Lecture 11 - Wildlife Population Surveying: Principles, Methodologies, and Technological Innovations, and Ethical Considerations

Context and Fundamentals of Wildlife Population Surveys

• The lecture is presented as an introduction to measuring populations and measuring population change, serving as a precursor to a guest lecture by postdoctoral scientist Dr. Amanda Lacaccio, who focuses on technological applications in wildlife surveys.

• Determining population sizes through time and comparing populations in different geographic areas is essential for understanding the following ecological factors:

  • The species niche and the importance of habitat variations.

  • Resource levels: Identifying critical thresholds where populations cannot persist below a certain resource level.

  • Carrying capacity of various environments.

  • The impact of specific threats, such as invasive species, climate change, or pollution, which may vary across a species’ distribution.

• Monitoring is vital for conservation management actions, including determining if populations are in decline, tracking the progress of revegetation, or evaluating the success of reintroductions and population supplementations.

Objectives and Justifications for Wildlife Monitoring

• Evaluating Pest Management: A significant amount of resources is spent on controlling pest animals and invasive species. However, monitoring the success of these programs is often neglected.

  • Lack of data regarding the success of control programs leads to poor ecological outcomes.

  • Ethical and controversial actions, such as poisoning or shooting animals to protect other species, require data to maintain "social licence" (public support and acceptance for management interventions).

• Sustainable Harvesting: Surveys are necessary to ensure that the harvesting of animals or plants is sustainable and does not lead to over-exploitation.

• Determining Intervention Success: Monitoring allows practitioners to see if a management action, such as habitat restoration or predator control, has actually affected target populations in the intended way.

Conceptual Frameworks: Precision vs. Accuracy

• Every survey method includes a set of assumptions regarding data collection and usage, which directly impact the reliability of the resulting data. Assumptions range from being extremely strict to being flexible without undermining statistical reliability.

• A critical consideration in wildlife surveying is the distinction between precision and accuracy, often conceptualized using a target analogy:

  • Low Accuracy, Low Precision: The worst-case scenario where estimates are scattered and far from the true value (the bull’s eye).

  • Low Accuracy, High Precision: Estimates are consistent and closely grouped, but they consistently miss the true value. This provides a consistent approximation over time, even if the absolute number is wrong.

  • High Accuracy, Low Precision: Points are near the true value on average but are scattered, making individual surveys less reliable.

  • High Accuracy, High Precision: The ideal scenario where each survey provides a result that is both correct and consistent.

• The importance of accuracy versus precision depends on the specific conservation question:

  • For a critically endangered species, accuracy is paramount. A difference of $20$ individuals between a count of $50$ and $100$ is massive.

  • For large populations, such as $13,000$ or $15,000$ animals, high accuracy is less crucial than precision, which helps track relative trends and ballpark figures ($14,000$, $15,000$, or $16,000$).

Direct Survey Methodologies

• Direct methods involve surveying or capturing the actual animals, allowing researchers to calculate true density (exact numbers per unit area) or indices of abundance.

• Census: Attempting to count every individual in a population. This is difficult for mobile wildlife but easier for immobile organisms:

  • Mobile Wildlife: Counting kangaroos in farmland paddocks in Western Melbourne is a "nightmare" due to their rapid movement over large ranges.

  • Immobile Wildlife: Estimating the density of trees in a defined area (e.g., 500\,\text{meters} \times 500\,\text{meters}) or marking barnacles on a rocky shoreline is straightforward.

  • Case Study Northern Heronized Wombat: A census was possible because the population in Epping Forest dropped to only $35\,\text{individuals}$.

• Advanced Technology - Satellite Imagery: Used to count animals from space without disturbing them:

  • Emperor Penguins in Antarctica: Traditional counting is dangerous and logistically challenging. Researchers used satellite imagery to identify "blobs" representing colonies.

  • Findings: They previously estimated between $270,000$ and $350,000\,\text{penguins}$. Satellite data revealed nearly $600,000\,\text{penguins}$ and identified $7$ additional colonies. This is vital for the conservation of a species threatened by climate change.

Quantitative Capture and Removal Techniques

• Mark-Release-Recapture (Capture-Recapture): A method involving multiple rounds of trapping to solve for $N$ (the total population size).

  • Formula: $N = \frac{M \times n}{m}$ where:

  • $M$ is the total number of animals captured and marked in the first round.

  • $n$ is the total number of animals captured in the second round.

  • $m$ is the number of animals in the second round that were already marked.

  • Example (Kangaroos): If $20$ kangaroos are marked on night one ($M = 20$), and on night two, $30$ kangaroos are caught ($n = 30$) with $20$ of those being marked ($m = 20$), the calculate population is $N = \frac{20 \times 30}{20} = 30$.

  • Assumptions: No births, no deaths, and no immigration or emigration during the trapping period. These assumptions are more likely to be met over short periods (e.g., $2$ to $5$ days).

• Known to be Alive (KTBA): Used when trapability varies over time. If an animal is caught in survey period $1$ and survey period $3$, it must have been alive during period $2$, even if it was not caught then.

  • Example: Surveys from March $6 - 8$, April $12 - 14$, and May $4 - 6$. If $8\,\text{individuals}$ are caught but patterns show $4$ others "must have been alive," the KTBA is $12$.

