Figure 5 - uploaded by Andrew Digby
Content may be subject to copyright.
1.1: The laughing owl (Sceloglaux albifacies), an extinct predator of little spotted kiwi. Credit: Keulemans, John Gerrard, 1842-1912. Keulemans, John Gerrard, 1842-1912 :Sceloglaux albifacies. J. G. Keulemans del.
Source publication
Contexts in source publication
Context 1
... Figure 5.3.1). False positives were rare, and from a range of sources: occasionally song thrush (Turdus philomelos), tui (Prosthemadera novaeseelandia) or morepork calls were classified as male LSK, but noise fluctuations were also mistaken for kiwi calls. The software detected some calls that were not visible on the spectrogram, but were just audible. False negatives were mainly due to very faint calls, but also res- ulted from acoustic interference from other kiwi and other species. These occurred when bird song overlapped kiwi calls, or when two kiwi calls of the same sex were separated by a gap of less than 5 s, in which case they appeared as only one event due to merging criteria ( § ...
Context 2
... recording location was on a spur in a central region of Zealandia (41.29776 S, 174.74432 E; 190 m elevation), a fenced 225 ha mammalian predator-free reserve in Wellington, New Zealand ( Figure 5. periodicity, frequency modulation of each element, change in amplitude across ele- ments, bandwidth, and signal-to-noise ratio. A weighted combination of these scores was used to produced a combined score, and events above a given threshold of this score are marked as candidate kiwi calls. The scores and their weighting were op- timised to detect LSK calls, whilst minimising false positives from other crepuscular or nocturnal species with similar calls that share LSK habitat, such as kaka (Nestor meridionalis; a parrot) , weka (Gallirallus australis; a rail), and morepork (Ninox novaeseelandiae; an owl). A minimum length of 7 s was imposed for all events. To avoid duplicate detections from calls containing pauses, a Python script was used to merge events if they were within 5 s of another event in the same frequency band and with a similar combined score. Processing time on a Windows PC (4-core Xeon 2.4 GHz; 4 GB RAM) was approximately two minutes for each hour of ...
Context 3
... validation showed that residuals conformed to assumptions of normality and homogeneity. Auto-correlation functions indicated very minor lack of independ- ence at some lags ( Figure 5.3.4), but no residual patterns as seen in the models without correlation structure. A temporal variogram confirmed that there were no significant independence problems ( Figure ...
Context 4
... validation showed that residuals conformed to assumptions of normality and homogeneity. Auto-correlation functions indicated very minor lack of independ- ence at some lags ( Figure 5.3.4), but no residual patterns as seen in the models without correlation structure. A temporal variogram confirmed that there were no significant independence problems ( Figure ...
Context 5
... call rates for both sexes showed a decrease with increasing temper- ature and cloud cover. Calls decreased significantly with rain and wind speed, al- though there were differences in how the sexes were affected by these variables ( Figure 5.3.3). There was a weak tendency for calls to increase with rising humidity, and for lower call rates at intermediate air pressures. Calls initially increased sharply from dry to moist ground conditions. Wind direction substantially affected the num- ber of calls detected at the study site, with fewer calls during southerly winds, and more during ...
Similar publications
Citations
... Turning a call density into the underlying animal density D a requires either the ability to identify individuals in an unsupervised way, or prior knowledge of their call rate μ, i.e., mean number of calls per individual per hour. The first is generally extremely difficult, and the second may well be hard to identify (see e.g., Digby (2013) for the little spotted kiwi, the species we focus on in the experimental section). However, if the true call rate μ was known, the animal density would be estimated aŝ D a =D/μτ , where τ is the survey duration . ...
... To show how SCR can be used to evaluate design tradeoffs in a practical setting, we conducted an acoustic survey of little spotted kiwi (Apteryx owenii) in Zealandia, a 225 hectare wildlife sanctuary in Wellington, New Zealand. The likely population size is around 200 kiwi, with 40 kiwi translocated to this site in 2001, and approximately 100 estimated in 2010 (Digby 2013). This species is nocturnal, and both sexes call intermittently throughout the night, sometimes in male-female duets. ...
... The number of syllables and exact pitch vary between calls. Further details on LSK acoustics can be found in Digby (2013) and Digby et al. (2014). There are other night callers at the study site, some of which can be confused with the kiwi pukupuku, particularly ruru (Ninox novaeseelandiae, a native owl). ...
