Publications and products

This page lists and summarizes our publications and other scientific products, since about 1999. First is a description of our books, then a list of other peer-reviewed publications with descriptions or abstracts, and then a list of some of our earliest products and presentations. Graduate student theses are not included here but are listed at this page.

According to Google Scholar, as of April 2024 the products listed here have been cited over 20,000 times.

Some of the journal articles are available from Bret Harvey's web site; or you can email Steve Railsback for papers not available here.


Agent-based and Individual-based Modeling: A Practical Introduction, Second Edition, Railsback, S. F., and V. Grimm. Princeton University Press, Princeton, New Jersey. 2019. The first edition was published in 2012. This is the first hands-on textbook for learning individual-based modeling, covering the topics addressed by our book Individual-based Modeling and Ecology (below) in a hands-on way. It includes instruction in implementing models with the NetLogo modeling platform. Much more information and supplementary materials are at the book's web site

Individual-based Modeling and Ecology, Grimm, V., and Railsback, S. F. Princeton University Press, Princeton, New Jersey. 2005. This was the first full monograph on individual-based modeling, and is now a standard reference on IBMs. It addresses the basics of ecological modeling and why we use IBMs, the "pattern-oriented modeling" strategy for designing IBMs with the right level of complexity, developing and testing theory for individual behavior, a conceptual framework for IBMs, and how to implement, analyze, and publish IBMs. A translation of this book to Simplified Chinese will soon be available from Higher Education Press.

Modeling Populations of Adaptive Individuals, Railsback, S. F., and Harvey, B. C. Princeton Monographs in Population Biology. 2020. This book addresses a fundamental issue of theoretical and applied ecology: how to model populations and communities in ways that include the effects of individual adaptive behavior, especially tradeoff behaviors such as foraging under predation risk. The book illustrates and provides extensive guidance for one approach: "state- and prediction-based theory" (SPT). Like conventional behavior theory, SPT assumes decisions are made to maximize a specific measure of future individual fitness such as probability of survival or expected number of offspring. But unlike conventional theory, SPT can work in individual-based models that include feedbacks of behavior (due, e.g., to competition among individuals) and environmental variability because instead of being solved with optimization it uses explicit predictions, approximations, and updating. Publication is expected in spring, 2020. Supplementary materials for this book are at this site.

Journal articles, technical reports, and conference proceedings

Harvey, B. C., J. L. White, R. J. Nakamoto, and S. F. Railsback. 2024. Empirical and model-based evaluation of a step-pool stream restoration project: Consequences for a highly valued fish population. North American Journal of Fisheries Management 44:637-649.

This article describes a combination field and simulation experiment to (a) evaluate the benefits of a step-pool stream restoration project, especially for steelhead, and (b) test the ability of our inSALMO model to predict those benefits. Field monitoring found steelhead abundance to increase dramatically immediately after the restoration project was completed but then decline to about twice the pre-restoration abundance. The model predicted restoration to result in an approximate doubling of abundance, corresponding to the final field observations.

Railsback, S. F. and B. C. Harvey. 2023. Can thermal refuges save salmonids? Simulation of cold pool benefits to trout populations. Transactions of the American Fisheries Society 152:381-512.

In this featured article, we applied InSTREAM 7 to the question of how stream trout popultaion abundance and persistence depend on the availability of cold-pool refuges, as temperatures increase. We simulated summer survival and growth, and long-term persistence and abundance, under four levels of refuge availability and four levels of temperature. We found that refuges could provide population persistence, but long-term abundance depended strongly on refuge size. The experiment did not support the conceptual model of refuges as letting fish "hang on" at high densities during peak temperatures, but instead suggested that competition for resources inside refuges was a strong driver of summer survival. For this paper, we created a special version of InSTREAM 7 that represents deep, slow habitat cells as refuges with lower temperatures; this version is available in the Zenodo repository (

Railsback, Steven F.; Harvey, Bret C.; Ayllón, Daniel. 2023. InSTREAM 7 user manual: model description, software guide, and application guide. Gen. Tech. Rep. PSW-GTR-276. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station. 306 p.

This is the official citation for a complete description of InSTREAM 7 (and, therefore, for most of InSALMO 7). The report also contains guidance on using the model software and applying the model to new sites, and includes a summary of InSTREAM has been validated. However, the most recent unofficial versions of the User Manual are at the InSTREAM 7 page.

Railsback, S. F. 2023. Spatial scales in instream flow modeling: why and how to use ecologically appropriate resolutions. River Research and Applications 39:987–992.

Conventional, widely used instream flow assessment models mishandle spatial scales in ways that can strongly bias results. This short article outlines the basics of selecting and using appropriate spatial resolutions for models of various kinds of fish. The article could serve as a brief guide to selecting spatial resolutions for ecological models in general. A pre-publication version is here.

Railsback, S. F. 2022. Suboptimal foraging theory: how inaccurate predictions and approximations can make better models of adaptive behavior. Ecology 103:e3721.

This Ecology Concepts and Synthesis article discusses how we can use approximations to model adaptive behavior in contexts--like most individual-based models--too complex for traditional optimization-based theory. Useful approximations often include that individuals base decisions on predictions of future conditions; even predictions that are usually inaccurate can produce good adaptive behaviors. The article advocates the methods explained in detail by Railsback and Harvey's book (2020, above) as a type of theory useful when optimization is not.  A pre-publication version is here.

Railsback, S. F. 2022. What we don't know about the effects of temperature on salmonid growth. Transactions of the American Fisheries Society 151:3-12.

Bioenergetics models have proven very useful for predicting how temperature affects fish growth, but their results can conflict with assumptions commonly used in salmonid management. A standard Rainbow Trout bioenergetics model indicates that: (a) there is no “optimal temperature for growth” unless food is unnaturally abundant; in fact, (b) we cannot predict how changes in temperature affect growth without also assuming how food intake changes; (c) temperature can have stronger effects on growth in cool seasons than in summer; and (d) salmonids genetically adapted to tolerate high temperatures are more, not less, susceptible to effects of temperature on growth. Despite their widespread use, we still have no data sets that allow comprehensive testing and calibration of bioenergetics models. A pre-publication version is here.

