Probability Of Movement Research Articles

Modeling the invasive emerald ash borer risk of spread using a spatially explicit cellular model

The emerald ash borer (EAB, Agrilus planipennis) is decimating native ashes (Fraxinus sp.) throughout midwestern North America, killing millions of trees over the years. With plenty of ash available throughout the continent, the spread of this destructive insect is likely to continue. We estimate that the insect has been moving along a “front” at about 20 km/year since about 1998, but more alarming is its long-range dispersal into new locations facilitated by human activities. We describe a spatially explicit cell-based model used to calculate risk of spread in Ohio, by combining the insect’s flight and short-range dispersal (“insect flight”) with human-facilitated, long-range dispersal (“insect ride”). This hybrid model requires estimates of EAB abundance, ash abundance, major roads and traffic density, campground size and usage, distance from the core infested zone, wood products industry size and type of wood usage, and human population density. With the “insect flight” model, probability of movement is dependent on EAB abundance in the source cells, the quantity of ash in the target cells, and the distances between them. With the “insect-ride” model, we modify the value related to ash abundance based on factors related to potential human-assisted movements of EAB-infested ash wood or just hitchhiking insects. We attempt to show the advantage of our model compared to statistical approaches and to justify its practical value to field managers working with imperfect knowledge. We stress the importance of the road network in distributing insects to new geographically dispersed sites in Ohio, where 84% were within 1 km of a major highway.

Feasibility of Template-Guided Attenuation Correction in Cat Brain PET Imaging

Attenuation correction (AC) is important in quantitative positron emission tomography (PET) imaging of medium-sized animals such as the cat. However, additional time for transmission (TX) scanning and tracer uptake is required in PET studies with animal-dedicated PET scanners because post-injection TX scanning is not available in these systems. The aim of this study was to validate a template-guided AC (TGAC) method that does not require TX PET data for AC in cat 2-deoxy-2-[F-(18)fluoro-D-glucose (FDG) brain PET imaging. PET scans were acquired using a microPET Focus 120 scanner. TX data were obtained using a (68)Ge point source before the injection of FDG. To generate the attention map (mu-map) template for the TGAC, a target image of emission (EM) PET was selected, and spatial normalization parameters of individual EM data onto the target were reapplied to the corresponding mu-maps. The inverse transformations of the mu-map template into the individual spaces were performed, and the transformed template was forward projected to generate the AC factor. The TGAC method was compared with measured AC (MAC) and calculated AC (CAC) methods using region of interest (ROI) and SPM analyses. The ROI analysis showed that the activity of the TGAC EM PET images strongly correlated with those of the MAC data (y = 0.98x + 0.01, R(2) = 0.96). In addition, no significant difference was observed in the SPM analysis. By contrast, the CAC showed a significantly higher uptake in the deep gray regions compared to the MAC (corrected P < 0.05). The ROI correlation with MAC was worse than with the TGAC (R(2) = 0.84). In SPM analysis for the voxel-wise group comparisons between before and after the induction of deafness, only the TGAC showed equivalent results with the MAC. The TGAC was reliable in cat FDG brain PET studies in terms of compatibility with the MAC method. The TGAC might be a useful option for increasing study throughput and decreasing the probability of subject movement. In addition, it might reduce the possible biological effects of long-term anesthesia on the cat brain in investigations using animal-dedicated PET scanners.

The Interaction of Body Morphology, Directional Kinematics, and Environmental Structure in the Generation of Neonatal Rat (Rattus norvegicus) Locomotor Behavior

We show that modeling directional kinematics is especially important for understanding spatial patterns of behavior in an arena. In rats, although patterns of locomotion appear to be under sensory control, they may be largely if not entirely explained, during early locomotor development, by the directional kinematics of random movement together with morphological/environmental constraints. Different relative probabilities of forward movement versus lateral movement (lateral:forward ratio; LFR) have dramatic consequences for the patterns of movement of modeled agents in an arena. These patterns of movement, produced by rigid-bodied agents, are also modified by increasing morphological realism in our model, wherein agents are able to bend and flex like rat pups. For both morphological designs, agents run with empirically derived LFRs matched metrics of movement for 7-day-old pups very well and 10-day-old rat pups fairly well. We conclude that LFR is an important parameter in modeling (a) patterns of dynamical behavior in an arena and (b) trajectories of behavioral development.