Passive acoustic surveys provide a convenient and cost-effective way to monitor animal populations, and methods for conducting and analysing such surveys are undergoing rapid development. However, no standard metric exists to evaluate the proposed changes. Furthermore, the metrics that are commonly used are specific to a single stage of the survey workflow, and may not reflect the overall effects of a design choice. Here, we attempt to define the effectiveness of acoustic surveys conducted in two common frameworks of population inference—occupancy modelling and spatially explicit capture-recapture (SCR). Specifically, we investigate precision as a possible metric of survey performance, but we observe that it does not lead to generally optimal designs in occupancy modelling. In contrast, the precision of the SCR density estimate can be optimised with fewer experiment-specific parameters. We illustrate these issues using simulations. We further demonstrate how SCR precision can be used to evaluate design choices on a field survey of little spotted kiwi (Apteryx owenii). We compare call recognition by software and human experts. The resulting tradeoff between missed calls and faster data throughput was accurately captured with the proposed metric, while common metrics failed to identify optimal improvements and could be inflated by deleting data. Due to the flexibility of SCR framework, the approach presented here can be applied to a wide range of different survey designs. As the precision is directly related to the power of subsequent inference, this metric evaluates design choices at the application level and captures tradeoffs that are missed by stage-specific metrics, enabling reliable comparison of survey methods.
... The parameter D then corresponds to call density, and turning this into the underlying animal density D a requires either the ability to identify individuals in an unsupervised way, or prior knowledge of their call rate µ. The first is generally extremely difficult, and the second may well be hard to identify (see e.g., Digby (2013) for kiwi, the species we focus on in the experimental section). However, even if the true call rate µ was known, the animal density would be estimated asD a =D/µ . ...
... To show how SCR can be used to evaluate design tradeoffs in a practical setting, we con- (Digby, 2013). Kiwi pukupuku has a life expectancy of >30 years; once individuals become adult, mortality is low, and pairs produce a 1-2 egg clutch each year (Colbourne, 1992;Robertson and Colbourne, 2004;Jolly, 1989). ...
... The number of syllables and exact pitch vary between calls. Further details on kiwi acoustics can be found in Digby (2013) and Digby et al. (2014). ...
Passive acoustic surveys provide a convenient and cost-effective way to monitor animal populations. Methods for conducting and analysing such surveys, especially for performing automated call recognition from sound recordings, are undergoing rapid development. However, no standard metric exists to evaluate the proposed changes. Furthermore, most metrics that are currently used are specific to a single stage of the survey workflow, and therefore may not reflect the overall effects of a design choice.
Here, we attempt to define and evaluate the effectiveness of surveys conducted in two common frameworks of population inference – occupancy modelling and spatially explicit capture-recapture (SCR). Specifically, we investigate precision (standard error of the final estimate) as a possible metric of survey performance, but we show that it does not lead to generally optimal designs in occupancy modelling. In contrast, precision of the SCR density estimate can be optimised with fewer experiment-specific parameters. We illustrate these issues using simulations.
We further demonstrate how SCR precision can be used to evaluate design choices on a field survey of little spotted kiwi ( Apteryx owenii ). We show that precision correctly measures tradeoffs involving sampling effort. As a case study, we compare automated call recognition software with human annotations. The proposed metric captured the tradeoff between missed calls (8% loss of precision when using the software) and faster data through-put (60% gain), while common metrics based on per-second agreement failed to identify optimal improvements and could be inflated by deleting data.
Due to the flexibility of SCR framework, the approach presented here can be applied to a wide range of different survey designs. As the precision is directly related to the power of detecting temporal trends or other effects in the subsequent inference, this metric evaluates design choices at the application level, and can capture tradeoffs that are missed by stage-specific metrics, thus enabling reliable comparison between different experimental designs and analysis methods.
... A male tuatara had been recorded outside the nest burrow on 22 November 2011, 15 days after the male LSK began incubating (see Table S1 in Supplementary Material), but had not been recorded since that time. At the time the interaction described below was recorded, the nest was known to contain the male LSK's chick, which was estimated to be 8 days old (Digby 2013;HRT unpubl. data). ...
