Cam trapping: ten simple steps to process all those amazing photos

Workflow

  1. Read some cam trap papers.
  2. Check camtrapR package and see what it does to decide if it suits your specific needs.
  3. Open folder for each site, each day, each rep.  Do a folder ‘get info’ to count # of total pics.  These are your ‘reps’ within reps, i.e. literally total number of snapshots (or use command line to get dir info for all your photo data).
  4. I would honestly just paste 0’s all the way down because many will be ‘false hits’.
  5. Then, open them all up and scroll through.
  6. Every time a positive hit, overwrite 0 in ‘animal.capture’ vector and in ‘animal’ vector record what it is.
  7. Also, copy all positive hits photos into a separate folder for additional analyses.  Use a folder structure or ID system that keeps track of the place and time that photo was from.  For instance, have a folder entitle positive-hits for each site, day, location, rep or aggregate into a single positive hit folder but use a mechanism to ensure we know where/when photo was taken. Do not cut and paste, copy. This is a backup mechanism for additional analyses and sharing data.
  8. We also want to know when animals are most active, or not; and hence, check timestamps and paste down in that column too. Ideal is actual time but morning, afternoon, night is absolutely adequate and more rapid if we cannot automate the scraping using R-package.
  9. If timestamp is incorrect, do a light-dark assessment to code as night or day – this is a very rapid process.
  10. Record observations if more than one animal or if the same animal was recaptured from previous instance. Record anything of note ecologically to calibrate the quantitatives and link photo-capture processing to data mapping/translation. The goal is to accurately map photos onto numbers that represent the dynamics of the system in study.

Outcomes

The goal is have to have both an evidence folder of positive hits and a dataframe that can then be wrangled to estimate relative efficacy of sampling, frequencies of different animals, spatiotemporal dynamics, and differences between structured treatments in the implementation of trapping.

Meta-data for manual processing spreadsheet workflow

Attribute is the column headers.

attribute description
year we have many years for Carrizo (evil laugh) so good to list here in a vector
region MNP for Mojave, CNM for Carrizo
site if you have more than one site, put name of site
calendar date dd-m-year
microsite larrea, buchhorn, ephedra or open depending on region
day this is census day, 1,2,3, to however many days sampled
rep if more than one rep per day
photo rep just cut and paste to total number of photos each cam took on one day, could be 10 to 10000
animal.capture binary 0 = false hit, 1 = animal present
animal list animal as ‘none’ if false hit, then animal name if one was there
timeblock with animal telemetry work, morning, afternoon, night is usually sufficient 6am to noon, noon to 6pm, them night time
night.day back up if timestamps are incorrect – just do by night and day using light and darkness in photos. very quick
filename.ID.positive.hits
optional depending on your filing system, copy all positive hits to a separate folder. somehow, keep track of positive locations and times for subsequent analyses
observations
record observations of anything ecological that pops such as if there was another animal in the photo OR if it was the same animal repeatedly recaptured

CSEE 2017 Highlights

This year the ecoblender lab attended CSEE 2017. The conference was great and covered four days of talks, workshops, and networking events. I attended a free workshop that taught some basics in mapping spatial data and different packages to use in R. There was also a wide range of talks that mostly seemed interdisciplinary. This included discussions of uncertainty in ecology, estimate the value of natural resources, and developing models of habitat selection. Here are some of the highlights I took away from the conference:

Modelling:

There was discussion over the usage and power of mechanistic vs. phenomenological models. This is a topic discussed often in ecology (see of that discourse here), but can be defined here as:

mechanistic: includes a process (physical, biological, chemical, etc) that can be predicted and described.

phenomenological: Is a correlative model that describes trends in associated data but not the mechanism linking them.

The discussion mostly described the relationship between phenomenological and Mechanistic models as not binary and rather a gradient of different models that describe varying amounts of a particular system. However, it did touch upon models such as GARP and MaxEnt that are often used for habitat selection or SDM but neglect the mechanism that is driving species occurrence. Two techniques I would like to learn more about are Line Search MCMC and HMSC which is a newly developed method for conducting joint species distribution models.