Railsback, S. F., D. Ayllón, and B. C. Harvey. 2021. InSTREAM 7: Instream flow assessment and management model for stream trout. River Research and Applications, DOI: 10.1002/rra.3845.

This paper introduces the newest version of inSTREAM and how it differs from previous versions and from other methods for assessing effects of flow and temperature regimes on salmonid populations. It provides an example assessment of flow and temperature regimes, illustrating how the model can evaluate the cumulative, population-level effects of processes such as how temperature affects growth and behavior, how temperature affects egg survival and development rates, and how behavior mediates relations among hydraulic habitat, growth, and survival. A pre-publication version is here.

Railsback, S. F., B. C. Harvey, and D. Ayllón. 2021. Importance of the daily light cycle in population-habitat relations: a simulation study. Transactions of the American Fisheries Society 150:130-143.

We simulated trout population response to instream flow and temperature regimes using 3 model versions differing in the number of daily light phases they represent: day-only, day and night, and day, night, dawn, and dusk. Predicted population responses differed in ways indicating that we need to consider all these phases in relating fish populations to habitat conditions.

Harvey, B. C. and S. F. Railsback. 2021. “All fish, all the time”: a good general objective for fish passage projects? Fisheries 46:119-124.

InSTREAM simulations indicate that trout populations can be sustained over time and space in a stream network containing movement barriers that are only rarely passable. This result implies that trout populations may not be harmed by habitat improvements for other species that partially block passage (e.g., artificial beaver dams to enhance frog habitat).

Ayllón, D., S. F. Railsback, V. Grimm, and others. 2021. Keeping modelling notebooks with TRACE: good for you and good for environmental research and management support. Environmental Modelling & Software 136:104932.

The practice of keeping modeling notebooks and its benefits are explained and advocated for environmental modelers, especially those supporting management decision-making.

Railsback, S. F., B. C. Harvey, and D. Ayllón. 2020. Contingent tradeoff decisions with feedbacks in cyclical environments: testing alternative theories. Behavioral Ecology 31:1192-1206.

This paper describes and tests a way to model how animals decide whether to feed or hide during each of four daily light phases (dawn, day, dusk, night) as a tradeoff between growth and predation risk when both depend on light levels as well as competition and habitat variables.

Grimm, V., S. F. Railsback, and 17 others. 2020. The ODD protocol for describing agent-based and other simulation models: A second update to improve clarity, replication, and structural realism. Journal of Artificial Societies and Social Simulation 23:7.

This is the latest guidance on ODD, a widely used standard for describing (and thinking about, and designing) individual- and agent-based models. The key part of this publication is Supplement 1, which provides detailed guidance and checklists for writing an ODD description (and thinking about a model's design).

Forbes, V. E., S. Railsback, C. Accolla, B. Birnir, R. J. F. Bruins, V. Ducrot, N. Galic, K. Garber, B. C. Harvey, H. I. Jager, A. Kanarek, R. Pastorok, R. Rebarber, P. Thorbek, and C. J. Salice. 2019. Predicting impacts of chemicals from organisms to ecosystem service delivery: A case study of endocrine disruptor effects on trout. Science Of the Total Environment 649:949-59.

Forbes, V., C. Salice, B. Birnir, R. Bruins, P. Calow, V. Ducrot, N. Galic, K. Garber , B. Harvey, Y. Jager, A. Kanarek, R. Pastorok, S. Railsback, R. Rebarber, and P. Thorbek. 2017. A framework for predicting impacts on ecosystem services from (sub)organismal responses to chemicals. Environmental Toxicology and Chemistry 36:845-59.

These two papers resulted from our participation in a National Institute for Mathematical and Biological Synthesis (NIMBioS; University of Tennessee) working group on ecotoxicology modeling. The group used our inSTREAM trout model in a case study of modeling effects of a contaminant (an estrogen compound that inhibits reproduction in male trout) on fish communities and the ecosystem services they provide.

Ayllón, D., S. F. Railsback, B. C. Harvey, I. G. Quirós, G. G. Nicola, B. Elvira, and A. Almodóvar. 2019. Mechanistic simulations predict that thermal and hydrological effects of climate change on Mediterranean trout cannot be offset by adaptive behaviour, evolution, and increased food production. Science Of the Total Environment:133648.

Ayllón, D., G. G. Nicola, B. Elvira, and A. Almodóvar. 2019. Optimal harvest regulations under conflicting tradeoffs between conservation and recreational fishery objectives. Fisheries Research 216:47-58.

Ayllón, D., S. F. Railsback, A. Almodóvar, G. G. Nicola, S. Vincenzi, B. Elvira, and V. Grimm. 2018. Eco-evolutionary responses to recreational fishing under different harvest regulations. Ecology and Evolution 8:9600-13.

Ayllón, D., S. F. Railsback, S. Vincenzi, J. Groeneveld, A. Almodóvar, and V. Grimm. 2016. InSTREAM-Gen: Modelling eco-evolutionary dynamics of trout populations under anthropogenic environmental change. Ecological Modelling 326:36-53.

These papers use Daniel Ayllón Fernandez's "InSTREAM-Gen": a modification of inSTREAM that includes simulated genetic evolution of two trout traits. Daniel did this work as a researcher at the Helmholtz Centre for Environmental Research UFZ and at Universidad Complutense Madrid, Spain. They describe the model version and its application to management issues such as angler regulation. Dr. Ayllón continues to work with us closely on development and application of inSTREAM.

Railsback, S., D. Ayllón, U. Berger, V. Grimm, S. Lytinen, C. Sheppard, and J. Thiele. 2017. Improving execution speed of models implemented in NetLogo. Journal of Artificial Societies and Social Simulation 20:3.