... This behaviour is seemingly similar to the reluctance of fairy prions/tītī wainui to enter nest burrows when tuatara are present (Corkery et al. 2015). Meanwhile, the LSK chick's calls are typical of 'deterministic chaos', random, non-linear vocalisations hypothesised to be used by animals in situations where it is important not to be ignored (Digby 2013) and thus could be a distress call. In light of the above, it is difficult to interpret the sequence of events described here as anything other than antagonistic. ...
Human-induced reductions in species’ ranges have resulted in the geographic separation of some previously sympatric species that interacted historically. Some previously co-occurring species are now being reconnected via translocation. However, interactions between these species can be difficult to predict, particularly in extreme instances where all populations of previously co-occurring species have become completely separated from each other. Here, we present video footage that, for the first time, captures an interaction between two species separated for centuries due to human disturbance, but that are now being reconnected via translocations; little spotted kiwi (LSK) (Apteryx owenii) and tuatara (Sphenodon punctatus). The video shows an aggressive interaction, apparently caused by competition for a burrow being used by the LSK for nesting. This footage suggests we have much to learn about how these species may have co-existed prior to human arrival in New Zealand.
... Several studies have shown that automating the process of detecting calls using software packages has potential in this regard (Bardeli et al., 2010;Brandes, 2008;Digby, 2013;Graff, 2014;Steer, 2010). ...
The inferences that can be made from any study are limited by the quality of the sampling design. By bad luck, when monitoring species that are difficult to detect (cryptic), sampling designs become dictated by what is feasible rather than what is desired. We calibrated and conducted a cost‐benefit analysis of four acoustic recorder options that were being considered as potential solutions to several sampling restrictions experienced while monitoring the Australasian bittern, a cryptic wetland bird. Such sampling restrictions are commonly experienced while monitoring many different endangered species, particularly those that are cryptic. The recorder options included mono and stereo devices, with two sound file processing options (visual and audible analysis). Recording devices provided call‐count data similar to those collected by field observers but at a fraction of the cost, which meant that “idealistic” sampling regimes, previously thought to be too expensive, became feasible for bitterns. Our study is one of the few to assess the monetary value of recording devices in the context of data quality, allowing trade‐offs (and potential solutions) commonly experienced while monitoring cryptic endangered species to be shown and compared more clearly. The ability to overcome challenges of monitoring cryptic species in this way increases research possibilities for data deficient species and is applicable to any species with similar monitoring challenges.
... Female kiwis, weka, and morepork sounds were more attenuated when the bird was close to the ground (0.25 m) than at 3 m height (Figure 3), the attenuation in the forest was always lower than in the open field ( Figure S4.2), and the difference was higher during the night (Figure 2b). The calling posture of at least two kiwi species, brown and little spotted (Digby, 2013), could be their natural adjustment to this: Both adopt a unique calling posture, extending the neck and pointing the bill upwards so that their sounds can probably avoid some ground attenuation. The North Island robin is a ground forager and hihi use the ground for foraging and copulating, and hence, both species mainly call under the canopy. ...
The use of automatic acoustic recorders is becoming a principal method to survey birds in their natural habitats, as it is relatively noninvasive while still being informative. As with any other sound, birdsong degrades in amplitude, frequency, and temporal structure as it propagates to the recorder through the environment. Knowing how different birdsongs attenuate under different conditions is useful to, for example, develop protocols for deploying acoustic recorders and improve automated detection methods, an essential part of the research field that is becoming known as ecoacoustics. This article presents playback and recapture (record) experiments carried out under different environmental conditions using twenty bird calls from eleven New Zealand bird species in a native forest and an open area, answering five research questions: (1) How does birdsong attenuation differ between forest and open space? (2) What is the relationship between transmission height and birdsong attenuation? (3) How does frequency of birdsong impact the degradation of sound with distance? (4) Is birdsong attenuation different during the night compared to the day? and (5) what is the impact of wind on attenuation? Bird calls are complex sounds; therefore, we have chosen to use them rather than simple tones to ensure that this complexity is not missed in the analysis. The results demonstrate that birdsong transmission was significantly better in the forest than in the open site. During the night, the attenuation was at a minimum in both experimental sites. Transmission height affected the propagation of the songs of many species, particularly the flightless ones. The effect of wind was severe in the open site and attenuated lower frequencies. The reverberations due to reflective surfaces masked higher frequencies (8 kHz) in the forest even at moderate distances. The findings presented here can be applied to develop protocols for passive acoustic monitoring. Even though the attenuation can be generalized to frequency bands, the structure of the birdsong is also important. Selecting a reasonable sampling frequency avoids unnecessary data accumulation because higher frequencies attenuate more in the forest. Even at moderate distances, recorders capture significantly attenuated birdsong, and hence, automated analysis methods for field recordings need to be able to detect and recognize faint birdsong.
