Global Animal Diel Activity Results
For details on how to use this app, click on the tab 'Start Here'.
Overview
Using a collaborative dataset that combines camera trap projects throughout the world, our aim is to evaluate the diel activity of wild, non-volant mammals by
estimating the diel phenotype (e.g., diurnal, nocturnal, crepuscular, cathemeral) and probabilities of activity (daytime, nighttime, twilight).
Definitions and Hypotheses
We define daytime as the time between the end of morning nautical twilight and the start of evening nautical twilight. We define nighttime as between midnight of that day and the start of morning astronomical twilight or between the end of evening astronomical twilight and midnight of the following day.
Lastly, we define twilight as the combination of dawn or dusk, where dawn is the time between the start of morning astronomical twilight and the end of morning nautical twilight, while dusk is the time between the start of evening nautical twilight and the end of evening
astronomical twilight.
The definitions of diurnal, nocturnal, and crepuscular are genearlly where an animal's probability of activity largely occurs during the daytime, nighttime, and twilight, respectively.
Cathemerality is generally where more than one diel period (daytime, nighttime, twilight) is used to a relatively large degree.
Importantly, there are three fundamental hypothesis sets (Maximizing, Traditional, and General) that results can be chosen from:
Maximizing: Crepuscular Max, Diurnal Max, and Nocturnal Max
Traditional: Crepuscular, Diurnal, Nocturnal, Traditional Cathemeral
General: Crepuscular, Diurnal, Nocturnal, General Cathemeral, Crepuscular-Nocturnal, Diurnal-Nocturnal, and Diurnal-Crepuscular
Quantitative definitions are visualized in the 'Hypothesis Sets' tab, which can be found under the 'Start Here' tab. Additional details can be found in the manuscript outling the methods
and logical of the hypothesis sets:
Unit of Analysis
Our fundamental sampling unit was a 28- or 58-day sampling period for a given project. For each sampling unit, we calculated the independent (>15 min) detection frequencies of each mammal species
during daytime, nighttime, and twilight.
Data Summary
Select Results
Use the left panel to filter results to summarize and plot.
Step 1:
Hypothesis Set: choose a hypothesis set. Details are in the tab 'Hypothesis Sets'.
Step 2:
Filter by: subset the results by:
- Country: Filters results by country, as selected in the 'Select Country' drop-down.
- Specific Project: Filters results by project, as selected in the 'Select Project' drop-down. Project names
are a grouping of the country name, a project indicator name, and the last name of the data contributor contact or public available dataset name.
- Species (across units): Organizes unit-level results by species from all projects and countries.
- Species (aggregated units): Organizes species-level results aggregated across units from all projects and countries for each species. Aggregation refers to the combining of all data for a species together for a single analysis and result, as such there is one result for each species.
Step 3:
Sort species names by: select 'Latin' or 'Common' and then type the species name within 'Type Species Name'. A full list of paired Latin and common names are provided in the 'Species list' tab.
Step 4:
Unit types to include: See the tab '3. Analysis Unit' for details on the three types: '28day', 56day, or'allday'.
Step 5:
P(Hypothesis)>: select the minimum posterior probability of support for a diel hypothesis (i.e., phenotypes). For example, '0' will present all results, while '0.8' will subset the results to only hypotheses that are supported at that amount or greater.
All results are the most supported model, given the data. So, for a given project it might be helpful to select '0' to see every result. However, if you intend to summarize results across analysis units, you should limit the analysis units with the clearest support, so
choosing a lower limit of '0.8' or '0.9'.
Step 6:
Sample size: select the minimum sample size of independent detections for each analysis unit. For example, '10' will include all analysis unit results (since units with < 10 detections were excluded), while '50' will subset the results to only analysis units with 50 or more independent detections.
Step 7:
Analysis Unit: select a specific analysis unit by unique (and arbitrary) numeric identifier. By default, all subsetted analysis units are provided. This is not available for the 'Species (aggregated)' organization.