• Removal and Catch Effort: Often used lethal or non-lethally for pest management (e.g., cane toads into a wetland). The number of captures typically drops each successive day.

  • Day 1: $59\,\text{animals}$.

  • Day 2: $33\,\text{animals}$ (cumulative $92$).

• Line Transects: Used for wide-ranging species like kangaroos. Investigators fly or drive along a line and record the distance of the animal from the line. Detection decreases with distance; the resulting "slope" of sightings is used to derive a population estimate. This prevents "capsulomyopathy" (stress-induced muscle damage/death) by avoiding physical capture.

Indirect Survey Methods: Indices of Abundance

• When absolute density is too costly or unethical, researchers use indices to measure change. Consistency and precision are prioritized over exact numbers.

• Trapping Success: Calculated as the number of animals caught per trap. High success (e.g., $80\,\%$) implies high abundance compared to low success (e.g., $20\,\%$).

• Bioacoustics (Calls):

  • Birds: Difficult due to mimicry (e.g., Lyrebirds mimicking Whipbirds or Cockatoos) and overlapping individual calls.

  • Frogs: Much easier because frog calls are distinct "isolating mechanisms for reproduction." Each species has a unique call.

  • Citizen Science Connection: The Frog ID Project allows the public to record frog calls via a mobile app to build a national dataset on frog dynamics.

• Sign Analysis (Tracks and Scats):

  • Tracks: Using "sand pads" (fine sediment) to record footprints over multiple days. Useful before and after baiting foxes to protect bandicoots. If tracks drop after poisoning, the intervention is deemed successful.

  • Scats: Counting pellets on a transect. Researchers calculate typical "defecation rates" (pellets per species per day) to estimate total individuals from a scat count.

  • EDNA: Collecting genetic material from soil, water, or air to identify species presence and, with enough sampling, genetic signatures of population size.

• Challenges with Sign:

  • Persistence: Rain or wind can erase tracks/scats on sand pads.

  • Detectability: Identifying if many tracks belong to many foxes or one very active fox is difficult.

  • Identification: Distinguishing between similar species, such as fox tracks and small Dingo pup tracks.

Ethics, Risks, and Behavioral Considerations in Sampling

• Trap Shyness vs. Trap Happiness:

  • Trap Shy: Animals avoid traps, potentially due to past negative experiences or instinct (e.g., crocodiles avoiding baited traps in Sri Lanka reservoirs).

  • Trap Happy: Animals enter traps repeatedly for rewards. Possums and bandicoots might follow a researcher down a transect line, entering a second cage immediately after being released from the first for the peanut butter and apple bait.

  • Ethics and Handling: Researchers must minimize stress. Improper handling can lead to:

  • Capsulomyopathy: Muscle damage in highly stressed mammals (kangaroos).

  • Chemical Sensitivity: Frog skin is highly sensitive to humans wearing fly repellent, sunscreen, or perfume.

  • Heat Stress: Birds held incorrectly in mist nets can quickly die from thermal build-up.

• Neophobic Behavior: Some animals fear new objects. Researchers may use "free feeding" (wiring traps open for several nights) to get animals comfortable with the trap before setting it to trigger.

Statistical Integrity and Long-Term Data

• Modeling Trends: Data is often analyzed through simple linear regression (where the slope denotes density), though ecological data often follows polynomials or humped-shaped quadratic relationships.

• Importance of Time Scales: Snapshot surveys can lead to false conclusions. The $50$-year study of Moose and Wolves illustrates this:

  • Short-term surveys might show stable populations, while long-term data reveals precipitous declines and predator-prey linkages.

  • Arid zones in Australia require longer monitoring (over $10\,\text{years}$) due to erratic rainfall cycles compared to forest environments.

• Type II Errors: Failing to detect a change that has actually occurred. This often happens if the population declined before monitoring began, leaving the investigator unaware of the magnitude or the cause (e.g., an $80\,\%$ drop in the past).

• Calibration: Refining indirect methods by running them concurrently with intensive direct methods in the same habitat. Once an index (like scat count) is calibrated against an absolute density (trapping), the cheaper index can be rolled out elsewhere.

Questions & Discussion

• Literature Review Assignment: The lecturer confirms the assignment is "imminently due" and reminds students that extensions must be officially requested. It focuses on one single species (rubric penalization applies for families of species).

• Assessment Roadmap: Following the literature review, the next assessment is a video assignment (group or individual details to follow).

• Drones: Lakshita and Alejandra ask about drones and the pioneer Lianping Co. Drones minimize disturbance if flown at optimum altitudes/noise levels, but waterbirds are still sensitive.

• Genetics and Extinction: Lakshita asks about "genetically edited direwolves" by the company Colossal. The lecturer clarifies these are genetically modified grey wolves, but the scientific achievement provides hope for using genetic tools to support endangered species or "de-extinction."

• Marine Species: Alejandra notes that some marine species change entire appearances or sexes through life cycles, complicating some indices.

• Methodological Revision: The introduction of camera traps over the last $15$ to $20\,\text{years}$ has revealed that traditional trapping (cage/pitfall) often missed certain species entirely, forcing a re-evaluation of historical ecological data.