... Great Spotted Kiwi and North Island Brown Kiwi vocalisations were more highly individualised than Little Spotted Kiwi, this suggests that acoustic based identification may be most effective in these species. Digby et al. (2014) speculated that this disparity in vocal individuality may derive from reduced genetic diversity in the Little SpottedKiwi, or high variability in recording quality(Digby 2013) ...
The ability to discriminate between individuals on the basis of call features has proven to be a useful tool for monitoring rare, nocturnal and cryptic bird species. In this study, vocal individuality was assessed in a population of Great Spotted Kiwi (Apteryx haastii) in the Hawdon Valley, New Zealand. Acoustic recorders were installed near the nest-sites of seven pairs between November 2012 and March 2013. Recorders were again installed at the new nesting sites for three of these same pairs in the following season (September-December 2013). Temporal and spectral parameters of calls were measured and stepwise discriminant function analysis was used to determine whether these parameters were effective in discriminating between individuals. The analyses indicated that Great Spotted Kiwi vocalisations were highly individualised. Male individuals were classified with an accuracy of 95.7%, on the basis of seven parameters, and females with an accuracy of 90%, on the basis of five call parameters. Individuals could also be accurately identified between years, suggesting that the individual features of calls are temporally stable. These findings suggest that Great Spotted Kiwi vocalisations could be used for acoustic identification purposes.
... Little Spotted Kiwi have a long-range 'whistle' call (hereafter 'calls'), which is produced by both sexes, usually during the first few hours of the night, although timing of calling varies during incubation (Digby 2013). The inter-sexual differences in these calls are not a result of size dimorphism but suggest functional call-divergence, with male calls more suited to territory defence and those of females better suited for short-range functions, such as mate-contact . ...
... Calls were recorded serendipitously as often as possible, but playback was also used to elicit calls. This was to ensure a sufficient sample size, because LSK, especially females, call infrequently (Digby 2013). During playback, male and then female LSK calls from the same population were broadcast in the vicinity of a LSK using a digital recorder (Sony PCM-M10P; Sony Corporation, Tokyo, Japan) and small 8-W loudspeaker (AJ Productions Ltd, Hamilton, New Zealand). ...
... These were avoided by retaining only syllables between the 15th and 35th syllables in each call. These limits were chosen by inspection of the 95% range of frequencies of all calls (Digby 2013). After calls were trimmed in this way, those containing fewer than five remaining syllables were discarded to ensure sufficient sample size. ...
The ability to identify individuals by call is particularly important for monitoring cryptic or nocturnal species. It provides key benefits in conservation management, such as allowing correction of biases in censuses and monitoring individual survival. Call surveys are a key tool in conservation of kiwi (Apterygidae), yet individuality of calls has previously been assessed in just one of the five species, the Brown Kiwi (formerly North Island Brown Kiwi, Apteryx mantelli). We have made the first test of whether Little Spotted Kiwi (Apteryx owenii; LSK) can be individually identified by call. Three classification methods (discriminant function analysis, support vector machines and statistical classifiers) were used to assign identity from either call or syllable variables. Using syllables, classification was significantly better than expected by chance for both males and females. Individual LSK can therefore be identified acoustically, suggesting that vocalisations may be used for conspecific identification. Classifier performance was worse using call variables, with no better than chance assignment for females and low accuracy for males. Significant differences in classification ability between the sexes support the hypothesis that the function of calls of male and female LSK differ. Furthermore, identification by call was much less reliable in LSK than in the more genetically diverse Brown Kiwi. This indicates that call diversity may be related to genetic diversity in kiwi, which, if confirmed, could provide a powerful conservation tool for rapid assessment of genetic differences among populations of these threatened species.