Abundance and movement patterns of age-0 juvenile lake sturgeon in the Peshtigo River, Wisconsin

In Great Lakes tributaries, age-0 juvenile lake sturgeon Acipenser fulvescens use riverine nursery habitats during their first summer of life and migrate to adjacent lakes after a seasonal decline in water temperature. We used mark-recapture data collected during this river-residency phase to monitor patterns in juvenile abundance and movement in the Peshtigo River, Wisconsin, during 2006 and 2007. Jolly-Seber and multistate models were used to estimate abundance and describe the probability of movement between river sections. Juvenile abundance was higher in 2007 than in 2006; however, in both years, abundance at the end of the sampling season was lower than at the beginning. Downstream movements were observed more frequently than upstream movements, and the probability of a downstream movement was higher than the probability of an upstream movement. The lower abundance later in the sampling season and propensity for downstream movements suggests that some age-0 juvenile lake sturgeon may leave the natal river before a seasonal decline in water temperature.

Prediction of breakdown in ultrathin SiO2 films with fractal distribution of traps

An approach to modeling of thin SiO2 film degradation based on fractal distribution of traps is presented, which shows a high predictive power for ultrathin oxides breakdown. The fractal patterns of traps are produced by a modified version of the diffusion limited aggregation (DLA) model introduced in literature as to describe irreversible colloidal aggregation. In the present version of the DLA model, the trap generation in SiO2 is not random. It occurs at an interface or the trap aggregates to another existing trap. The movement probabilities of particles creating traps are defined through two parameters, which describe the lateral and straight movement. The two parameter values have been set as to obtain a certain fractal dimension (1.7) derived from optical inspections on dielectrics already reported in literature. The model has been calibrated using breakdown data obtained on samples fabricated ad hoc. It is coherent with the Weibull statistics describing the oxide film failure. Finally, the results of the model have been compared with experimental data of breakdown taken from literature.

Single-cell quantification of Bax activation and mathematical modelling suggest pore formation on minimal mitochondrial Bax accumulation

Mitochondrial outer membrane permeabilisation (MOMP) during apoptosis is triggered by the activation and oligomerisation of Bax and Bak, but a quantification of these processes in individual cells has not yet been performed. Single-cell imaging of Bax translocation and oligomerisation in Bax-deficient DU-145 cells expressing CFP-Bax and YFP-Bax revealed that both processes started only minutes before or concomitantly with MOMP, with the majority of Bax translocation and oligomerisation occurring downstream of MOMP. Quantification of YFP-Bax concentrations at mitochondria revealed an increase of only 1.8 + or - 1.5% at MOMP onset. This was increased to 11.2 + or - 3.6% in bak-silenced cells. These data suggested that Bax activation exceeded by far the quantities required for MOMP induction, and that minimal Bax or Bak activation may be sufficient to trigger rapid pore formation. In a cellular automaton modelling approach that incorporated the quantities and movement probabilities of Bax and its inhibitors, activators and enablers in the mitochondrial membrane, we could re-model rapid pore formation kinetics at submaximal Bax activation.

The Influence of Hemorrhagic Shock on the Electroencephalographic and Immobilizing Effects of Propofol in a Swine Model

Hemorrhagic shock increases the hypnotic effect of propofol, but the influence of hemorrhagic shock on the immobilizing effect of propofol is not fully defined. Twenty-four swine (30.3 +/- 3.6 kg) were anesthetized by inhalation of isoflurane and randomly assigned to either a control (n = 12) or a hemorrhagic shock (n = 12) group. Animals in the shock group were bled to a mean arterial blood pressure of 50 mm Hg and maintained at this level for 60 min. After isoflurane inhalation was stopped, propofol was infused at 50 mg x kg(-1) x h(-1) until no movement was observed after application of a dewclaw clamp every 2 min. Arterial samples for measurement of the propofol concentration were collected just before each use of the dewclaw clamp and the Bispectral Index (BIS) was also recorded. Analysis of the pharmacodynamics was performed using a sigmoidal inhibitory maximal effect model for BIS versus effect-site concentration and a logistic regression analysis for the probability of movement versus effect-site concentration. The propofol doses needed to reach a 50% decrease from baseline BIS, and no movement after noxious stimuli were reduced by hemorrhagic shock by 54% and 38%, respectively. Hemorrhagic shock decreased the effect-site concentration that produced 50% of the maximal BIS effect from 11.6 +/- 3.8 to 9.1 +/- 1.7 microg/mL and that producing a 50% probability of movement from 26.8 +/- 1.0 to 20.6 +/- 1.0 microg/mL. The results show that hemorrhagic shock increases both the hypnotic and immobilizing effects of propofol due to pharmacokinetic and pharmacodynamic alterations, with the changes in pharmacodynamics occurring to a similar extent for both effects.