Camera traps:

There was also a morning session that described benefits and tools for using camera traps. These sessions are always great as they give a chance to see some wildlife without disturbance. Topics focus around deer over abundance harming caribou populations, how wildlife bridges do not increase predation through the Prey-Trap Hypothesis and techniques for using wildlife cameras or drones. One talk that was particularly interested used call back messages when triggered to see how animals respond to noises such as human’s talking or a mating call.

One of the more useful things I believe to have taken out of the session is how to estimate animal abundance and movement when the animals in your camera traps are unmarked. One modelling technique using Bayesian modelling and was found to be equivalent to genetic surveys of animal fur for estimating animal abundance. This is in contrast to the more frequent spatial capture-recapture (SCR) methods that either mark individuals or supplement camera trap data with other surveys. I also discovered there the eMammal project at the Smithsonian that is an Open Access project for the management and storage of camera trap data.

Ecology and climate change:

Climate change as always is a big topic at these conferences. There was a good meta-analysis out of the Vellend lab that show artificial warming of plant communities does not result in significant species loss. However, there was evidence that changes in precipitation does significant impact plant communities. The results are very preliminary, but I look forward to seeing more about it in the future. I also liked a talk that is now a paper in Nature that models networks in the context of climate change. The punchline of the results being that species composition in communities is dependent on dispersal, and high dispersal rates can maintain network structure although members of the community may change.

I presented results from our upcoming paper modelling positive interactions in desert ecosystems:

Overall I learned a lot from the CSEE 2017 conference and thought it was a health balance of size and events. Victoria was also a great city and made hosting the conference very easy. Next year it will in the GTA and I plan on connecting with the organization committee to potentially host an R workshop at the beginning of the conference. Until then!

Microenvironmental change in Cuyama Valley 2017.2 goals

In 2016, we deployed micro-environmental data logger arrays to monitor global change dynamics at very fine scales. We also structured measurements to ensure we can infer and link to a biotic interaction signal between common plants within this region.

This is very important region to study for at least two reasons ecologically.

(1) Water issues with people, plants, and agricultural are critical here.

(2) Cuyama Valley is an excellent set of sites or mesocosm for the San Joaquin Valley at large. The San Joaquin Valley is still sinking (NASA report). We need to understand temperature, precipitation, and soil moisture availability patterns at many scales within the region.

This season, 2017, is a relative boom year in terms of precipitation. Here are the immediate sampling goals for this season.

1. April (mid). At peak flowering, count burrows, re-measure shrub sizes, sample annual vegetation, and collect biomass.
2. May (mid). Retrieve all logger units, download data, and check functionality. These data capture two growing seasons – one drought, one wet.
3. May (mid). Re-deploy and re-initialize loggers. Rationale – need data on shrub effects when it matters for animals like lizards and hoppers etc and when it is really hot.
4. Sept (end). Retrieve loggers and sensors, download data, end experiment.
5. Oct. Design and test a missing-data strategy to address missing sensor and logger failures. I will likely implement a within-site, resampling data strategy associated with central tendency measures to fill gaps.

UTM to longitude latitude R-code

UTM-long.lat conversions

Option 1. Work with original dataframe that has locations as UTM

library(rgdal)

utms <- SpatialPoints(data[, c(“long”, “lat”)], proj4string=CRS(“+proj=utm +zone=10”)) #create UTM matrix

longlats <- spTransform(utms, CRS(“+proj=longlat”)) #transform

Option 2. Generate a new dataframe and use coordinates function instead (preserves other vectors in dataframe)

#convert UTM to long.lat

mapdata <- data

coordinates(mapdata) <- ~long+lat #similar to SpatialPoints

proj4string(mapdata) <- CRS(“+proj=utm +zone=10”) #assign projection and coordinate reference system

longlats <- spTransform(mapdata, CRS(“+proj=longlat”)) #transform

Waiting for the rain

Rainfall updates

The growing season of 2015-2016 has come and gone with disappointing results from the supposed El Niño year. The 2016-2017 season is approaching and a few had feared that it would just continue the current pattern of drought. I was especially fearful having battling drought four years in a row in my study of plant interactions. It would be nice to have a chance with at least “average” precipitation amounts. Half way through the rain season and this year looks promising. Areas of California have been seeing some pretty significant precipitation including some potential floods. While this is great news in terms of drought relief for coastal cities and the Sierra Nevada snow pack, I wonder what the consequences will be for the deserts? In particular, the Mojave always seems to be in the unluckiest of rain shadows, missing most of the precipitation that the rest of the state experiences. I took a snap shot of the rainfall and average temperatures since seeding at the end of October. Here are the results:

Interpretation

The right combination of rain, temperature, and timing are absolutely crucial in desert ecosystems in regards to how the plant composition will respond. In an older paper by Beatley (1974) is a description of how these three variables determine plant composition. From this and my own experience, the absolute minimum rain to see any annual vegetation on the ground is 2.5 cm. However, these plants usually die within a month if there is no subsequent rain. I have seen this occur in multiple years where Halloween rain is not followed by any other precipitation until mid-January. The result? Many dead plants, and a new representation for plant communities. The Mojave has seen enough rain to begin germination and this rain has all occurred within the last 3 weeks. This, plus continued cold temperatures, should encourage the persistence of annuals for at least another month. If at least one other major rain storm passes through in that time I would expect to see these plants make it to flowering. On the more westerly side of the state, my sites have been seeing fairly consistent rain. This is great news for my Panoche Hills site that likely has passed its precipitation threshold that guarantees emerged plants to flowering.

Fingers crossed as always!

Journals for synthesis

Colleagues and I were checking through current journal listings that either explicitly focus on synthesis such as systematic reviews or include a section that is frequently well represented with synthesis contributions. Most journals in ecology, evolution, and environmental science that publish primary standard, research articles nonetheless also offer the opportunity for these papers too, but it can be less frequent or sometimes less likely to accept different forms of synthesis (i.e. systematic reviews in particular versus meta-analyses).

List

Diverse synthesis contributions very frequent
Conservation Letters (Letters)
Perspectives in Science
Perspectives in Plant Ecology, Evolution and Systematics
Diversity & Distributions
Ecology Letters
TREE
Oikos
Biological Reviews
Annual review of ecology, evolution, systematics
Letters to Nature
Frontiers in Ecology and the Environment
PLOS ONE (many systematic reviews)
Environmental Evidence
Biology Letters
Quarterly Review of Biology

Frequent synthesis contributions with some diversity in formats
Global Ecology and Biogeography
Annals of Botany
New Phytologist
Ecography
Ecological Applications
Functional Ecology
Proceedings of the the Royal Society B
Ecology and Evolution

Progress Report – Fall 2016

Several weeks ago I completed my first progress report for my MSc program. This involved a giving a short presentation (slides above) followed by a question/discussion period. My thesis focuses on pollination facilitation – non-competitive pollinator sharing between plant species that improves the reproductive success of at least one of the participants. I will be investigating these interactions in the Mojave Desert, a biodiversity hotspot supporting 659 species of bees and 1680 annual plants.

Why spatial? The study of ecology is normally separated into hierarchies, however, we know that these different levels are integrated and interact despite studying them in isolation. All interactions take place in space, and so explicitly including spatial dimensions to a study can be a way of connecting these levels, leading to a deeper understanding of the observed interactions.

It can be a little intimidating to stand in front of your committee and tell them your ideas, but they are there to support you. I received some great feedback which I am using retool my experimental design in preparation of the upcoming field season. Advice: Be careful about your clipart choices! I used a picture of queen honeybee (they don’t pollinate!!) in an interaction diagram explaining pollination facilitation. This isn’t as bad as the infamous biology textbook “Bees of the World” showing a pollinating fly on the cover, but it was noticed right away.

ESA 2016 – my highlights

Functional traits

I have known for a while that functional traits are at the core of ecology. Unfortunately, for most of my experiments I focus solely on biomass and abundance. However, at ESA this year I noticed many researchers who measured community abundance and used a secondary database such as TRY for extracting functional traits per plots. This is a great re-purposing of already collected data to be reused to answer different questions. I am definitely going to consider analyzing data sets I currently have the species composition for to answer questions about functional diversity rather than typical species richness.

functional traits

Statistics

There was a considerable amount of NMDS usage this year, furthering that ecological analyses are becoming more complicated and requiring ordinations. Although ecologists may love NMDS, there is a preference in the statistical world to avoid it. I won’t get into the pros and cons here, but I believe ordinations such as PCoA, CA and RDA can accomplish the same as NMDS with less limitations. I will need to explore this further but I found this document a good starting point. Permuted ANOVAs were also another particularly popular statistical test and one I need to find a good R package for.

