HMSC
Hierarchical Modelling of Species Communities (HMSC) is a model-based approach for analyzing community ecological data (Ovaskainen et a.2017a).

The obligatory data for HMSC-analyses includes a matrix of species occurrences or abundances and a matrix of environmental covariates. Additional and optional data include information about species traits and phylogenetic relationships, and information about the spatiotemporal context of the sampling design. HMSC partitions variation in species occurrences to components that relate to environmental filtering, species interactions, and random processes. HMSC yields inference both at species and community levels. It can be used to generate simulated communities under given environmental conditions, and thus its predictions can be compared to independent validation data, and it can be used for scenario simulations.

Versions

The current version of the R-package Hmsc is found on CRAN. For a description of the software, see https://doi.org/10.1111/2041-210X.13345

The development version is found on Github.

The old matlab and R versions of HMSC can be found here.

HMSC book

The first copies of the book now exist, fresh from the printing press!

The official publication date (for both the printed book and the ebook) for the UK and Europe has been set as 11th June 2020. Further global publication dates will depend on shipping schedules and will be available on www.cambridge.org.

Click here to order from the publisher's site - use offer code JSDM2020 to get 20% off.

Executable scripts

As supporting information to the book, we provide here R-scripts and data files for the three real data examples of the book: the plant example of Section 6.7, the fungal example of Section 7.9, and the bird example of Section 11.1. The scripts were originally developed by Otso Ovaskainen and Nerea Abrego, and Jari Oksanen has made further improvements to them.

 

Forthcoming HMSC courses

We will organize the next joint species distribution course 15-19 August 2022 in the context of the Jyväskylä Summer School (https://www.jyu.fi/en/research/summer-and-winter-schools/jss). This will be a hybrid event, so you can take part either physically (in Jyväskylä, Finland), or through online zoom sessions. The main aim of the course is to teach the participants how to apply joint species distribution modelling in practice with the R-package Hmsc. The course will be a mixture of basic and advanced topics, so it is suitable both as the first course on this topic, as well as giving more profound understanding for those who have already taken a previous course on this topic.

The deadline for officially registering for the course was April 30th. If you plan to participate only remotely, you can register here (google form) until August 1st. 

For the course description, see here and the text below.

BIO3: Joint Species Distribution Modelling with Hierarchical Modelling of Species Communities (HMSC)
Time: 15-19.8.2022
Study mode: Hybrid
Participants: Master students, PhD students, post docs, senior researchers
Lecturer(s): Otso Ovaskainen, Nerea Abrego, Mirkka Jones, Gleb Tikhonov, Jari Oksanen, Sara Taskinen, Jenni Niku, and others
Coordinator(s): Otso Ovaskainen and Nerea Abrego
Code: BENS7007
Modes of study: Lectures, computer demonstrations, computer exercises, self-reading of course book, small-group discussions
Course book: Ovaskainen, O. and Abrego, N. 2020. Joint Species Distribution Modelling – With Applications in R. Cambridge University Press.
Credits: 3 ECTS
Evaluation: Pass / fail, based on project work report

Contents: This course is designed for students and researchers who are interested in analysing community data in a way that allows placing their results in the context of modern theory. The course follows the book of Ovaskainen and Abrego (2020), thus covering a comprehensive treatment of Hierarchical Modelling of Species Communities (HMSC), including both the technical detail of the statistical methods, as well as the ecological interpretation of the results. With the help of worked out examples, the participants learn how to conduct and interpret statistical analyses in practice with the R-package Hmsc, providing a fast starting point for applying HMSC to their own data. The participants are encouraged to bring also their own data so that they can get hands-on support on HMSC-analyses of their own projects.

Learning outcomes: The participants will learn how to set up R-scripts that run the entire Hmsc pipeline from constructing and fitting the models to producing the result tables and figures, as well as how to write texts for how the analyses were done (for the Material and Methods section for your manuscript) and how to report the results (the Results section of your manuscript).