Application of a Multistate Model to Estimate Culvert Effects on Movement of Small Fishes

AbstractWhile it is widely acknowledged that culverted road‐stream crossings may impede fish passage, effects of culverts on movement of nongame and small‐bodied fishes have not been extensively studied and studies generally have not accounted for spatial variation in capture probabilities. We estimated probabilities for upstream and downstream movement of small (30‐120 mm standard length) benthic and water column fishes across stream reaches with and without culverts at four road‐stream crossings over a 4‐6‐week period. Movement and reach‐specific capture probabilities were estimated using multistate capture‐recapture models. Although none of the culverts were complete barriers to passage, only a bottomless‐box culvert appeared to permit unrestricted upstream and downstream movements by benthic fishes based on model estimates of movement probabilities. At two box culverts that were perched above the water surface at base flow, observed movements were limited to water column fishes and to intervals when runoff from storm events raised water levels above the perched level. Only a single fish was observed to move through a partially embedded pipe culvert. Estimates for probabilities of movement over distances equal to at least the length of one culvert were low (e.g., generally ≤0.03, estimated for 1‐2‐week intervals) and had wide 95% confidence intervals as a consequence of few observed movements to nonadjacent reaches. Estimates of capture probabilities varied among reaches by a factor of 2 to over 10, illustrating the importance of accounting for spatially variable capture rates when estimating movement probabilities with capture‐recapture data. Longer‐term studies are needed to evaluate temporal variability in stream fish passage at culverts (e.g., in relation to streamflow variability) and to thereby better quantify the degree of population fragmentation caused by road‐stream crossings with culverts.

SU-FF-T-147: Intra Fractional Motion in Clinical IMRT Prostate Treatments, Warrants the Use of Faster Treatment Techniques

Purpose: To estimate intrafracional motion of the prostate in a routine clinical setting and its impact on margin reduction and treatment time. Method and Materials: External beam treatment for cancer of the prostate using an IMRT technique was evaluated. Fifteen patients underwent a marker match procedure ensuring correct positioning at time of treatment. For all fields intra‐treatment images were obtained, yielding 5 to 7 images per fraction. IMRT was delivered using a dynamic sliding window technique. The obtained images were processed to remove IMRT information. The markers were detected in the image using an automated methodology. Every image was timestamped and chronologically adjacent images were backprojected to yield 3D marker coordinates. Allowing to calculate the position of the prostate during the treatment delivery at specific time instances. Using a Poisson model for the probability of movement we can determine the maximal allowable time frame within which to perform this treament. Results: The maximal treament time measured was 1460s, the shortest lasted 343s. The times were measured starting from the last image in the marker‐match procedure and includes the decision and adjustment process. Depending on the elapsed time we noticed an increase in positional confidence level from 5.8mm to 7.6mm. The delivery of the fields are of the order of 250s. Conclusion: We note a significant increase in probability of prostate movement in our treaments as time elapses. This limits the amount of margin reduction possible. There are two strategies possible to reduce this time. 1) Increase the marker match speed, or 2) increase the delivery speed. A good candidate to do this is the use of a volumetric Arc technique (VMAT) which is implemented in our department with RapidArc™. The latter is able to deliver the same or even better dose distribution in under 2 minutes (8 patients).

Gap‐crossing decisions of forest birds in a fragmented landscape

AbstractHabitat loss and fragmentation are recognized as primary drivers of biodiversity loss worldwide. To understand the functional effects of habitat fragmentation on bird populations, data on movement across gaps in habitat cover are necessary, although rarely available. In this study, we used call playback to simulate a conspecific territorial intruder to entice birds to move through the landscape in a predictable and directional manner. We then quantified the probability of movement in continuous forest and across cleared gaps for two forest‐dependent species, the grey shrike‐thrush (Colluricincla harmonica) and the white‐throated treecreeper (Cormobates leucophaeus). Fifty‐four playback trials were conducted for each species across distances ranging from 25 to 480 m in continuous forest and 15–260 m across gaps in a forest‐agricultural landscape in southern Victoria, Australia. The probability of movement was significantly reduced by gaps in forest cover for both species. Shrike‐thrushes were six times more likely to move 170 m in continuous forest than to cross 170‐m gaps. The mean probability that treecreepers would cross any gap at all was less than 0.5, and they were three times less likely to move 50 m across a gap than through continuous forest. Both species displayed non‐linear responses to increasing gap distance: we identified a gap‐tolerance threshold of 85 m for the shrike‐thrush and 65 m for the treecreeper beyond which individuals were most unlikely to cross. The presence of scattered paddock trees increased functional connectivity for the shrike‐thrush, with individuals crossing up to 260 m when scattered trees were present. We conclude that gaps in habitat cover are barriers to movement, and that characteristics of the intervening matrix influence landscape permeability.