We and collaborators published paper to show that our preferred software platform, NetLogo, is capable for the kind of large models we often build, and to show how to improve execution speed. There is a web page with updates to the article here.

Harvey, B. C., and J. L. White. 2017. Axes of fear for stream fish: water depth and distance to cover. Environmental Biology of Fishes 100:565–73.

Harvey, B. C., and J. L. White. 2016. Use of cover for concealment behavior by rainbow trout: influences of cover structure and area. North American Journal of Fisheries Management 36:1308-14.

Harvey, B. C., and R. J. Nakamoto. 2013. Seasonal and among-stream variation in predator encounter rates for fish prey. Transactions of the American Fisheries Society 142:621-27.

While these are not modeling papers, they illustrate the field and laboratory experiments we do to support and test our models.

Grimm, V., D. Ayllón, and S. F. Railsback. 2017. Next-generation individual-based models integrate biodiversity and ecosystems: yes we can, and yes we must. Ecosystems 20:229-36.

This article was invited for inclusion in the 20th anniversary issue of Ecosystems. The abstract: Ecosystem and community ecology have evolved along different pathways, with little overlap. However, to meet societal demands for predicting changes in ecosystem services, the functional and structural view dominating these two branches of ecology, respectively, must be integrated. Biodiversity–ecosystem function research has addressed this integration for two decades, but full integration that makes predictions relevant to practical problems is still lacking. We argue that full integration requires going, in both branches, deeper by taking into account individual organisms and the evolutionary and physico-chemical principles that drive their behavior. Individual-based models are a major tool for this integration. They have matured by using individual-level mechanism to replace the demographic thinking which dominates classical theoretical ecology. Existing individual-based ecosystem models already have proven useful both for theory and application. Still, next-generation individual-based models will increasingly use standardized and re-usable submodels to represent behaviors and mechanisms such as growth, uptake of nutrients, foraging, and home range behavior. The strategy of pattern-oriented modeling then helps make such ecosystem models structurally realistic by developing theory for individual behaviors just detailed enough to reproduce and explain patterns observed at the system level. Next generation ecosystem scientists should include the individual-based approach in their toolkit and focus on addressing real systems because theory development and solving applied problems go hand-in-hand in individual-based ecology.

Railsback, S. F., B. C. Harvey, S. J. Kupferberg, M. M. Lang, S. McBain, and H. H. J. Welsh. 2016. Modeling potential river management conflicts between frogs and salmonids. Canadian Journal of Fisheries and Aquatic Sciences 73:773-84.

We describe FYFAM, our new model of river management effects on yellow-legged frog breeding, and use the model to compare unregulated flows and temperatures vs. those managed primarily for salmon populations. The abstract: Management of regulated rivers for yellow-legged frogs and salmonids exemplifies potential conflicts among species adapted to different parts of the natural flow and temperature regimes. Yellow-legged frogs oviposit in rivers in spring and depend on declining flows and warming temperatures for egg and tadpole survival and growth, whereas salmonid management can include high spring flows and low-temperature reservoir releases. We built a model of how flow and temperature affect frog breeding success. Its mechanisms include adults selecting oviposition sites to balance risks of egg dewatering by decreasing flow versus scouring by high flow, temperature effects on development, habitat selection by tadpoles, and mortality via dewatering and scouring. In simulations of a regulated river managed primarily for salmonids, below-natural temperatures delayed tadpole metamorphosis into froglets, which can reduce overwinter survival. However, mitigating this impact via higher temperatures was predicted to cause adults to oviposit before spring flow releases for salmonids, which then scoured the egg masses. The relative timing of frog oviposition and high flow releases appears critical in determining conflicts between salmonid and frog management.

Railsback, S. F., and B. C. Harvey. 2016. Understanding anadromy as an individual adaptive behaviour: theory and its consequences. Pages 41-52 in G. S. Harris, editor. Sea Trout: Science & Management. Proceedings of the 2nd International Sea Trout Symposium, October 2015, Dundalk, Ireland.

This proceedings paper is from a keynote presentation at the conference, and reports results of simulation experiments using the steelhead trout version of our inSALMO model. The abstract: Facultative anadromy, the apparent ability of individual fish to decide if and when to migrate to the ocean, has been addressed as a genetic tendency, a population-level adaptation and as an individual adaptive behaviour. We developed a model, inSALMO-FA, that represents anadromy as an individual behaviour that maximizes expected reproductive output at the next spawning. In the model, each juvenile salmonid decides if and when to migrate to the ocean by considering its current size, recently experienced growth and predation risk. The conclusions from several simulation experiments include: 1) individual variation in experience can produce both anadromous and resident individuals under many conditions, even at the same site, 2) predation risk may be more important than growth in driving anadromy, 3) habitat enhancement can produce more of both smolts and residents instead of causing one life history to dominate at a site, and 4) individuals with higher metabolic rates and dominance are not consistently more likely to smolt. Because models like inSALMO let us examine the consequences of alternative hypotheses about anadromy under controlled and observable conditions that still include much natural complexity, they are an important supplement to field and laboratory experiments for understanding anadromy.

Belarde, T. A., and S. F. Railsback. 2016. New predictions from old theory: Emergent effects of multiple stressors in a model of piscivorous fish. Ecological Modelling 326:54-62.