Prerequisites: Some prior experience on the use of linear models, and some prior experience on the use of the R-environment

Past HMSC courses

Europe, Finland, Helsinki, and online. Five-day-course 2-6 November. Teachers Otso Ovaskainen, Nerea Abrego, Jari Oksanen, and others: Hierarchical Modelling of Species Communities with the R-package Hmsc:

  • NOTE: DUE TO COVID, THE ENTIRE COURSE WILL BE ARRANGE REMOTELY WITH ZOOM. ALL LECTURES WILL BE RECORDED, AND BREAK-OUT GROUPS WILL BE ARRANGED IN EUROPEAN, AMERICAN, AND AUSTRALASIAN TIME ZONES.
  • Teachers: Otso Ovaskainen, Nerea Abrego, Mirkka Jones, Jari Oksanen, Øystein Opedal, Gleb Tikhonov & Bess Hardwick.
  • Course dates: November 2nd-6th 2020
  • Venue: zoom
  • Course Overview: This course is designed for students and researchers who are interested in analysing data on community ecology in a way that allows placing their results in the context of modern theory. The course covers a comprehensive treatment of HMSC, including both the technical detail of the statistical methods, as well as the ecological interpretation of the results. With the help of worked out examples, the participants learn how to conduct and interpret statistical analyses in practice with the R-package Hmsc, providing a fast starting point for applying HMSC to their own data. The participants are encouraged to bring also their own data so that they can get hands-on support on HMSC-analyses of their own projects. Ideally, you will come back from the course with R-scripts that run the entire Hmsc pipeline from constructing and fitting the models to producing the result tables and figures, as well as draft texts for how the analyses were done (for the Material and Methods section for your manuscript) and what the results are (for the Results section of your manuscript).
  • Course timing: how do we manage global participation? We have 123 registrations from Europe+Africa, 77 from America, and 39 from Australasia. All times in this document are GMT+0. The baseline course will run every day from 9am to 3pm, with one hour allocated to breaks, so 5 hours of active teaching for each of the five days. All lectures and computer demonstrations will be recorded, so those for whom these times are difficult can watch the recorded versions. Break-out groups will be organized also in the evening (GMT 4pm-6pm; should be convenient for America) and in the morning (GMT 6am-8am; should be convenient for Australasia). The entire event will happen in zoom. Questions in plenary session are to be asked primarily by chat.
  • COURSE PROGRAMME:
    • All times in this document are GMT+0.
    • Zoom link for lectures: https://helsinki.zoom.us/j/61939913297? (You should have received the password by email, contact us if you did not receive it)
    • We will take a one-hour break daily 11:30 AM to 12:30 PM, with additional short breaks between lectures
    • Lecture slides will be here:
    •  
    • Monday 2nd November: 9am-3pm
    • Tuesday 3rd November: 9am-3pm
    • Wednesday 4th November: 9am-3pm (+4pm-6pm)
      • R-demonstration 3. Bird case study. skipped
      • R-demonstration 4. Selected case studies from the data provided by the participants. https://youtu.be/slP4BCbiriA
      • Break-out groups 1 - Europe (1pm-3pm) Possibility to discuss with teachers in small groups in any topics of your interest. In particular, demonstrations of what was done/can be done with your own data, especially for those who submitted pilot data. Links to break-out groups are in the emailed word document.
      • Break-out groups 1 – America (4pm-6pm)
        • Links to break-out groups are in the emailed word document.
    • Thursday 5th November: 9am-3pm (+6am-8am and 4pm-6pm)
      • Break-out groups 1 – Australasia (6am-8am)
        • Links to break-out groups are in the emailed word document
      • R-demonstration 5. Selected case studies from the data provided by the participants. https://youtu.be/Q_m3q6Y2Jho
      • Lecture 8. How to describe the HMSC analyses in the “Material and Methods” section of your manuscript (based on R-demonstrations 4 and 5). https://youtu.be/qPPDb6TMb48
      • Lecture 9. How to describe the HMSC analyses in the “Results” section of your manuscript (based on R-demonstrations 4 and 5). https://youtu.be/nuv3UBzbOP4
      • Break-out groups 2. - Europe (1pm-3pm) Possibility to discuss with teachers in smaller groups in any topics of your interest. In particular, demonstrations of what was done/can be done with your own data, especially for those who submitted pilot data.
      • Break-out groups 2 – America (4pm-6pm)
    • Friday 6th November: 9am-3pm (+6am-8am)
  • Registration is now closed. Participants will be prioritized according to registration order and relevance.
  • You can receive a signed certificate stating that you have successfully completed this 2 ECTS course - hopefully your university will then agree to register the credits.
    • To achieve the study credits, you should turn in a learning diary at the end of the course (2 5 pages in total, pdf file), where you summarize in your own words what you learned during each of the course days. Send the pdf file to hmsc-course@helsinki.fi, with email subject learning_diary_surname_firstname , e.g. learning_diary_ovaskainen_otso
    • If you are a student at Helsinki, we can register your credits directly.
  • The main challenge will be that we wish to provide one-on-one guidance on how to analyze your own data, for those participants who plan to come with their own data. As the main tool to achieve this, we will check already before the course that your data are technically ready for HMSC-analyses, including fitting a pilot HMSC-model to your data. This will ensure that during the course we can focus on the actual science, rather than wondering e.g. why a specific error message appears.
  • Thus, if you wish for support with analyzing your own data with HMSC, send your data after registering! As the first step, read carefully through INSTRUCTIONS.txt (found from the zip-file below) which explains in which format and where you should send your data. The other files in the zip-folder are templates that you should follow when preparing your own data. We will start setting up the pilot HMSC-models in the order that we receive the data, so even if there is no deadline of when you should send your data, we recommend you to send it as soon as possible to ensure that we will have the time to process it before the course.
  • If you have any questions, email to hmsc-course (at) helsinki.fi with email subject "question_surname_firstname", as explained in more detail in INSTRUCTIONS.txt.