Development of a Multi-Scale Electromigration Simulator Based on a Combination of Ultra Accelerated Quantum Chemical Molecular Dynamics and Kinetic Monte Carlo Methods Application to Cu Interconnects Lifetime Simulation

We present a novel study on electromigration (EM) phenomena in Cu interconnects using a newly developed multi-scale simulator that consists on a combination of a device scale simulator based on a kinetic Monte Carlo (KMC) method and an atomic scale simulator based on ultra accelerated quantum chemical molecular dynamics (UA-QCMD). We have firstly demonstrated the simulation of the lifetime of Cu interconnects using the newly developed device scale simulator setting some suitable KMC probabilities for the void movement according to the regions in which it can be divided, i.e., the crystal grain and the grain boundary. The simulated values are shown to be in good agreement with experimental values. In an attempt to connect the device scale studies to quantum chemical instances of the system – since the correlation of probability of the void movement with, for example, activation energies or diffusion coefficients is important – we have developed an atomic scale simulator based on our original UA-QCMD method. In this atomic scale simulation, the electron wind force was evaluated using our original electrical conductivity prediction simulator based on KMC method which uses the electronic states from tight-binding quantum chemical (TBQC) calculation. Using this atomic scale simulator under the conditions of 475 K of temperature and 2.5 ×1010 A/m2 of current density, we were able to successfully simulate the migration of a Cu atom from a lattice site to a vacant site by evaluating the electron wind force.

Modelling the manipulation of C60 on the Si(001) surface performed with NC-AFM

We present a theoretical model of manipulation of theC60 molecule on the Si(001) surface with a non-contact atomic force microscope(NC-AFM). The model relies on the lowering of the energy barrier for theC60 manipulation due tothe interaction of the C60 with an AFM tip and the subsequent thermal movement of the molecule over this barrier. Weperformed numerical simulations of these energy barriers for a series of tip positions relative tothe molecule to show how the barriers change with the tip position. The values of thesebarriers are then used in kinetic Monte Carlo simulations to estimate the probability of theC60 movement for different tip positions and temperatures. Virtual atomic forcemicroscope simulations, which include the kinetic Monte Carlo treatment of theC60 movement, are then performed to describe in real time the process of movement of theC60 molecule during an NC-AFM scan. Our results demonstrate that manipulation of theC60 molecule, which is covalently bound to the surface, is possible with NC-AFM, even though thereis no continuous tip–molecule contact, which is known to be a necessary requirement for theC60 manipulation with scanning tunnelling microscopy. We show that the manipulation eventcan be identified in real NC-AFM experiments as an abrupt change in the distance of thetip closest approach (topography), and as spikes in the frequency shift and dissipationsignals.

An Individual-based Model for Reviewing Marine Reserves in the Light of Fisheries-induced Evolution in Mobility and Size at Maturation

Size-selective fishing mortality may induce evolutionary changes to smaller size at maturation in fish stocks. Marine reserves can be a useful management tool dealing with the ecological and evolutionary effects of fisheries. We examine the effect of size-selective harvesting and the implementation of marine reserves on the evolution of size at maturation using individual-based models. The mobility of individual fish is included as an evolving trait to account for possible behavioural changes in dispersal patterns. The evolutionary change in maturation size can be prevented inside the reserve, leading to higher abundance there. The results show the benefit of reserves on yield, especially at intermediate movement probability. Reserve implementation may cause mobility of individuals to decrease, increasing the possibility of local adaptation in maturation strategy. Evolutionary changes in mobility through marine reserve implementation increase the protection from evolution to small maturation size within the reserve, but may lead to overestimation of yield benefits due to reduced spill-over.