This paper by former MS student Tyler Belarde analyzes a model of river fluctuation effects on juvenile Colorado pikeminnow, using the model as a testbed for hypotheses about cumulative effects of multiple stressors. The abstract:  Predicting cumulative effects is an important challenge of theoretical and management ecology. If a population will be exposed to multiple stressors (e.g., toxins, introduced competitors, climate change), will their cumulative effects be independent and hence multiplicative (the population survival rates due to each stressor can be multiplied together to determine the total reduction in abundance), synergistic (cumulative effects are greater than multiplicative), or antagonistic (stressors offset each other so cumulative effects are less than multiplicative)? Further, the effects of each stressor can vary with such factors as habitat quality, population density, and weather. It is difficult to predict cumulative effects with traditional population-level models because such models must assume the type and strength of stressor interactions a priori, and measuring stressor effects and interactions empirically is rarely practical. Instead, we used an individual-based model in which cumulative effects emerge from how each stressor affects the growth and survival of individuals, and how individuals interact. Our model is in fact based on theoretical concepts explored in the landmark 1980 paper of DeAngelis et al. (Cannibalism and size dispersal in young-of-the-year largemouth bass: experiment and model, Ecol. Model. 8, 133–148): in a community of fish that eat each other, initial differences in size among individuals have strong effects on subsequent abundance and size distributions. We model survival and growth of juvenile Colorado pikeminnow (Ptychocheilus lucius) during their first year, and two stressors they are subject to. The first stressor is a daily cycle of flow fluctuations imposed by an upstream hydroelectric dam; these fluctuations affect habitat area, food supply, and temperature, which then affect juvenile fish growth. Second is an introduced fish species that competes with pikeminnow for food, while both species can prey on each other via the size-based mechanism described by DeAngelis et al. We simulated the effects of the 36 combinations of six levels of these two stressors in each of 28 sites and weather year to produce 840 scenarios, using 7 weather year datasets as replicates. Emergent cumulative effects were multiplicative in 69% of these scenarios, synergistic in 22%, and antagonistic in 9%. Therefore, any a priori assumption about stressor interactions would be wrong in many situations. Synergistic effects were most common in deeper and larger habitats favorable to the introduced species; antagonistic effects were most common in smaller habitats where the introduced species had low growth, because flow fluctuations further reduced the small food supply.

Stillman, R. A., S. F. Railsback, J. Giske, U. Berger, and V. Grimm. 2015. Making predictions in a changing world: the benefits of individual-based ecology. BioScience 65:140-50.

Our salmonid modeling program is one of several case studies of individual-based ecological research. The abstract: Ecologists urgently need a better ability to predict how environmental change affects biodiversity. We examine individual-based ecology (IBE), a research paradigm that promises better a predictive ability by using individual-based models (IBMs) to represent ecological dynamics as arising from how individuals interact with their environment and with each other. A key advantage of IBMs is that the basis for predictions—fitness maximization by individual organisms—is more general and reliable than the empirical relationships that other models depend on. Case studies illustrate the usefulness and predictive success of long-term IBE programs. The pioneering programs had three phases: conceptualization, implementation, and diversification. Continued validation of models runs throughout these phases. The breakthroughs that make IBE more productive include standards for describing and validating IBMs, improved and standardized theory for individual traits and behavior, software tools, and generalized instead of system-specific IBMs. We provide guidelines for pursuing IBE and a vision for future IBE research.

Railsback, S. F., B. C. Harvey, and J. L. White. 2015. Effects of spatial extent on modeled relations between habitat and salmon spawning success. Transactions of the American Fisheries Society 144:1220-36.

The abstract: We address the question of spatial extent: how model results depend on the amount and type of space represented. For models of how stream habitat affects fish populations, how do the amount and characteristics of habitat represented in the model affect its results and how well do those results represent the whole stream? Our analysis used inSalmo, an individual-based model of salmon spawning, incubation, and juvenile rearing. The model was applied to 12 sites, totaling 4.0 km length, on Clear Creek, California, treating the simulated 4.0 km as a synthetic whole stream. Simulation experiments examined responses of salmon spawning and rearing success to habitat variables such as flow and temperature, when the model included each individual site, all sites, and random combinations of two to nine sites. Some responses, such as temperature effects on egg incubation, were insensitive to spatial extent. Other responses, including effects of flow on production of large juveniles, varied sharply among sites and varied with spatial extent. Most small sites had little effect on overall results, but one small site provided exceptionally good juvenile rearing habitat and strongly affected responses of the entire stream. Larger sites (length > 15 channel widths) in distinct habitat types (e.g., highly disturbed and recently restored) also had strong effects. Including more or longer sites generally increased model representativeness but not consistently. Results highly representative of the entire stream could also be obtained by combining large sites in typical habitat with "hot spots" of especially productive habitat. Finally, sites lower in the watershed appear to be more important to model results and salmon spawning success because more juveniles migrate through them.

Penaluna, B. E., J. B. Dunham, S. F. Railsback, I. Arismendi, S. L. Johnson, R. E. Bilby, M. Safeeq, and A. E. Skaugset. 2015. Local variability mediates vulnerability of trout populations to land use and climate change. PLoS ONE 10:e0135334.

Penaluna, B. E., S. F. Railsback, J. B. Dunham, S. Johnson, R. E. Bilby, and A. E. Skaugset. 2015. The role of the geophysical template and environmental regimes in controlling stream-living trout populations. Canadian Journal of Fisheries and Aquatic Sciences 72:893-901.

These papers are products of Brooke Penaluna's PhD research at Oregon State University. She applied inSTREAM to a variety of headwater streams and used it to study relative effects of several stressors.

Railsback, S. F., and M. D. Johnson. 2014. Effects of land use on bird populations and pest control services on coffee farms. Proceedings of the National Academy of Sciences of the United States of America 111:6109-14.

This is the second of two papers based on our coffee farm model; it uses the model to draw conclusions about how land management affects bird populations, pest infestation, and coffee production. The abstract: Global increases in both agriculture and biodiversity awareness raise a key question: should cropland and biodiversity habitat be separated, or integrated in mixed land uses? Ecosystem services by wildlife make this question more complex. For example, birds benefit agriculture by preying on pest insects, but other habitat is needed to maintain the birds. Resulting land use questions include: what areas and arrangements of habitat support sufficient birds to control pests, whether this pest control offsets the reduced cropland, and what the benefits are of "land sharing" (mixed cropland and habitat) vs. "land sparing" (separate areas of intensive agriculture and habitat). Such questions are difficult to answer using field studies alone, so we use a simulation model of Jamaican coffee farms, where songbirds suppress the coffee berry borer (CBB). Simulated birds select habitat and prey in five habitat types: intact forest, trees (including forest fragments), shade coffee, sun coffee, and unsuitable habitat. The trees habitat type appears especially important: it provides efficient foraging and roosting sites near coffee plots. Small areas of trees (but not forest alone) could support enough birds to suppress CBB in sun coffee; the degree to which trees are dispersed within coffee had little effect. In simulations without trees, shade coffee supported sufficient birds to offset its lower yield. High areas of both trees and shade coffee reduced pest control because CBB was less often profitable prey. Because of the pest-control service by birds, land sharing was predicted more beneficial than land sparing in this system.

