Bayesian clustering with AutoClass explicitly recognises uncertainties in landscape classification

Clustering of multivariate data is a commonly used technique in ecology, and many approaches to clustering are available. The results from a clustering algorithm are uncertain, but few clustering approaches explicitly acknowledge this uncertainty. One exception is Bayesian mixture modelling, which treats all results probabilistically, and allows comparison of multiple plausible classifications of the same data set. We used this method, implemented in the AutoClass program, to classify catchments (watersheds) in the Murray Darling Basin (MDB), Australia, based on their physiographic characteristics (e.g. slope, rainfall, lithology). The most likely classification found nine classes of catchments. Members of each class were aggregated geographically within the MDB. Rainfall and slope were the two most important variables that defined classes. The second-most likely classification was very similar to the first, but had one fewer class. Increasing the nominal uncertainty of continuous data resulted in a most likely classification with five classes, which were again aggregated geographically. Membership probabilities suggested that a small number of cases could be members of either of two classes. Such cases were located on the edges of groups of catchments that belonged to one class, with a group belonging to the second-most likely class adjacent. A comparison of the Bayesian approach to a distance-based deterministic method showed that the Bayesian mixture model produced solutions that were more spatially cohesive and intuitively appealing. The probabilistic presentation of results from the Bayesian classification allows richer interpretation, including decisions on how to treat cases that are intermediate between two or more classes, and whether to consider more than one classification. The explicit consideration and presentation of uncertainty makes this approach useful for ecological investigations, where both data and expectations are often highly uncertain.     

Preliminary results of a classification of fifty-one selected northeastern Wisconsin Lakes (USA) using indicator diatom species

Diatom species lists were generated for 51 lakes in northeastern Wisconsin, and then classified by two-way indicator species analysis (TWINSPAN). The lakes were initially divided into two main groupings: Group I lakes which were alkaline lakes of moderate to high productivity, and Group II lakes which were acid lakes of low productivity. Group I lakes were further divided into two subgroupings, and four levels. In total, twelve indicator diatom species were recognized during the TWINSPAN lake classification, these were: Achnanthes exigua Grun., A. lanceolata var. dubia Grun., Amphora ovalis var. Affinis (Kütz.) V.H. ex Det., A. perpusilla (Grun.) Grun., Cymbella cistula (Ehrenb.) Kirch., Diploneis elliptica (Kütz.) Cl., Eunotia pectinalis (O.F. Mull.) Rabh., Fragilaria vaucheriae (Kütz.) Peters., Frustulia rhomboides var. crassinervia (Bréb. ex W. Sm.) Ross, Navicula capitata Ehrenb., N. decussis Østr., and N. scutelloides W.Sm. ex Gregory.     

Vegetation responses to landscape structure at multiple scales across a Northern Wisconsin,USA, pine barrens landscape

Increasing awareness of the importance of scale and landscape structure to landscape processes and concern about loss of biodiversity has resulted in efforts to understand patterns of biodiversity across multiple scales. We examined plant species distributions and their relationships to landscape structure at varying spatial scales across a pine barrens landscape in northern Wisconsin, U.S.A. We recorded plant species cover in 1×1 m plots every 5 m along a 3575 m transect, along with variables describing macro- and micro-landscape structure. A total of 139 understory plant species were recorded. The distributions of many species appeared to be strongly associated with landscape structural features, such as distinct management patches and roads. TWINSPAN and detrended correspondence analysis (DCA) identified three groups of species that overlapped extensively in the ordination, possibly reflecting the relatively homogeneous nature of disturbance in the pine barrens landscape. Distribution of understory plants did not reflect all of the patch types we identified along the transect; plot ordination and classification resulted in three to five plot groups that differed in niche breadth. Wavelet transforms showed varying relationships between landscape features and plant diversity indices (Shannon–Weiner, Simpson's Dominance) at different resolutions. Wavelet variances indicated that patterns of Shannon diversity were dominated by coarse resolutions ranging from 900–1500 m, which may have been related to topography. Patterns of Simpson's Dominance were dominated by 700 m resolution, possibly associated with canopy cover. However, a strong correspondence between overstory patch type and diversity was found for several patch types at ranges of scales that varied by patch type. Effects of linear features such as roads were apparent in the wavelet transforms at resolutions of about 5–1000 m, suggesting roads may have an important impact on plant diversity at landscape scales. At broad scales, landscape context appeared more important to diversity than individual patches, suggesting that changes in structure at fine resolutions could alter overall diversity characteristics of the landscape. Therefore, a hierarchical perspective is necessary to recognize potential large-scale change resulting from small-scale activities.     