... In this study we investigated factors aff ecting nocturnal avian vocal behaviour, using LSK as a study species. Both sexes of LSK produce long-range ' whistle ' calls comprised of a series of 20 – 30 repeated syllables (Digby et al. 2013a). We made continuous recordings of a single population over a three-year period. ...
... Th e recorder location (41.29776S, 174.74432E, 190 m elevation) was not changed during the study, to avoid confounding eff ects from variation in acoustic characteristic and kiwi densities between sites. Th e estimated detection radius of at least 300 m (Digby et al. 2013b) extended into at least 14 LSK territories, as determined from night sightings and a New Zealand Dept of Conservation census (H. ...
... Th e classifi er assigned a kiwi/non-kiwi classifi cation and confi dence score (0 – 1) to each event. Sex was determined by the event frequency, since male and female LSK calls occupy non-overlapping frequency ranges (Digby et al. 2013a). Th e full procedure is described in Digby et al. 2013b. ...
Temporal and environmental variation in vocal activity can provide information on avian behaviour and call function not available to short-term experimental studies. Inter-sexual differences in this variation can provide insight into selection effects. Yet factors influencing vocal behaviour have not been assessed in many birds, even those monitored by acoustic methods. This applies to the New Zealand kiwi (Apterygidae), for which call censuses are used extensively in conservation monitoring, yet which have poorly understood acoustic ecology. We investigated little spotted kiwi Apteryx owenii vocal behaviour over 3 yr, measuring influences on vocal activity in both sexes from time of night, season, weather conditions and lunar cycle. We tested hypotheses that call rate variation reflects call function, foraging efficiency, historic predation risk and variability in sound transmission, and that there are inter-sexual differences in call function. Significant seasonal variation showed that vocalisations were important in kiwi reproduction, and inter-sexual synchronisation of call rates indicated that contact, pair-bonding or resource defence were key functions. All weather variables significantly affected call rates, with elevated calling during increased humidity and ground moisture indicating a relation between vocal activity and foraging conditions. A significant decrease in calling activity on cloudy nights, combined with no moonlight effect, suggests an impact of light pollution in this species. These influences on vocal activity provide insight into kiwi call function, have direct consequences for conservation monitoring of kiwi, and have wider implications in understanding vocal behaviour in a range of nocturnal birds.
... With spectral features such as NLP, two-voicing and formants infrequent, and low inter-individual call variation (Digby 2013), it is feasible that LSK do not use calls for recognition at all. As socially monogamous birds defending year-round territories, these kiwi may have little requirement for vocal individual identification and may assess every nearby calling intruder as a potential threat, irrespective of identity (Bard et al. 2002). ...
Non-linear phenomena (NLP) in vocalizations may have adaptive functions, such as increasing acoustic impact or conveying fitness or identity information. Yet despite their potential to impart behavioural information, these spectral features have not been studied in most bird species and many genera. This applies to the New Zealand kiwi (Apterygidae), with NLP-like features indicated in one species, but not explored further. Yet as nocturnal and cryptic birds heavily reliant on vocal communication, kiwi are ideal species in which to assess the function of these structures. Furthermore, kiwi are acoustically typical but taxonomically and ecologically unique among birds, so can provide an important reference for determining the behavioural relevance of NLP in other bird species. We have assessed the occurrence of NLP in little spotted kiwi (Apteryx owenii) calls, in the first detailed study of such features in a ratite. NLP in the form of subharmonics are common in this species. We tested for possible adaptive functions of these features by comparing their occurrence with call spectral parameters and conditions. Subharmonics increased with frequency and during calls produced in territorial contexts, indicating that these features may provide acoustic impact, possibly to convey aggression or fitness information. Our results also suggest that NLP are unlikely to provide identification information in this kiwi species.
Revealing behavioral secrets in extinct species
Extinct species had complex behaviors, just like modern species, but fossils generally reveal little of these details. New approaches that allow for the study of structures that relate directly to behavior are greatly improving our understanding of the lifestyles of extinct animals (see the Perspective by Witmer). Hanson et al. looked at three-dimensional scans of archosauromorph inner ears and found clear patterns relating these bones to complex movement, including flight. Choiniere et al. looked at inner ears and scleral eye rings and found a clear emergence of patterns relating to nocturnality in early theropod evolution. Together, these papers reveal behavioral complexity and evolutionary patterns in these groups.
Science , this issue p. 601 , p. 610 ; see also p. 575