Railsback, S. F., and M. D. Johnson. 2011. Pattern-oriented modeling of bird foraging and pest control in coffee farms. Ecological Modelling 222:3305-19.

This is the first paper on the Jamaica coffee farm model. It was also designed as an explicit example of using "pattern-oriented modeling" to design individual-based models and test their theory for individual behavior. The paper (1) describes nine patterns observed in the field by Matt Johnson and his students, which characterize bird foraging and pest consumption; (2) describes the IBM designed so these patterns could emerge from it; and (3) tests four alternative theories for bird foraging behavior and identifies one that best reproduces the observed patterns. Some classical foraging theory is shown not to be useful at the level of realism of this model.

Railsback, S. F., and B. C. Harvey. 2013. Trait-mediated trophic interactions: is foraging theory keeping up? Trends in Ecology & Evolution 28:119-25.

This review resulted from our work developing foraging theory for individual-based models. The abstract: Many ecologists believe that there is a lack of foraging theory that works in community contexts, for populations of unique individuals each making trade-offs between food and risk that are subject to feedbacks from behavior of others. Such theory is necessary to reproduce the trait-mediated trophic interactions now recognized as widespread and strong. Game theory can address feedbacks but does not provide foraging theory for unique individuals in variable environments. "State- and prediction-based theory" (SPT) is a new approach that combines existing trade-off methods with routine updating: individuals regularly predict future food availability and risk from current conditions to optimize a fitness measure. SPT can reproduce a variety of realistic foraging behaviors and trait-mediated trophic interactions with feedbacks, even when the environment is unpredictable.

Railsback, S. F., B. C. Harvey, and J. L. White. 2014. Facultative anadromy in salmonids: linking habitat, individual life history decisions, and population-level consequences. Canadian Journal of Fisheries and Aquatic Sciences 71:1270-78.

The abstract: Modeling and management of facultative anadromous salmonids is complicated by their ability to select anadromous or resident life histories. Conventional theory for this behavior assumes individuals select the strategy offering highest expected reproductive success but does not predict how population-level consequences such as a stream's smolt production emerge from the anadromy decision and habitat conditions. Our individual-based population model represents juvenile growth, survival, and anadromy decisions as outcomes of habitat and competition. In simulation experiments that varied stream growth and survival conditions, we examined how many simulated juveniles selected anadromy vs. residence and how many of those choosing anadromy survived until smolting. Due to variation in habitat and among individuals, the within-population frequency of anadromy changed gradually with growth and survival conditions instead of switching abruptly. Higher predation risk caused more juveniles to select anadromy but fewer survived long enough to smolt. Improving growth appears a much safer way to increase smolt production compared to reducing freshwater survival. Smolt production peaked at high growth and moderately high survival, conditions that also produced many residents.

Harvey, B. C., and S. F. Railsback. 2014. Feeding modes in stream salmonid population models: is drift feeding the whole story? Environmental Biology of Fishes 97:615-25.

This paper resulted from our participation at an American Fisheries Society symposium on drift-feeding models. It draws two important conclusions. First, our inSTREAM trout model successfully predicted differences in adult trout growth and how growth was affected by stream flow, in a small stream. Second, the model was more accurate when "search" feeding was included as an alternative to drift feeding; search feeding appears to be important at least in low-flow situations.

Railsback, S. F., M. Gard, B. C. Harvey, J. L. White, and J. K. H. Zimmerman. 2013. Contrast of degraded and restored stream habitat using an individual-based salmon model. North American Journal of Fisheries Management 33:384-99.

This is one product of the inSALMO version of our salmonid IBMs. The abstract: Stream habitat restoration projects are popular, but can be expensive, and can be difficult to evaluate. We describe inSALMO, an individual-based model designed to predict habitat effects on freshwater life stages (spawning through juvenile outmigration) of salmon. We used inSALMO in a demonstration evaluation of habitat restoration on Clear Creek, California, by simulating production of total and large (> 5 cm fork length) Chinook salmon Oncorhynchus tshawytscha outmigrants at a restored and an unrestored site. The calibrated model reproduced observed redd locations and outmigrant timing and size. In simulations, the restored site had much higher production of fry that established and grew before outmigration; its habitat appears to provide high survival and positive growth due to moderate velocities, shallow depths, and cover for feeding and hiding. However, the restored site did not produce more total outmigrants because at both sites spawning gravel was sufficient and the vast majority of fry moved downstream soon after emergence. Simulations indicated that at both sites increasing food and cover availability could further increase production of large, but not total, outmigrants; while spawning gravel, temperature, and flow appear near-optimal already. Further gravel addition was predicted to increase total fry production but have little or even negative effect on production of large outmigrants, illustrating that actions benefitting one life stage can negatively affect others. The model predicted that further enhancements (e.g., in cover availability) would be more beneficial at the restored site than at the unrestored site; restoration efforts may be most effective when concentrated in "hot spots" with good habitat for growth and predator-avoidance as well as spawning. In contrast to the traditional notion of "limiting factors", the model indicates that multiple factors have strong effects. The model, while uncertain, provided more understanding of restoration effects than can field studies alone and appears useful for designing projects to meet specific restoration objectives.