This selection is helpful when wanting to only plot a single result, rather than a collection of results. The analysis unit is only unique to the Results Table and changes with different data filtering.
Step 8: Note: there is no 'Run' or 'Submit' button. Once all selections have been made, proceed to viewing the results tabs.
Results Summary
Results are summarized in the top-level tabs:
Results Table
This table will have species and project information, and diel phenotype results. The table will be populated with all the species selected for a given project or country; this table can be downloaded using the blue 'Download' button.
Below are descriptions of important columns.
- Hypothesis: The most supported empirically estimated diel phenotype; depends on the hypothesis set.
- P.Hypothesis: The estimated probability of support for the diel phenotype based on Bayes factors.
- Sample Size: The summed number of detections for twilight, daytime, and nightime.
- Latitude/Longitude: The mean lat/long in decimal degrees of all camera locations for an individual analysis unit.
- Ref.Activity: Literature supported diel phenotype from Cox et al. 2021.
- Ref.Act.Flex: An indicator (0 or 1) of diel flexibility from the literature and summarized by Cox et al. 2021.
Cox, D. T. C., Gardner, A. S., & Gaston, K. J. (2021). Diel niche variation in mammals associated with expanded trait space. Nature communications, 12(1), 1753.
Descriptions of each column can be found by downloading this file:
What is an 'Analysis Unit'
Our fundamental sampling unit was a 28- or 58-day sampling period for a given project. For each sampling unit, we calculated the independent (>15 min) detection frequencies of each mammal species during daytime, nighttime, and twilight.
Users can not select how to filter detections by type, but rather filter units based on rules that have already been applied.
For each project,
- We removed detections to enforce a minimum of 15 minutes between detections of the same species at a given camera location.
- We combined camera traps within 150 m of one another to a single site.
For each species and project, we organized detections into three types of analysis units,
- '28day': 28 days of continuous sampling across a camera trap project with at least ten detections.
- '56day': if a '28day' unit was not possible for a species and project then we used 56 days of continuous sampling across a camera trap project with at least ten detections.
- 'allday' if neither a '28day' or '56day' unit was not possible (due to too few detections in these periods) then we selected detections from all days of the camera trap project. This was intended to make inference on the diel activity of hard to detect and rare species.
However, it should be recognized that the time span of the detections could be quite long and thus could confound temporal changes in diel activity.
Our goal was to produce enough detections for rare and common species to make inference to a species diel activity for each project during a short enough time period (i.e., 28 days). If this was not
possible, we expanded to a 56 day period. To provide results on the very rare species that had few detections, we captured these in the 'allday' units.
Hypothesis Sets
There are three hypothesis sets used to estimate the diel phenotype of each analysis unit: Maximizing, Traditional, and General.
Each set is defined for a different objective and are specified by the combined probabilities of activity during twilight, daytime, and nighttime. For full details
read the manuscript descirbing the logic and results of these (and more),
Maximizing
This hypothesis set includes three hypotheses (Crepuscular Max, Diurnal Max, and Nocturnal Max) with the objective to evaluate which diel period (twilight, daytime, nighttime) is used most.
As such, there is no hypothesis about activity across multiple time periods (i.e., cathemeral).
Species example: the Coyote (Canis latrans) is estimated to be active during twilight, daytime, and nightime at probabilities of 0.20, 0.30, and 0.50, respectively. Since the nighttime probability is used the most, this data would support the 'Nocturnal Maximization' phenotype hypothesis.
Traditional
This hypothesis set includes four hypotheses (Crepuscular, Diurnal, Nocturnal, Traditional Cathemeral) that aim to capture the general interpretation of these hypotheses from the literature. Crepuscular, Diurnal, and Nocturnal are defined based on having at
least 0.80 probability in their respective diel periods (twilight, daytime, nighttime). If an animal is not mostly active in one period than it is defined as Traditional Cathemeral; this occurs when either two or three diel periods are used more than
1-0.80 = 0.20. The logic behind the threshold of 0.80 is that an animal uses a clear majority of time in one diel period, but is not so strict that there is not some moderate amount of activity outside of this period.