References

Core papers about HMSC

Ovaskainen, O., Tikhonov, G., Norberg, A., Blanchet, F. G., Duan, L., Dunson, D., Roslin, T. and Abrego, N. 2017a. How to make more out of community data? A conceptual framework and its implementation as models and software. Ecology Letters 20, 561-576. https://doi.org/10.1111/ele.12757

Ovaskainen, O., Tikhonov, G., Dunson, D., Grøtan, V., Engen, S., Sæther, B.-E. and Abrego, N. 2017b. How are species interactions structured in species rich communities? A new method for analysing time-series data. Proceedings of the Royal Society B: Biological Sciences, 284, 20170768. https://doi.org/10.1098/rspb.2017.0768

Tikhonov, G., Abrego, N., Dunson, D. and Ovaskainen, O. 2017. Using joint species distribution models for evaluating how species-to-species associations depend on the environmental context. Methods in Ecology and Evolution 8, 443-452. https://doi.org/10.1111/2041-210X.12723

Tikhonov, G., Opedal, Ø. H., Abrego, N., Lehikoinen, A., de Jonge, M. M., Oksanen, J. and Ovaskainen, O. 2020a. Joint species distribution modelling with the R-package Hmsc. Methods in Ecology and Evolution, in early view, https://doi.org/10.1111/2041-210X.13345

Tikhonov, G., Duan, L., Abrego, N., Newell, G., White, M., Dunson, D. and Ovaskainen, O. 2020b. Computationally efficient joint species distribution modelling of big spatial data. Ecology, e02929. https://doi.org/10.1002/ecy.2929

Ovaskainen, O. and Abrego, N. 2020. Joint Species Distribution Modelling – With Applications in R. Cambridge University Press, in press.

