# Talks

Selected Talks

EE Holmes, N Kumar, S Maity, BR Smitha, S Cubelio, C Wilson, N Saravarane, N Kumar. 2016. Multivariate time-series forecasting of the NE Arabian Sea Oil Sardine fishery using satellite covariates. Univ of WA Quant Seminar. Oct 2016.

*Talk on the Indian oil sardine forecasting project. The project was still at an exploratory stage in Oct 2016, and the analysis changed substantially in the next year.*PDFN Tolimieri, EE Holmes, and G Williams. Multivariate Autoregressive State-Space Models (MARSS): PVA for data-poor species. 2016.

*Nick and I gave versions of this talk a few different places. It’s an overview of Tolimieri et al. (2017)*PDFEE Holmes. Inferring and forecasting community responses to climate drivers using multivariate time-series analysis. Univ of Idaho, ID. Dec 2015.

*Overview talk for the graduate student seminar in the Statistics Dept at Univ of Idaho.*PDFEE Holmes, E Ward, M Scheurell, S Hampton, L Scheef, E Wolkovich, D Pendleton, T Francis. Statistical Advances in Inferring and Forecasting Community Responses to Climate Drivers, using time-series analysis. Nov 2012.

*Overview talk on our CAMEO grant work given at the Cary Institute.*PDFEE Holmes, E Ward, M Scheurell, S Hampton, L Scheef, D Pendleton, T Francis. Univ of WA Quant Seminar. Nov 2011.

*Overview talk on our CAMEO grant. The last few slides talk about the problems of B matrix estimation.*PDFBX Semmens, EE Holmes, EJ Ward, CV Pattengill-Semmens, and BI Ruttenberg. Integrating time-series of reef fish community monitoring data. AFS Conference, Portland, OR. Aug 2011.

*This was a talk on Pattengill-Semmens et al. (2011) about combining different types of surveys using MARSS models.*PDFE Ward, EE Holmes and K Balcomb. 2009. Marine Mammal Bieenial Conference, Quebec City, Quebec. Quantifying the effects of prey limitation on killer whale reproduction. PDF of poster

S Viscido and EE Holmes. 2008. Multivariate autoregressive modeling of multi-species time-series data. Univ of WA Quant Seminar. Feb 2008.

*Overview talk about Steven’s work. Slides 28-37 summarize his simulations on trying to break the MAR estimation using short time series, observation error, and unknown environmental drivers. Take-home what that the approach is fairly robust.*PDFEE Holmes, Ecological Society of America Meetings, Memphis, TN. Parsimonious stochastic models for vertebrate population processes. August 2006.

*This was a talk where I summarized the philosophy and justification for a generic statistical model to describe the extinction dynamics for large classes of stochastic population models. The talk in Miami, July 2005, is where I first tried to enunciate these ideas. This represents the progress made in a year – or a 15 minute version at least. Nine months later, these ideas became the paper “A statistical approach to quasi-extinction forecasting”.*PDFEE Holmes. A statistical approach to PVA validation: testing quasi-extinction estimation and the IUCN red list criteria. 2006.

*Another early talk on the work I was doing that became Holmes et al. (2007). This has some of my comparisons of the Kalman filter versus ‘slope’ method. It illustrates how ‘slope’ performs better for short time series. The Kalman filter approach needs 20+ years.*PDFEE Holmes, LW Fritz, AE York, & K Sweeney. Evidence of continuing declines in fecundity of Steller sea lions. I’ve given various versions of this talk: March 2006 April 2006 June 2006.

EE Holmes, LW Fritz, AE York, & K Sweeney. Evidence of continuing declines in fecundity of Steller sea lions. Marine Mammal Society Meetings. San Diego, CA. December 2005. PDF of poster.

J Payne, EE Holmes(speaker), and AE Edwards, University of Idaho, Biology Dept. Ecosystem management: vision versus practice. November 2005

EE Holmes, University of Florida at Miami, Math Dept. Patterns of first passage probabilities in population monitoring data. July 2005.