Estimation of survival and detection probabilities for multiple tagged salmon stocks with nested migration routes, using a large-scale telemetry array

Mark–recapture models for estimating survival and detection probabilities of tagged animals that migrate past successive receiver stations can incorporate multiple, linked stocks to improve detection probability estimates. When multiple release groups are analysed in a common framework, detection information can be shared to compensate for small sample sizes and provide generality beyond single-stock approaches. Methods for structuring complex detection history data and applying standard mark–recapture models are presented, allowing for information sharing among multiple stocks under nested migration route structures, where some portions of routes are shared with other stocks and other portions are unique. Possible biases from split-route migration patterns within a release group are described, along with a simple method of correcting these biases using stock-specific parameters that incorporate movement probabilities. Environmental covariates can be paired with stock-specific run timing data to model stock-specific detection probabilities that change seasonally. Finally, a method for assessing the redundancy of receivers on a multiple-receiver detection line is described, based on considerations of receiver line geometry. Examples are drawn from detection data of juvenile salmon on the Pacific Ocean Shelf Tracking Project (POST) array, but the methods presented here are transferable to other systems.

The Expand Probability Analysis of a Surrounding Large Truss-Board Antenna

Successful unfolding in the space is the first step for a satellite antenna to work normally, so that it is important to study the movement reliability of a large deployable satellite antenna. By studying the movement principium, a mechanical analysis model of the deployment mechanism and calculating formulae for the internal forces of trusses were presented. From the performance function of the truss unfolding to meet, a movement reliability analysis model of the mechanism was also given. Synthetically considering the dimension errors and the impact of space environmental factors, we treated the mechanism's movement as a function of random variables and derived the reliability calculation formula by using the method of second-order moment. Finally, the movement probability of the deployment mechanism was simulated by the computer within whole spreading process. The trend of the computing results agrees with the truss model experiments, which shows that the method is reasonable and effective.

Statistics of lateral atom manipulation by atomic force microscopy at room temperature

We perform vacancy-mediated lateral manipulations of Si adatoms on the $\text{Si}(111)\text{\ensuremath{-}}(7\ifmmode\times\else\texttimes\fi{}7)$ surface by atomic force microscopy at room temperature. A variety of line profiles associated with different atom hopping processes is observed in successive topographic line scans for the atom manipulations. Atom manipulation statistics show stochastic behavior in atom movements. The probability of atom movement increases with a decrease in the tip-surface distance. Moreover, it depends on the scan direction, even with crystallographic equivalence, because of tip apex asymmetry. In addition, the hopping rate reflects the difference in stability of the adsorption sites of the manipulated atoms.

Linking colonies to fisheries: Spatio-temporal overlap between common murres ( Uria aalge) from Tatoosh Island and coastal gillnet fisheries in the Pacific Northwest, USA

The population-level impact of seabird bycatch is difficult to assess because colony-of-origin is often unknown. As an alternative and complementary approach to ship-derived observations, we assessed the relative potential for bycatch of a known seabird population by quantifying spatio-temporal overlap with local fisheries. Common murres ( Uria aalge) have been reported as the most abundant seabird inadvertently caught in Washington and British Columbia coastal gillnet fisheries. In 1999–2001, we tracked 48 common murres from Tatoosh Island, the closest colony to the fisheries, during post-breeding. Using capture-mark-recapture models, we estimated weekly murre movement probabilities to/from three strata (offshore of, centered around, and inshore of Tatoosh Island). Based on movement probabilities and population size, we projected strata- and week-specific murre abundance. We created an index of overlap by calculating the product of murre abundance × gillnet fishing effort as a function of strata and time. The majority of murres (80%) moved inshore, where fishing effort was consistently the highest, suggesting that up to 4000 Tatoosh murres were vulnerable to bycatch. Index scores in the inshore stratum were 5–25 times higher relative to the offshore and Tatoosh strata, respectively. Overlap was sensitive to phenology, as index scores increased by 50% when dispersal was shifted four weeks earlier, while a two weeks delay decreased scores by 20%. Until the long-term impact of cumulative mortality in gillnet fisheries is determined, we believe a precautionary approach is warranted in the inshore stratum where the potential for bycatch was highest. We advocate the use of visible netting in inshore fisheries, a proven solution that reduces murre bycatch while maintaining fishing efficiency.