Another great talk from the conference discussed the advantages of the Negative Binomial distributions particularly for species abundance distributions. The talk discussed the tendency to fit discrete count data into log-normal distributions or Poisson. Often these distributions do not fit the data, while negative binomial does. The author A. Rominger promised me that he would be publishing an article detailing this commonality later in the year. I will hold him to that! In the mean time, here is an R package that he developed for species abundance distributions and the negative binomial distribution called pika.negative binomial distribution

Species distribution modeling

I had the opportunity to sit in on a talk by C. Merow that discussed used “expert maps” to further refined predictions of species distributions. It was a great talk and introduced a new concept to refine species predictions. Using the Map of Life (MOL), ecologists create species boundary maps that are would go into an SDM. The maps are compatible with MaxEnt and were shown to be effective at better predicting species occurrence. While a much better technique, I think the major limitation is that specific species will not have these maps delineated yet. However, until MOL catches up, big picture questions can be addressed with a greater degree of accuracy. The talk also reminded me I need to learn more Point Process Poisson modeling

The other instance SDMs that I thought was extremely informative was by Leung discussing co-occurrence models for invasion. This models have strong similarities with our work except instead of testing how an invasive species co-occurs with native species, we are trying to determine how a facilitating benefactor species interacts with neighbouring species. The way this is done is either using a proxy measurement or by using multi-species models. The example Leung used for a proxy measurement was boat movement in Ontario as a function of invasibility for a particular non-native species. I though this was a great idea and something I will consider for my next SDM experiment. The multi-species models is something that was a bit more complicated and that I will need to explore further.

Species distributions

General

Overall ESA went well! The attendance appeared a bit lower than last year, but I felt I was still super busy. It was a great experience and I learned a lot more that the highlights I am listing here. We also received some great feedback on micronet on how to improve it and I go some ideas to develop an R package that hopefully satisfies everyone’s micro-environmental challenges. A full list of my participation at ESA below:

A test of the stress-gradient hypothesis including both abiotic stress and consumer pressure during an extreme drought year

How to Set Up Automated Sensors Arrays for Measuring Micro-Environmental Characteristics and Synthesize to Larger Scales

Microenvironmental change as a mechanism to study global change

The use of shrubs as a tool for re-establishing native annuals to an invaded arid shrub land

 

esa2016

Precipitation mediates the mechanism of facilitation in a Californian Desert

ESA 101 at Fort Lauderdale is coming up! I will be presenting on recent findings from an experiment we conducted over two years. I am extremely excited for both the presentation and the results! Here is the slide deck, statistical analyses and program outline.

https://afilazzola.github.io/water.consumer/

Background/Question/Methods
The stress gradient hypothesis original purposed the frequency of plant interactions along countervailing gradients of abiotic stress and consumer pressure. However, research to date has studied these two stressors in isolation rather than together, thereby potentially neglecting the interaction of these factors on plant composition. In the arid central valley of California, we artificially manipulated a soil moisture gradient and erected animal exclosures to examine the interactions between dominant shrubs and the subordinate annual community. We conducted this experiment in an extreme drought year (2014) and a year of above-average rainfall (2016).

Results/Conclusions

Shrubs positive affected the abundance and biomass of the annual community at all levels of soil moisture and consumer pressure. In the drought year, shrub facilitation and water addition produced similar positive effect sizes on plant communities; however, the shrub facilitation effect was significantly greater in watered plots. During the year with higher rainfall, there was no observed water or exclosure effect, but shrubs still significantly increased biomass of the subordinate plants. Shrubs and positive interactions maintain productivity of annual plant communities at environmental extremes despite reductions in droughts stress or consumer pressure and these positive effects are even more pronounced with water addition. The relationship between consumer pressure and abiotic stress on plant interactions is non-linear particularly since shrubs can facilitate understorey plants through a series of different mechanisms.