Abrego, N., Norberg, A. and Ovaskainen, O. 2017. Measuring and predicting the influence of traits on the assembly processes of wood-inhabiting fungi. Journal of Ecology 105, 1070-1081. https://doi.org/10.1111/1365-2745.12722

Ovaskainen, O., Roy, D., Fox, R. and Anderson, B. 2016. Uncovering hidden spatial structure in species communities with spatially explicit joint species distribution models. Methods in Ecology and Evolution 7, 428-436. https://doi.org/10.1111/2041-210X.12502

Ovaskainen, O., Abrego, N., Halme, P. and Dunson, D. 2016. Using latent variable models to identify large networks of species-to-species associations at different spatial scales. Methods in Ecology and Evolution 7, 549-555. https://doi.org/10.1111/2041-210X.12501

A review of joint species distribution modelling:

Warton, D., Blanchet, G., O’Hara, R., Ovaskainen, O., Taskinen, S., Walker, S. and Hui, F. K. C. 2015. So many variables: joint modelling in community ecology. Trends in Ecology and Evolution 30, 766-779. https://doi.org/10.1016/j.tree.2015.09.007

How HMSC results link to the underlying community assembly processes:

Ovaskainen, O., Rybicki, J. and Abrego, N. 2019. What can observational data reveal about metacommunity processes? Ecography 42, 1877-1886. https://doi.org/10.1111/ecog.04444

The predictive performance of HMSC in comparison to other SDM methods:

Norberg, A., Abrego, N., Blanchet, F. G., Adler, F., Anderson, B., Anttila, J., Araújo, M., Dallas, T., Dunson, D., Elith, J., Foster, S., Fox, R., Franklin, J., Godsoe, W., Guisan, A., O'Hara, B., Hill, N., Holt, R. D., Hui, F., Husby, M., Kålås, J., Lehikoinen, A., Luoto, M., Mod, H., Newell, G., Renner, I., Roslin, T., Soininen, J., Thuiller, W., Vanhatalo, J., Warton, D., White, M., Zimmermann, N., Gravel, D. and Ovaskainen, O. 2019. A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levels. Ecological Monographs 89, e01370. https://doi.org/10.1002/ecm.1370

Some empirically oriented papers applying HMSC:

Norberg, A., Halme, P., Kotiaho, J. S., Toivanen, T. and Ovaskainen, O. 2019. Experimentally induced community assembly of polypores reveals the importance of both environmental filtering and assembly history. Fungal Ecology 41, 137-146. https://doi.org/10.1016/j.funeco.2019.05.003

Minard, G., Tikhonov, G., Ovaskainen, O. and Saastamoinen, M. 2019. The microbiome of the Melitaea cinxia butterfly shows marked variation but is only little explained by the traits of the butterfly or its hostplant. Environmental Microbiology 21, 4253-4269. https://doi.org/10.1111/1462-2920.14786

Rocha, R., Ovaskainen, O., López-Baucells, A., Farneda, F. Z., Sampaio, E. M., Bobrowiec, P.E.D, Cabeza, M., Palmeirim, J. M. and Meyer, C. F. J. 2018. Secondary forest regeneration benefits old-growth specialist bats in a fragmented tropical landscape. Scientific Reports 8, 3819. https://doi.org/10.1038/s41598-018-21999-2

Häkkilä, M., Abrego, N., Ovaskainen, O. and Mönkkönen, M. 2018. Habitat quality is more important than matrix quality for bird communities in protected areas. Ecology and Evolution 8, 4019-4030. https://doi.org/10.1002/ece3.3923

Lammel, D. R., Barth, G., Ovaskainen, O., Cruz, L. M., Zanatta, J. A., Ryo, M., Souza, E. M. and Pedrosa, F. O. 2018. Direct and indirect effects of a pH gradient bring insights into the mechanisms driving prokaryotic community structures. Microbiome 6, 106. https://doi.org/10.1186/s40168-018-0482-8

Abrego, N., Dunson, D., Halme, P., Salcedo, I. and Ovaskainen, O. 2017. Wood-inhabiting fungi with tight associations with other species have declined as a response to forest management. Oikos 126, 269-275. https://doi.org/10.1111/oik.03674

Rocha, R., Ovaskainen, O., López-Baucells, A., Farneda, F. Z., Ferreira, D. F., Bobrowiec, P.E.D, Cabeza, M., Palmeirim, J. M. and Meyer, C. F. J. 2017. Design matters: an evaluation of the impact of small man-made forest clearings on tropical bats using a before-after-control-impact design. Forest Ecology and Management 401, 8-16. https://doi.org/10.1016/j.foreco.2017.06.053