Dosing of Remifentanil to Prevent Movement During Craniotomy in the Absence of Neuromuscular Blockade

In neuroanesthesia practice, muscle relaxants may at times need to be avoided to facilitate intraoperative motor pathway monitoring. Our study's objective was to determine the optimal dose of remifentanil required to prevent movement after neurosurgical stimulation. After Institutional Review Board approval and written informed consent, 132 patients undergoing elective craniotomy randomly received one of 12 remifentanil dose regimens (0.10 to 0.21 microg/kg/min). Remifentanil was started before induction with propofol and succinylcholine. Anesthesia was maintained with isoflurane (0.6% end-tidal) in air/oxygen. During the study, movement was assessed on predetermined criteria by the anesthesiology, nursing, and neurosurgical teams. Heart rate and blood pressure were recorded every 5 minutes. We assessed the relationship between movement, hypotension, bradycardia, and dose using probit analysis and logistic regression. Sixty-five percent of the patients moved in response to surgical stimuli [95% confidence interval (CI): 49%-79%] at a remifentanil infusion rate of 0.10 microg/kg/min, and movement decreased to 21% (95% CI: 11-35) at 0.21 microg/kg/min. The probability of movement was 50% at an infusion rate (95% CI) of 0.13 (0.10 to 0.15) microg/kg/min remifentanil and decreased to 25% at an infusion rate of 0.19 (0.17 to 0.29) microg/kg/min. The probability of hypotension and bradycardia was 50% at 0.13 (0.10 to 0.15) microg/kg/min and 0.17 (0.15 to 0.21) microg/kg/min, respectively. Higher doses of remifentanil lessen the risk of movement in the absence of muscle relaxants with surgical stimulation for elective craniotomy. Hypotension and bradycardia were common at higher doses. Even at the maximum dose (0.21 mcg/kg/min) there was a 20% chance of movement. Adjunctive therapy is needed to ablate movement reliably, and to counteract the hypotensive effect of remifentanil. These findings may be helpful for clinicians administering remifentanil and isoflurane during procedures, where muscle relaxants may not be desirable.

Movement of Reservoir‐Stocked Riverine Fish between Tailwaters and Rivers

AbstractThe movement of fish from onstream impoundments into connected streams and rivers has traditionally been overlooked in fish stocking decisions but is critical to the ultimate impact of stocking riverine species into reservoirs. Hybrid saugeyes (female walleye Sander vitreus × male sauger S. canadensis) stocked into Deer Creek Reservoir, Ohio, readily move from the reservoir to the tailwater below. Downstream movement of these saugeyes from the tailwater may have consequences for native prey species and parental stocks downstream. We used fixed‐station radiotelemetry to quantify the temporal movement patterns of 203 reservoir‐stocked saugeyes from the tailwater of the reservoir, the stream flowing from the tailwater, and the river into which the stream flowed. From October 1998 through July 2000, most (75%) saugeyes never left the tailwater, and those that left returned 75% of the time. Overall, saugeyes spent 90% of their time in the tailwater, 7‐8% of their time downstream in small streams, and 2‐3% of their time farther downstream in the Scioto River (45 km downstream). No radio‐tagged saugeyes moved to the Ohio River (155 km downstream). The probability of downstream movement generally increased with increasing flow and when dissolved oxygen dropped to lethal levels in summer. The probability of movement was highest in winter and spring, when it was probably related to spawning, and low in summer (except when dissolved oxygen was low) and fall. The patterns of movement seemed to reflect the relative suitability of tailwater over stream habitat. The predominant use of and return to tailwater habitat after downstream movement limited overall stream and river residence time. Although the daily movement probability for an individual was low, when we apply these rates to all of the stocked saugeyes in the Ohio River drainage, we cannot safely conclude that only small numbers move from reservoir tailwaters to downstream river systems. We recommend that managers refrain from stocking systems for which there are concerns about native species in connected drainages.

Fitting Probability Distributions to Animal Movement Trajectories: Using Artificial Neural Networks to Link Distance, Resources, and Memory

Animal movement paths are often thought of as a confluence of behavioral processes and landscape patterns. Yet it has proven difficult to develop frameworks for analyzing animal movement that can test these interactions. Here we describe a novel method for fitting movement models to data that can incorporate diverse aspects of landscapes and behavior. Using data from five elk (Cervus canadensis) reintroduced to central Ontario, we employed artificial neural networks to estimate movement probability kernels as functions of three landscape-behavioral processes. These consisted of measures of the animals' response to the physical spatial structure of the landscape, the spatial variability in resources, and memory of previously visited locations. The results support the view that animal movement results from interactions among elements of landscape structure and behavior, motivating context-dependent movement probabilities, rather than from successive realizations of static distributions, as some traditional models of movement and resource selection assume. Flexible, nonlinear models may thus prove useful in understanding the mechanisms controlling animal movement patterns.