Harvey, B. C., R. J. Nakamoto, J. L. White, and S. F. Railsback. 2014. Effects of streamflow diversion on a fish population: combining empirical data and individual-based models in a site-specific evaluation. North American Journal of Fisheries Management 34:247–57.

This paper describes a successful validation of inSTREAM in a small stream. The abstract: Resource managers commonly face the need to evaluate the ecological consequences of specific water diversions of small streams. We addressed this need by conducting 4 years of biophysical monitoring of stream reaches above and below a diversion and applying two individual-based models of salmonid fish that simulated different levels of behavioral complexity. The diversion of interest captured about 24% of streamflow between June and October but had little or no effect over the remainder of the year. The change in biomass of Rainbow Trout Oncorhynchus mykiss and steelhead (anadromous Rainbow Trout) over the dry season (June–October) favored the upstream control over the downstream diversion reach over 4 years (2008–2011). Dry-season growth did not differ consistently between the two reaches but did exhibit substantial annual variation. Longer-term observations revealed that in both reaches most fish growth occurred outside the period of dry-season diversion. After calibration to the upstream control reach, both individual-based models predicted the observed difference in fish biomass between control and diversion reaches at the ends of the dry seasons. Both models suggested the difference was attributable in part to differences in habitat structure unrelated to streamflow that favored the upstream reach. The two models both also reproduced the large seasonal differences in growth, small differences between reaches in individual growth, and natural distributions of growth among individuals. Both the empirical data and simulation modeling suggested that the current level of diversion does not threaten the persistence of the salmonid population. In multiyear simulations using the two models, the model incorporating greater flexibility in fish behavior exhibited weaker population-level responses to more extreme reductions in dry-season streamflow.We believe the application of individual-based models in this case has placed resource managers in a relatively strong position to forecast the consequences of future environmental alterations at the study site.

Grimm, V., J. Augusiak, A. Focks, B. M. Frank, F. Gabsi, A. S. A. Johnston, C. Liu, B. T. Martin, M. Meli, V. Radchuk, P. Thorbek, and S. F. Railsback. 2014. Towards better modelling and decision support: documenting model development, testing, and analysis using TRACE. Ecological Modelling 280:129-39.

This paper describes TRACE, a standard protocol for documenting (and therefore improving) the modeling process.

Yu, R., S. Railsback, C. Sheppard, and P. Leung. 2013. Agent-based fishery management model of Hawaii's longline fisheries (FMMHLF): model description and software guide. University of Hawaii at Manoa, School of Ocean and Earth Science and Technology and Joint Institute for Marine and Atmospheric Research, Honolulu, Hawaii. JIMAR Contribution 13-383.

This report is the main product of our collaboration with researchers at the University of Hawaii to model the Hawaiian long-line fishing fleet and how its vessels would respond to alternative sea-turtle protection regulations. It is available on-line.

Lytinen, S. L., and S. F. Railsback. 2012. The evolution of agent-based simulation platforms: a review of NetLogo 5.0 and ReLogo. in: Proceedings of the Fourth International Symposium on Agent-Based Modeling and Simulation, at the 21st European Meeting on Cybernetics and Systems Research (EMCSR 2012), Vienna, Austria, April 2012.

This paper compares the programming experience, documentation, and execution speed of two software platforms. NetLogo has continued to steadily increase in suitability for scientific applications since our 2006 review in Simulation (below). ReLogo is a new product of the Repast program; it implements NetLogo's primitives so they can be used in models programmed in the languages Groovy or Java, within the Eclipse development environment. Overall we found ReLogo to be substantially more cumbersome to use, lacking in documentation, and slower in execution than NetLogo. It was not clear how ReLogo could combine NetLogo-like code with Repast's libraries that are more specialized or general than NetLogo. This review is available here.

Harvey, B. C., and S. F. Railsback. 2012. Effects of passage barriers on demographics and stability properties of a virtual trout population. River Research and Applications 28:479-89.

Do barriers to upstream passage (low-head dams; culverts) affect trout population characteristics such as abundance and frequency of local extinction? How do such effects depend on the density and location of barriers? This simulation study, using inSTREAM to represent a large network of small to medium-sized stream reaches, produced a few surprising answers.

Grimm, V., and S. F. Railsback. 2012. Pattern-oriented modelling: a `multiscope' for predictive systems ecology. Philosophical Transactions of the Royal Society B 367:298-310.

This paper (based on a presentation to the Royal Society by V. Grimm) provides an excellent overview of the "pattern-oriented modeling" strategy for designing and testing IBMs.

Railsback, S. F., and B. C. Harvey. 2011. Importance of fish behaviour in modelling conservation problems: food limitation as an example. Journal of Fish Biology 79:1648–62.

This paper, based on a keynote talk by S. Railsback at the 2011 meeting of the Fisheries Society of the British Isles, uses inSTREAM to investigate the concept of "food limitation": at what level of food availability does a fish population no longer benefit from more? In the simulation experiments, when fish were assumed to use behavior to trade off feeding and predation avoidance, the traditional notion of food limitation was completely contradicted.

Grimm, V., U. Berger, F. Bastiansen, S. Eliassen, V. Ginot, J. Giske, J. Goss-Custard, T. Grand, S. Heinz, G. Huse, A. Huth, J. U. Jepsen, C. Jørgensen, W. M. Mooij, B. Müller, G. Pe’er, C. Piou, S. F. Railsback, A. M. Robbins, M. M. Robbins, E. Rossmanith, N. Rüger, E. Strand, S. Souissi, R. A. Stillman, R. Vabø, U. Visser, and D. L. DeAngelis. 2006. A standard protocol for describing individual-based and agent-based models. Ecological Modelling 198:115-296.

Grimm, V., U. Berger, D. L. DeAngelis, G. Polhill, J. Giske, and S. F. Railsback. 2010. The ODD protocol: a review and first update. Ecological Modelling 221:2760-68.

These papers established the widely used ODD protocol for describing agent- and individual-based models. This protocol has been used in hundreds of publications.

