 mm) were used as substrates, stiffened with wire or brazing rod. Ten stems were offered per aquarium to minimize the probability of false records due to larvae reoccupying their original positions. Copepods (Phyllognathopus volcanicus) were provided as prey. Larvae were introduced individually into containers. The aquaria were examined at 1, 24, and 48 hours to record perch occupation patterns. After 48 hours, the rear row of stems was moved to the front, initiating new experimental cycles. This methodology, which accounted for territorial defense behaviors, introduced a systematic bias against finding a site preference and is noteworthy. The experimental design meticulously controlled various parameters to allow for precise observation of site selection.

The study highlights the importance of site selection for X. zealandica larvae, a sit-and-wait predator that spends significant time in one place and depends on passing prey for food. They actively select particular site geometries (Chapter 8), restricting themselves to certain microlocalities (Chapter 8). Stems are used as fishing sites or vantage points from which to attack prey (Macan 1964, 1977; Lawton et al. 1980; Baker 1980). This active site selection, however, is unexpected given their high mobility and lack of investment in site modification. The apparent superabundance of suitable sites and a variable food source makes the persistence of territorial behavior puzzling. This highlights a need for deeper investigation into the ecological and evolutionary drivers of this species' territoriality, particularly when comparing with other less territorial species. The study's observation of habitat use in diverse environments provides insights into the ecological factors that influence site selection preferences in dragonfly larvae.