Camille Pagniello

Emerging, Low-Cost Technology

The vastness of the ocean and the pace at which it changes make it difficult to observe and document. As such, person-intensive methods are not sufficient to characterize marine ecosystems. “Do-it-yourself” technology that is designed specifically for monitoring and co-developed in an adaptative management context directly with conservation practitioners, resource managers and policymakers could help address this gap.

Camille is interested in designing and developing low-cost, non-invasive, multi-sensory monitoring platforms for ocean conservation and management using the “working backwards” ocean solutions development framework.

• Acoustics
• Animals
• Conservation
• Sustainable Management
• Technology

Publications

Seminars, Presentations, Posters and Reports

Bioacoustics

Listening to the biological components of the ocean’s soundscape is a promising method for monitoring animal biodiversity. Passive acoustic monitoring provides good spatial and temporal coverage enabling the mapping of the distribution and occurrence of animals within their habitat.Additionally, most sounds can be uniquely described by their spectral and temporal acoustic characteristics. When species-sound associations are known, passive acoustics can enable species-specific monitoring of marine communities.

Camille uses multi-element passive acoustic arrays in concert with optical imaging to answer three questions:

1. When are animals producing sound?
2. Where are animals producing sound?
3. Who are the animals producing sound?

• Acoustics
• Animals
• Oceanography

Publications

Seminars, Presentations, Posters and Reports

Machine Learning for Detection and Classification

Manually inspecting image and acoustic datasets on the order of terabytes for the presence and absence of different animals and sound types is time consuming. To facilitate the widespread use of image and acoustic methods to estimate marine biodiversity, scientists and conservation practitioners alike need open-source automated detection and classification tools that can identify, separate and classify individual images or signals at a species-level.

Camille's work utilizes physics- and statistical-based approaches in concert with new advances in deep learning to develop a generalized power-law detection algorithm for fish sounds, and train a convolutional neural network to classify multiple fish species’ sounds.

• Acoustics
• Animals
• Conservation
• Data Science
• Technology

Publications

Seminars, Presentations, Posters and Reports

Habitat Mapping

Kelp forests are a prominent feature along the world’s coastlines in temperate and subpolar regions. However, increasing ocean temperatures are threatening the persistence of these highly productive ecosystems. While the distribution of surface-canopy-forming kelp can be mapped using remote sensing, a large portion of the canopy is frequently submerged by currents and are missed by such techniques. Acoustic mapping can capture the sub-surface distribution of kelp.

Camille's work is using a low-cost, commercial “fish finder" to map the distribution of kelp forest. Using this "fish finder" we are able to distinguish surface and understory kelp.

• Acoustics
• Conservation
• Technology
• Sustainable Management

Publications

Acoustic Propagation in Biologics

The primary objective of this research is to develop a unified ocean/seabed acoustic propagation model including scattering from kelp for low- and mid-frequency sounds in shallow water. This characterization of the water column and seabed acoustic properties will be used to improve the ability to detect, localize, and track acoustic sources in coastal environments. Data collected from a controlled source and source of opportunity (i.e., a low-frequency fish chorus) during the 2018 Kelp Acoustic Propagation Experiment (Kelp APE) will be used to help develop and validate the acoustic propagation model.

• Acoustics
• Oceanography

Computational Geometry and Topology

Few invariant measures are defined for fractals. Development of such measures will increase fractal geometry’s utility. This investigation explored the possibility of developing a systematic approach for determining invariants. Our strategy was to work asymptotically towards developing scaled magnitude functions for fractals in R² by approximating a fractal with a finite sequence in a metric space, and attempting to take a limit of those measures. Using the Euclidean metric, it was found that the magnitude functions resulting from even low order fractals are very large and do not immediately lend themselves to asymptotic analysis. General formulae for calculating weights for two and three orbit fractals were developed. In addition, a general form of Leinster and Willerton’s (2009) Lemma 7 was proved. Solutions to some of the difficulties that were met when attempting to solve matrices and determine weights of higher order fractals will be presented.

• Geometry
• Topology

Seminars, Presentations, Posters and Reports

Atmospheric Physics

The CANDAC Rayleigh-Mie-Raman Lidar (CRL) at Eureka, Nunavut (80N, 86W) has been successfully used to measure aerosols and water vapour over the past three years. The field of view (FOV) of the telescope and the lasers alignment to the FOV have a dramatic affect on the overall performance of the lidar. This investigation attempted to characterize this relationship in two ways. First, the entire FOV of the telescope was be mapped, as this provided further information about the alignment of the laser beam for a range of FOVs and thus, determined the affect of the FOV on the optimal alignment. Second, simulations to determine the theoretical change in the background signal received and its relationship to the FOV were done in order to compare it to the experimentally determined relationship.

Camille also helped operate the Dalhousie Ground Station as part of the international BORTAS (Quantifying the impact of BOReal forest fires on Tropospheric oxidants over the Atlantic using Aircraft and Satellites) measurements campaign that studied the chemistry of biomass burning plumes. The experiment featured fly-overs of the British BAe-146 research aircraft, guided by the lidar measurements.