Railsback, S. F., B. C. Harvey, S. K. Jackson, and R. H. Lamberson. 2009. InSTREAM: the individual-based stream trout research and environmental assessment model. PSW-GTR-218, USDA Forest Service, Pacific Southwest Research Station, Albany, California.

This report documents version 4.2 of inSTREAM and can be downloaded here.

Harvey, B. C., and S. F. Railsback. 2009. Exploring the persistence of stream-dwelling trout populations under alternative real-world turbidity regimes with an individual-based model. Transactions of the American Fisheries Society 138:348-60.

Laboratory studies have shown that turbidity reduces the ability of trout to see and capture food, yet also reduces risk because the trout are more difficult for predators to see. What are the overall consequences of these opposing effects on trout populations? Sub-lethal effects of turbidity are difficult to evaluate, in part because turbidity varies widely and in part because effects on mortality, growth, and reproduction are very difficult to measure in rivers. We used the individual-based trout model, in combination with laboratory studies, to examine these issues and predict population-level consequences of individual-level turbidity effects. This paper is available from Bret Harvey's web site.

Grimm, V., and S. F. Railsback. 2009. Model the real, artificial, or stylized iguana? Artificial life and adaptive behavior can be linked through pattern-oriented modeling. Adaptive Behavior 17:309-12.

This invited commentary muses on the applicability of "artificial life" (computer simulations that are life-like but not explicitly related to any particular organism) to biological research. Pattern-oriented modeling (see Grimm et al. 2005 below) provides a way to learn about the adaptive behavior of real organisms from artificial life studies.

Harvey, B. C., and S. F. Railsback. 2007. Estimating multi-factor cumulative watershed effects on fish populations with an individual-based model. Fisheries 32:292-98.

This paper uses inSTREAM to investigate cumulative impacts and interactions among three stressors: elevated wet-season turbidity, elevated dry-season temperature, and loss of pools. Simulated effects were non-linear and non-additive: at high stress levels, cumulative effects were worse than predicted by assuming each stress acts alone. Such interacting effects are especially interesting for temperature and turbidity because they operated at different times of year. This paper is available from Bret Harvey's web site.

Grimm, V., E. Revilla, U. Berger, F. Jeltsch, W. M. Mooij, S. F. Railsback, H.-H. Thulke, J. Weiner, T. Wiegand, and D. L. DeAngelis. 2005. Pattern-oriented modeling of agent-based complex systems: lessons from ecology. Science 310:987-91.

Individual-based (or agent-based) models are an important tool for understanding complex systems, but science still needs a general strategy for designing, testing, and learning from such bottom-up models. This paper reviews examples of a strategy we call "pattern-oriented modeling". Using a variety of observed patterns helps scientists design and parameterize IBMs and to develop "algorithmic" theory for how system properties arise from characteristics of individuals and their environment. There is a link to this paper from Volker Grimm's web site.

Railsback, S. F., S. L. Lytinen, and S. K. Jackson. 2006. Agent-based simulation platforms: review and development recommendations. Simulation 82:609-23.

Which software platform is best for your individual-based model? This article reviews the most popular software platforms for individual- and agent-based modeling: MASON, NetLogo, Repast, and the Objective-C and Java versions of Swarm. The primary basis of the review is the authors' experience implementing a series of example models and teaching several platforms. The paper also compares the execution speed of the example models implemented in the different platforms. Conclusions include recommendations for future development of platforms for agent-based simulation.

Railsback, S. F., H. B. Stauffer, and B. C. Harvey. 2003. What can habitat preference models tell us? Tests using a virtual trout population. Ecological Applications 13:1580-94.

What do empirical observations of habitat selection (e.g., animal density) tell us about habitat quality? Which is a better predictor of population response to habitat alteration - an empirical model of density as a function of habitat, or a mechanistic understanding of how intrinsic habitat quality varies with habitat? In this paper we use our stream trout IBM as a virtual ecosystem to address these questions, with surprising results.

Railsback, S. F., B. C. Harvey, J. W. Hayse, and K. E. LaGory. 2005. Tests of theory for diel variation in salmonid feeding activity and habitat use. Ecology 86:947-59.

Digital appendices, including description of the trout IBM and animations of simulation experiments, are published in Ecological Archives here. The abstract: For many animals, selecting whether to forage during day or night is a critical fitness problem: at night, predation risks are lower but feeding is less efficient. Habitat selection is a closely related problem: the best location for nocturnal foraging could be too risky during daytime, and habitat that is safe and profitable in daytime may be unprofitable at night. We pose a theory that assumes animals select the combination of daytime and night activity (feeding vs. hiding), and habitat, that maximizes expected future fitness. Expected fitness is approximated as the predicted probability of surviving starvation and predation over a future time horizon, multiplied by a function representing the fitness benefits of growth. The theory’s usefulness and generality were tested using pattern-oriented analysis of an individual-based model (IBM) of stream salmonids and the extensive literature on observed diel behavior patterns of these animals. Simulation experiments showed that the IBM reproduces eight diverse patterns observed in real populations. (1) Diel activity (whether foraging occurs during day and/or night) varies among a population’s individuals, and from day to day for each individual. (2) Salmonids feed in shallower and slower water at night. (3) Individuals pack more tightly into the best habitat when feeding at night. (4) Salmonids feed relatively more at night if temperatures (and, therefore, metabolic demands) are low. (5) Daytime feeding is more common for life stages in which potential fitness increases more rapidly with growth. (6) Competition for feeding or hiding sites can shift foraging between day and night. (7) Daytime feeding is more common when food availability is low. (8) Diel activity patterns are affected by the availability of good habitat for feeding or hiding. We can explain many patterns of variation in diel foraging behavior without assuming that populations or individuals vary in how inherently nocturnal or diurnal they are. Instead, these patterns can emerge from the search by individuals for good tradeoffs between growth and survival under different habitat and competitive conditions.