Quantifying uncertainties in the microvascular transport of nanoparticles

The character of nanoparticle dispersion in the microvasculature is a driving factor in nanoparticle-based therapeutics and bio-sensing. It is difficult, with current experimental and engineering capability, to understand dispersion of nanoparticles because their vascular system is more complex than mouse models and because nanoparticle dispersion is so sensitive to in vivo environments. Furthermore, uncertainty cannot be ignored due to the high variation of location-specific vessel characteristics as well as variation across patients. In this paper, a computational method that considers uncertainty is developed to predict nanoparticle dispersion and transport characteristics in the microvasculature with a three step process. First, a computer simulation method is developed to predict blood flow and the dispersion of nanoparticles in the microvessels. Second, experiments for nanoparticle dispersion coefficients are combined with results from the computer model to suggest the true values of its unknown and unmeasurable parameters—red blood cell deformability and red blood cell interaction—using the Bayesian statistical framework. Third, quantitative predictions for nanoparticle transport in the tumor microvasculature are made that consider uncertainty in the vessel diameter, flow velocity, and hematocrit. Our results show that nanoparticle transport is highly sensitive to the microvasculature.     

Decline in zoobenthos densities in the profundal sediments of Lake Mendota (Wisconsin,USA)

High densities of zoobenthos inhabited Lake Mendota's profundal zone in the early 1900s through the mid-1940s. Chaoborus punctipennis was the most abundant organism during the winter, along with moderate densities of Chironomus spp., Pisidium sp., oligochaetes, and Procladius sp. By the early 1950s, Chaoborus punctipennis densities had declined to 10% of former levels, while Chironomus increased significantly. However, by the mid-1960s, Chaoborus, Chironomus, and Pisidium densities had decreased to very low population levels. By 1987–89, Pisidium was no longer found. Zoobenthos that had not decreased from earlier surveys were oligochaetes and Procladius, although further sampling of oligochaetes is needed to confirm current densities. These organisms are the most tolerant of severe anoxia.Four possible reasons for this decline were evaluated: (a) decline in food availability, (b) increase in fish predation, (c) use of toxic insecticides in the drainage basin, and (d) changes in the profundal sediment environment. Based on literature information and long-term data for Lake Mendota, a change in the profundal sediment environment is the most likely explanation for the decline in the less-tolerant zoobenthos species. Although the duration and extent of anoxia in the hypolimnion have not changed since the early 1900s, hypolimnetic ammonia and hydrogen sulfide concentrations apparently have increased as Mendota became more eutrophic after the mid-1940s. However, further study is needed to determine if these higher concentrations or other factors were responsible for the dramatic decline in lake Mendota's profundal zoobenthos.     

Response of Unionid Mussels to Dam Removal in Koshkonong Creek, Wisconsin (USA)

Dam removal is a potentially powerful tool for restoring riverine habitats and communities. However, the effectiveness of this tool is unknown because published data on the effects of dam removal on in-stream biota are lacking. We investigated the effects of a small dam removal on unionid mussels in Koshkonong Creek, Wisconsin (USA). Removal of the dam led to mortality both within the former impoundment and in downstream reaches. Within the former reservoir, mortality rates were extremely high (95&p e r c n t;) due to desiccation and exposure. Mussel densities in a bed 0.5km downstream from the dam declined from 3.80±0.56 musselsm^?2 in fall 2000 immediately after dam removal to 2.60±0.48 musselsm^?2 by summer 2003. One rare species, Quadrula pustulosa, was lost from community. Mortality of mussels buried in deposited silt was also observed at a site 1.7km below the dam. Silt and sand increased from 16.8 and 1.1% of total area sampled in fall 2000 to 30.4 and 15.9%, respectively, in summer 2003. Total suspended sediment concentrations in the water column were always higher downstream from the reservoir than upstream, suggesting that transport and deposition of reservoir sediments likely contributed to downstream mussel mortality. Thus, while benefits of the dam removal included fish passage and restoration of lotic habitats in the former millpond, these changes were brought about at some cost to the local mussel community. Pre-removal assessments of potential ecological impacts of dam removal and appropriate mitigation efforts should be included in the dam removal process to reduce short-term negative ecological effects of this restoration action.     

