<p> Although herbaceous communities are important components of floodplain ecosystems, the factors constraining their restoration and post-restoration dynamics are poorly understood. Over the decade following restoration of a 3.2 km reach of the Merced River and floodplain in California, we tracked herbaceous community composition to distinguish floodplain habitats and utilized perturbations from revegetation treatments and post-restoration flooding to generate community assembly rule hypotheses regarding treatment effectiveness and persistence, with a particular interest in native perennials capable of suppressing non-natives over time if undisturbed. Revegetation treatments comprised combinations of sowing a sterile cover crop, sowing native species, and inoculating mycorrhizae. Most surveyed floodplain areas comprised a low terrace characterized by exceptionally droughty soils, relatively deep groundwater, and occasional flooding lasting into summer. Few species could tolerate both flood and drought to this extent, and the flood year community was generally distinct from that in non-flood years. Both communities were dominated by ruderals capable of avoiding stress and re-establishing following disturbance, including many non-native annual grassland species. Only <i>Artemisia douglasiana</i> responded to the treatments, as most seeded native species failed to establish, including those native perennial grasses expected to suppress non-native annuals, while other seeded native species either established adequately from natural dispersal or failed to persist through moderate flooding. Neither the cover crop nor mycorrhizal inoculation had any meaningful effect. Restoration efforts in naturally ruderal-dominated habitats may be better spent allowing natural regeneration, addressing particularly noxious invasives, and identifying or constructing habitats supporting long-lived native perennials. </p><p> Although originally developed for population sizes and population growth rates, modern capture-recapture models can estimate demographic rates in complex situations: multistate models for multiple study sites and stage-structured populations, superpopulation entry probability models for recruitment, and multievent models when state assessments are uncertain. However, combinations of these complications, such as recruitment studies with uncertain state assessments, are common, yet no single model has explicitly incorporated all of these elements. Ultimately, these models estimate the same fundamental population process with the same general approach, and we combine them in a generalized hidden process model based upon a simple discrete state and transition population model with Poisson recruitment that can estimate how recruitment and survivorship rates vary with respect to measured covariates from uncertain state assessments for a stage-structured population at multiple sites. Although closely related to the motivating models, the generalized model relaxes the Markov assumption. While we provide the distributions necessary to implement Bayesian data augmentation methods, we also provide an efficient analytical likelihood with a compact parameter space that is applicable in the absence of density-dependent mortality. As a demonstration, we estimate the influence of several covariates on recruitment and survivorship rates from uncertain observations of <i>Salix gooddingii </i> seedlings at different locations along a riparian gradient, and we use simulations to examine variation in the precision of estimated parameters. </p><p> In Mediterranean climates, cottonwoods and willows often exhibit high germination and seedling mortality rates, with recruitment occurring primarily in the occasional year when favorable spring floods improve survivorship. However, along the Robinson Reach of the Merced River, both germination and mortality rates appeared to be atypically low. To understand why these rates were so low along this recently restored flow-regulated, gravel-bedded stream, we surveyed <i>Populus fremontii, Salix exigua,</i> and <i> Salix gooddingii,</i> estimated germination and survivorship rates, and examined their correlations with factors expected to constrain recruitment, namely seed release, seed arrival, moist germination beds, light levels, groundwater depth, groundwater recession rates, and shear stress. Germination/initial establishment rates were low due in part to low seed arrival rates. Only <i> Salix gooddingii</i> was abundant enough to model in detail, and while moist germination surfaces increased germination/initial establishment, rates were low overall. Survivorship rates for <i>Salix gooddingii</i> seedlings and for small individuals were not correlated with any examined covariates. Seedlings tolerated moderate competition, and the absence of major scouring, even during 6 year flows, enabled survival at sites with sufficiently shallow groundwater that seedlings were unaffected by groundwater recession rates.</p>

<p>Numerical clustering encompasses a group of exploratory multivariate statistical methods devoted to finding groups in data based on either responses of individual variables or dissimilarity measures calculated from the variables. Despite their popularity, there have been few controlled comparisons of methods using data of known clustering structure and which compare more than a few methods. This study utilizes simulated plant community data to assess what data properties affect the performance of numerical clustering methods used in vegetation classification, including properties that can be controlled during data collection and measured before statistical analysis. This was done by with simulation experiments varying properties of species assemblages themselves ? ?-diversity, ?-diversity, and the level and type of noise in the data ? or the clustering structure of sampling units (SUs) in environmental space ? number of SUs per group, equality of number of SUs or cluster dispersion among groups, the proximity of adjacent clusters, and the number of clusters. Cluster recovery was measured using the Adjusted Rand Index (ARI) ? a chance-corrected measure of the proportion of elements classified similarly in two clustering results. ARI is an approximation of the proportion of sites correctly classified, so scores near 1.0 indicate accurate cluster recovery, while scores near 0.0 indicate poor cluster recovery. Methods are robust if they have a mean ARI score near 1.0 despite variation in data properties. Methods tested include flexible beta clustering, TWINSPAN, average, complete, and single linkage, K-means, Partitioning Around Medoids, ISOPAM, OPTPART, OPTSIL, Noise Clustering, model-based EM clustering (Mclust), Fuzzy Analysis (FANNY), and Information Analysis. Where applicable, methods were tested using four combinations of standardization and dissimilarity, yielding 59 unique combinations of method, standardization, and dissimilarity. Across all experiments, a couple of general trends emerge. No methods are robust when either ?-diversity or ?-diversity are very low. When ?-diversity is lowered by including a second set of generalist species along with a set of specialists, mean ARI scores are considerably higher than when decreasing ?-diversity by increasing the range of all species in the data set. Most methods are less robust when implemented with Euclidean distances, except for Ward?s method, PAM, FANNY, and Information Analysis (which only uses the information statistic calculated from presence data as a dissimilarity measure). Nonhierarchical methods fail when the number of SUs is highly unequal between clusters, except for OPTSIL initiated form Flexible Beta clustering results. Hierarchical methods are more sensitive to intermediate and outlier sites, though Ward?s method, Flexible Beta, Information Analysis, and TWINSPAN all perform better than UPGMA, complete linkage, and single linkage. Sources of random error are unimportant individually, but may be more important when paired with other factors. The optimal choice of clustering method is a product of trade-offs between near optimal performance in most experiments and robustness where other methods fail. For this reason, I recommend using Flexible Beta clustering with possible refinement by OPTSIL as a standard clustering method for vegetation classification. Flexible Beta clustering achieved mean ARI scores that are among the highest in all experiments, while remaining robust to factors that nonhierarchical methods (equality of number of SUs) and other hierarchical methods (intermediate/outlier SUs) are not robust to. OPTSIL did not always drastically improve Flexible Beta results, but it also never made them worse. Nevertheless, in models with low ?-diversity and when adjacent clusters are close together, OPTSIL does improve Flexible Beta Results.