Species example: the Gray Fox (Urocyon cinereoargenteus) is estimated to be active during twilight, daytime, and nighttime at probabilities of 0.20, 0.80, and 0.00, respectively.
Under the 'Traditional' hypothesis set, the Diurnal phenotype hypothesis would be supported because of the high probability of activity during the daytime.
General
This hypothesis set includes seven hypotheses. The Diurnal, Crepuscular, and Nocturnal hypotheses are defined the same as in Traditional. The main difference is the separation of the probability space of
Traditional Cathemeral into four more specific hypotheses: General Cathemeral, Crepuscular-Nocturnal, Diurnal-Nocturnal, and Diurnal-Crepuscular. The General Cathemeral hypothesis --- which represents a subset of the parameter space
taken by the previously mentioned Traditional Cathemeral --- aims to define when an animal uses all three diel periods (twilight, daytime, nighttime) at equal to or more than a minimum amount (i.e., 0.80) and greater than or equal to lower amount (i.e., 0.10),
such that we consider it important to differentiate animal activity when a diel period is used at least this much. However, if only two diel periods are used above 0.10 then we classify this activity using one of the binomial hypotheses
(Crepuscular-Nocturnal, Diurnal-Nocturnal, and Diurnal-Crepuscular).
Species example: the Gray Fox (Urocyon cinereoargenteus) is estimated to be active during twilight, daytime, and nighttime at probabilities of 0.20, 0.80, and 0.00, respectively.
Under the 'General' hypothesis set, the Diurnal-Crepuscular phenotype hypothesis would be supported because of the high probability of activity during the daytime and twilight.
Private Data Contacts
Public Data Sources
This table summarizes the 'Results Table' by frequency of the supported hypotheses for a single species across all selected analysis units by country and season. Thus, the frequency number corresponds to the number of analysis units in the 'Results Table' for a given hypothesis, Country, and Season.
A plot of a single species' probability of activity in the twilight, daytime, and nighttime from the selected analysis unit's . The shaded space represents the literature reference category, as interpreted by our defintions of Traditional diel phenotypes. Probabilities outside of these shaded areas suggests that the species uses a different diel phenotype for the given analysis unit.
Note, this plot can not be manipulated.
Hover over the colored circle to display the median probability of activity, as well as relevant data information of the Analysis unit - Project Code - Season - Begining Sampling Date. In the top right corner of the plot area there are additional options, including to download the plot.
If more than 10 analysis units are selected, the legend is not displayed. For 10 or less analysis units, the legend labels correspond to the first column of the 'Results Table'.
The rectangular shading in the plot indicates the probability space of the Reference diel phenotype. Estimated probabilities outside of this space indicates empirical evidence that an analysis unit does not agree with the reference diel phenotype.
Description
Please be patient. The plot may take a moment to load.
A 3d plot of a single species' selected analysis units probability of activity in the twilight, daytime, and nighttime.
When more than 30 analysis units are selected, only the posterior medians (most probable value) of the probabilities are plotted and are colored
based on the supported diel phenotype. If less than 30 analysis units are selected, the posterior medians (most probable value; dark grey dots) are plotted with all posterior samples, representing uncertainty around the most probable value. If more than
30 analysis units are selected, only the posterior medians are plotted.
Note, this plot can be moved around the page and in 3-dimensions: 1) right click using the mouse on the plot to move the plot around the page, 2) left click using the mouse on the plot to turn the plot for a different angle.
In the top right corner of the plot area there are additional options, including to download the plot. 3) Hover over the posterior samples to display the probability of activity of twilight, daytime, and nighttime, as well as
relevant data information of the Analysis unit - Project Code - Season - Begining Sampling Date.
.