• Arctic Climate
• Air Quality

Publications

Seminars, Presentations, Posters and Reports

Zooplankton Biomass Size Spectrum

Quantifying the size-structured biomass of marine communities and their governing processes is directly relevant to ecosystem-based management of exploited resources, yet shallow, fluctuating, coastal ecosystems, such as the Northumberland Strait, receive little research effort despite the presence of valuable fisheries.

Camille investigated whether spatial and temporal variation in the size-structured biomass of the zooplankton community is related to variation in water mass structure based on oceanographic and plankton abundance-at-size data collected in the Northumberland Strait. Principal component analysis and hierarchical agglomerative clustering of physical variables were used to define distinct oceanographic zones within the Strait. Linear and quadratic normalized biomass size spectra (NBSS) were used to quantify the size-structured biomass of the associated zooplankton community. Correlations amongst the NBSS parameters showed that biological processes governed most of the size structure variation in the zooplankton community. In spring, the NBSS was primarily influenced by biological processes while in autumn physical factors had more influence on the size structure. Top-down processes, such as predation on larger zooplankton, appeared to explain the steeper NBSS slopes relative to the hypothetical steady-state slope of -1.00. Differences in all of the NBSS parameters, except slope, between seasons, were likely due to bottom-up processes such as increased nutrient input. Similarly, bottom-up processes may have resulted in the steeper slopes in coastal waters relative to core waters in late spring. The NBSS also varied between the open ocean and each of the upper and lower Strait. The interpretation of pattern requires caution due to the effects of aliased sampling. Limitations of the quadratic NBSS, when the shape of the spectrum deviates from the parabola, indicate that the linear NBSS better describes processes governing NBSS variation. Overall, the results of this thesis created the first quantitative zooplankton community baseline for detecting ecosystem change in the Strait.

• Oceanography
• Statistics
• Zooplankton

Seminars, Presentations, Posters and Reports

Avian Bioacoustics and Physiology

The “drumming” sound produced by Gallinago gallinago (Common Snipe) and Gallinago delicata (Wilson’s Snipe) were characterized. Both the drum durations and durations of five pulses, i.e. the 7th to the 3rd pulse from the end of the drum, followed a near normal distribution. Trends suggested that little variation occurred in these variables within the species, regardless of their geographic location. Unfortunately, inter-drum intervals proved to be a less than ideal measurement to characterize the drum of Wilson’s Snipe.

Additionally, Strigiformes specimens and vocal tracts were dissected and measured using dissecting scope.

• Acoustics
• Birds
• Physiology

Seminars, Presentations, Posters and Reports

Marine Chemistry

Thecosome pteropods are shell-forming planktonic gastropods, which are sensitive to ocean acidification due to their aragonite shells. This study investigated how continued ocean acidification will impact the geographic distribution, total abundance, species composition and species diversity of oceanic thecosome pteropods in the southeast Pacific.

Pteropods were collected at nine stations along the cruise track with a 333 µm-mesh meter net. At the same locations, hydrographic measurements and seawater samples were collected with a CTD and Niskin bottles, respectively. Pteropods were counted and speciated using a dissecting microscope. Total abundance, species composition and species diversity were compared with latitude, pH, total carbonate alkalinity (TCA), sea surface temperature (SST), dissolved inorganic carbon (DIC), and aragonite saturation state (Ω_aragonite).

Total abundance, species composition and species diversity were not significantly associated with SST, pH, TCA, DIC or Ω_aragonite. Species composition in the same general locale tended to be similar, however was primarily determined by the dominating species. Limacina spp. were the most common pteropods present along the cruise track and often dominated the species composition. Pteropod communities with low abundances typically had high species diversities, and no dominating species. The results of this experiment suggest that although ocean acidification is increasingly becoming a threat to pteropods, it is not currently affecting these organisms in their natural environment.

• Chemistry
• Oceanography
• Zooplankton

Seminars, Presentations, Posters and Reports

Acoustic Scattering from Marine Organisms

Both multi-frequency and broadband acoustic scattering techniques have been used to investigate the nature of a strong pervasive acoustic scattering layer observed during both day and night, at a depth of approximately 20-50 m in the northeast Pacific. Zooplankton composition and size were determined using net sampling techniques, and water properties were determined using conductivity, temperature, and depth sensors. Dominant scatterers have been identified using scattering models for zooplankton. A scattering model for Corolla was also developed to determine their contributions to the total scattering.

Corolla, and elastic-shelled pteropods, which both accounted for a small fraction of the total abundance and biomass, dominated the scattering at low and high frequencies, respectively. Fluid-like siphonophore body parts, which accounted for a small fraction of the total abundance, but a large fraction of the total biomass, dominated the scattering at mid frequencies. Contrarily to previous findings, copepods that accounted for most of the abundance, but little of the biomass, never dominated the scattering. The oceanographic characteristics of the locations analyzed may have created an ideal microenvironment for these organisms to aggregate in layers.

• Acoustics
• Animals
• Oceanography

Seminars, Presentations, Posters and Reports