Railsback, S. F., and B. C. Harvey. 2002. Analysis of habitat selection rules using an individual-based model. Ecology 83:1817-30.

Railsback, S. F., R. H. Lamberson, B. C. Harvey, and W. E. Duffy. 1999. Movement rules for spatially explicit individual-based models of stream fish. Ecological Modelling 123:73-89.

These two papers first document (1999) how we model adaptive tradeoff decisions in IBMs, and (2002) pattern-oriented testing of alternative theories for individual behavior.

Huse, G., S. Railsback, and A. Fernø. 2002. Modelling changes in migration patterns of herring: collective behaviour and numerical domination. Journal of Fish Biology 60:571-82.

Railsback, S. F. 2001. Concepts from complex adaptive systems as a framework for individual-based modelling. Ecological Modelling 139:47-62.

Railsback, S. F. 2001. Getting "results": the pattern-oriented approach to analyzing natural systems with individual-based models. Natural Resource Modeling 14:465-74.

Railsback, S. F., B. C. Harvey, R. H. Lamberson, D. E. Lee, N. J. Claasen, and S. Yoshihara. 2002. Population-level analysis and validation of an individual-based cutthroat trout model. Natural Resource Modeling 15:83-110.

Ropella, G. E. P., S. F. Railsback, and S. K. Jackson. 2002. Software engineering considerations for individual-based models. Natural Resource Modeling 15:5-22.

Products of historic interest

"How does individual behavior affect population resilience and stability in virtual trout?", S. F. Railsback and B. C. Harvey. Presentation at the 2006 Ecological Society of America meeting, August 7-11, Memphis. This paper was presented in the symposium Revisiting the "stability" icon: Upstart approaches to modeling resilience, organized by Volker Grimm, Donald DeAngelis, Uta Berger, and Steve Railsback.

"A sensitivity analysis of an individual-based trout model", P. M. Cunningham. Presentation at SwarmFest 2006, the Swarm Development Group agent-based modeling conference, July 23-24, University of Notre Dame.

"Juvenile Salmon Movement through the Sacramento-San Joaquin Delta: Challenges in Using Field Data to Validate Models", A. M. Dodd.   Poster presentation at the 2006 World Conference on Natural Resource Modeling, June 25-28, Norwegian School of Economics and Business Administration. Natural Resource Modeling special issue on individual-based models, now available. This issue contains papers from all speakers in our special symposium at the 2000 Ecological Society of America conference.

"Fish populations as complex adaptive systems", Geir Huse and Steve Railsback. Draft Manuscript. This paper discusses the new science of Complex Adaptive Systems and its application to understanding population ecology. It provides an introduction to key concepts of CAS and examples of how these concepts could change how we study marine and freshwater fisheries. This paper is the primary product of Dr. Huse's visit to Humboldt in the fall of 2000.

"Complex Adaptive Systems meets the real world: Making agent-based simulation work for ecological management and research". This seminar on our methods for building, testing, and doing science with individual-based models was presented by Steve Railsback to the Santa Fe Institute and the Center for Nonlinear Studies at Los Alamos National Laboratory.

"Advancing the Individual-based Modeling Approach: New Tools and Concepts ". Special symposium at the Ecological Society of America annual meeting, Snowbird, Utah, August 10, 2000. We organized this symposium, which presented progress on theoretical and software aspects of individual-based modeling. A special issue of the journal Natural Resource Modeling (see above) included papers from the symposium.

Presentations at SwarmFest 2002, the annual Swarm users conference, March 29-31, Seattle, Washington. Steve Railsback presented several examples of using individual-based models to test ecological theory. Steve Railsback and Tamara Grand presented an ecological perspective in a discussion of how agent-based simulation can contribute to theory in ecological and social sciences.

Presentations at SwarmFest 2001, the annual Swarm users conference, April 28-30, Santa Fe, New Mexico. Steve Jackson presented concepts for implementing simulations with multiple model swarms (separate models with differing time and space scales, and agents that pass among models). Steve Railsback led a panel discussion on publication of research based on agent-based simulation.

Presentations at SwarmFest 2000, March 11-13, Utah State University. Steve Jackson presented our method for automated experiment management (generating replicate simulations and scenario comparisons) in a discussion of alternative approaches. Steve Railsback presented the paper "Getting 'Results': The Pattern-oriented Approach to Analyzing Complex Systems with Agent-based Models".

"Individual-based Models: Progress Toward Viability for Fisheries Management", S. F. Railsback, R. H. Lamberson, and S. Jackson. Presentation at Spatial Processes and Management of Fish Populations, 17th Lowell Wakefield Symposium, Anchorage, Alaska, October 1999.

"Swarm Software for Individual-based Fish Models" presented by Steve Jackson, Steve Railsback, and Glen Ropella at SwarmFest99.

"Emerging Thoughts on Individual-based Fish and Wildlife Models" presented by Steve Railsback at SwarmFest99.

"Tools for Individual-based Stream Fish Models", report prepared by Steve Railsback and Steve Jackson. EPRI TR-114006, Electric Power Research Institute, Palo Alto CA, 1999. This document was developed in conjunction with the software to provide CIFSS users with guidance on building, testing, and using models. The report documents our conceptual approach to IBMs, and contains a complete user's guide to our trout model software. It outlines the software's structure and provides guidance on formulating models, implementing changes in model formulation in the CIFSS software, building input files, running and testing models, and conducting research and management experiments with CIFSS models.

"California Individual-based Fish Simulation System, Trout Instream Flow Model Formulation", Report prepared by Steve Railsback, Bret Harvey, Steve Jackson, Roland Lamberson, and Walt Duffy. This report documents the formulation of our first stream trout model, including how we simulate stream habitat and trout spawning and reproduction, movement, foraging and growth, and mortality. August, 1999.

"A Swarm-based System for Developing Individual-based Fish Models", S. F. Railsback, proceedings of the 1999 EcoHydraulics conference, Salt Lake City, Utah.