*This was an informal almost philosophical talk where I tried to ennunciate some of my ideas about the meaning of parsimony in population modeling.*PDFEE Holmes, L Fritz, AE York and K Sweeney, Society for Marine Mammalogy Meetings, San Diego, CA. poster: Evidence of continuing declines in fecundity in Steller Sea Lions. December 2005

*This work is an extension of Holmes and York 2002. Here we are updating that work to look how demographics have been changing recently.*PDFEE Holmes and B Semmens, Institute of Theoretical and Mathematical Ecology, Univ. of Miami, Workshop on Spatial Ecology: The Interplay between Theory and Data. Diffusion approximation approaches for metapopulation viability analysis. January 2005.

*This was a talk mainly to theoreticians working in spatial ecology. I was expanding on the ideas in Holmes and Semmens (2004).*PDFEE Holmes and AE York, Cooperative Fish and Wildlife Research Unit, WA Dept. of Fish and Wildlife. Monitoring the effect of management on long-lived species. December 2004.

*Every so often I give a talk that actually has pictures of animals in it. This is basically a spashy remake of my ConsBio talk in 2001.*PDFEE Holmes, University of Washington Fisheries Dept., SAFS Quantitative Seminar, University of Washington. State-space models for PVA. December 2004.

*The actual talk slides do not capture at all the nature of this talk, which was by and large philosophical in nature. I spent a good quarter of the time at the blackboard talking about the nature of the debate about variability in population processes and why Pimm’s paper would have been considered important in the context of the debates about population regulation. This talk is during a time when I was developing the ideas behind Holmes et al. (2007).*PDFEE Holmes, Ecological Society of America Meetings, Portland, OR, “From theory to application: diffusion approximations for Population Viability Analysis”. This was in a session I organized titled Emerging approaches for the analysis of stochastic ecological data: dealing with multiple error sources, hidden states, complex nonlinearities, and uncertainty. August 2004.

*Although this talk has the same title as Holmes (2004), there are no slides or figures from that paper. This talk explores ideas that came up post Holmes and Fagan (2002) and Holmes et al. (2005). This shows a few results from the next installment of the Holmes and Fagan collaboration. The ideas in this talk hint at the ideas in U Florida in July 2005.*PDFAdam U, EE Holmes and K Balcomb, University of Washington, SAFS. Analysis of Southern Resident fin growth rates 1977-2003. July 2004.

*This was talk given to a killer whale research group on a small project using measurements of photos of juveniles killer whale fins to look at fin growth rates over a 25 year period.*PDFEE Holmes, Northwest Fisheries Science Center, Seattle, WA, “Cross-validations of a diffusion approximation approach to risk estimation for salmonids”. This was in a workshop I organized at NWFSC titled Technical Workshop on Population Trends and Extinction Metrics which was required as part of the 2000 Snake River hydropower Biological Opinion, December 5, 2003.

*The talk was an overview of about 5 years of research. I also prepared a white paper discussing that research in more detail. I think my white paper, in the final report, on diffusion approximations in practice is a lot more lucid than Holmes 2004. The final report has point-counterpoint structure as the participants gave comments and critiques of diffusion approximations estimates as used in the Snake River ESU. It gives a sense of some of the behind the scenes discussions and debates.*TALK Final ReportEE Holmes and W Fagan, Western Fisheries Society Annual Meeting, Symposium on Model validation and uncertainty. A statistical approach to validation applied to diffusion approximation models. Summer 2002.

*Basically a talk on Holmes and Fagan (2002). The talk had animation which makes the pdf look bad.*PDFEE Holmes and AE York, Society for Conservation Biology Annual Meeting, Symposium on Designing Marine Reserves for Marine Mammals. Using age-structure to monitor long-lived marine mammals. Summer 2001.

*This was a talk on Holmes and York (2002). See the remake of this talk given in December 2004 above.*EE Holmes. Estimating risks using diffusion approximations. Univ of WA Quant Seminar. 2001.

*This was a 45 minute ‘chalk’ talk. This was after I’d come up with the ‘slope’ method in Holmes (2001). I latter abandoned that since it was ad-hoc and I wanted (for statistical reasons) to use a proper maximum-likelihood method so I could use model selection criteria and likelihood ratio tests, but it actually is quite robust and we used it for a lot of salmon risk assessments. See McClure et al. (2003).*PDF