The influence of arbuscular mycorrhizae and light on Wisconsin (USA) sand savanna understories 1

To explain the complex community composition found in Wisconsin (USA) oak savannas, we investigated potentially interacting effects of light gradients and arbuscular mycorrhizal fungi (AMF) on community composition in the greenhouse, using a fully randomized block experimental design. We used plant species, soil, and AMF from a remnant sand savanna in setting up the experiment, using two light and five AMF treatments. Eleven plant species were seeded into 80 microcosms, and they were grown together for 20 weeks. Plant numbers and biomass were measured, and Simpsons index was calculated for both. Data were analyzed using ANOVA and nonparametric ANOVA. We found significant light effects on biomass and on numbers of four species. There were no treatment effects on Simpsons index, and only Schizachyrium numbers showed a significant AMF effect. These findings are consistent with results from other studies of the sand savanna, and, collectively, these data suggest that plant community composition in this species-rich savanna is not strongly influenced by arbuscular mycorrhizae. This is a novel finding with important implications for understanding interactions between plant and AMF diversity in wild communities.     

Quantifying the influence of topographic position on cougar (Puma concolor) movement in southern California, USA

Large carnivores are frequently used as focal species for landscape-level planning and conservation purposes. Information on cougars Puma concolor , for example, is being used to predict movement corridors and linkage areas in habitats influenced by rapid urbanization. However, animal movement through habitat terrain is a function of multiple factors, including complex topographic features. To assess the use of topographic position during movements by cougars in the Santa Ana Mountain Range of southern California, we analyzed the travel paths of 10–17 radio-tagged individuals monitored during 44 overnight sessions. We examined selection for canyon bottom, gentle slope, steep slope and ridgeline topography at the scale of the movement session and at the scale of the home range. At both scales of selection, our results suggest that traveling or hunting cougars discriminated in their use of topographic position, that canyon bottoms and gentle slopes (<6°) ranked highest in compositional analyses of selection, and that these patterns were not highly confounded by the presence of preferred vegetation types. Ridgelines were used significantly less often than other positions. Our novel method of quantifying availability and use of topographic positions permits the assessment of terrain features, such as canyon bottoms, in facilitating cougar movements. For complex landscapes, models of animal movement should consider the topographic context that motivates patterns of habitat use, and should be developed using data obtained and analyzed at the appropriate spatial and temporal scales.     

Pyric tree spatial patterning interactions in historical and contemporary mixed conifer forests,California, USA

Tree spatial patterns in dry coniferous forests of the western United States, and analogous ecosystems globally, were historically aggregated, comprising a mixture of single trees and groups of trees. Modern forests, in contrast, are generally more homogeneous and overstocked than their historical counterparts. As these modern forests lack regular fire, pattern formation and maintenance is generally attributed to fire. Accordingly, fires in modern forests may not yield historically analogous patterns. However, direct observations on how selective tree mortality among pre‐existing forest structure shapes tree spatial patterns is limited. In this study, we (a) simulated fires in historical and contemporary counterpart plots in a Sierra Nevadan mixed‐conifer forest, (b) estimated tree mortality, and (c) examined tree spatial patterns of live trees before and after fire, and of fire‐killed trees. Tree mortality in the historical period was clustered and density‐dependent, because trees were aggregated and segregated by tree size before fire. Thus, fires maintained an aggregated distribution of tree groups. Tree mortality in the contemporary period was widespread, except for dispersed large trees, because most trees were a part of large, interconnected tree groups. Thus, postfire tree patterns were more uniform and devoid of moderately sized tree groups. Postfire tree patterns in the historical period, unlike the contemporary period, were within the historical range of variability identified for the western United States. This divergence suggests that decades of forest dynamics without significant disturbances have altered the historical means of pyric pattern formation. Our results suggest that ecological silvicultural treatments, such as forest restoration thinnings, which emulate qualities of historical forests may facilitate the reintroduction of fire as a means to reinforce forest structural heterogeneity.     

Quantifying spatial and temporal variation of North Pacific fin whale (Balaenoptera physalus) acoustic behavior

In order to help develop hypotheses of connectivity among North Pacific fin whales, we examine recordings from 10 regions collected in the spring and fall. We develop a Random Forest model to classify fin whale note types that avoids manual note classification errors. We also present a method that objectively quantifies the note and pattern composition of recordings. We find that fin whale recordings near Hawaii have distinctive patterns, similar to those found in other regions in the central North Pacific, suggesting potential migration pathways. Our results are consistent with previous studies that suggest there may be two different populations utilizing the Chukchi Sea and central Aleutians in the fall and mix to some degree in the southern Bering Sea. Conversely, we found little difference between spring and fall recordings in the eastern Gulf of Alaska, suggesting some residency of whales in this region. This is likely due to fine scale similarities of calls among the inshore regions of British Columbia, while offshore areas are being utilized by whales traveling from various distant areas. This study shows how our novel approach to characterize recordings is an objective and informative way to standardize spatial and temporal comparisons of fin whale recordings.     

Quantifying the lag time to detect barriers in landscape genetics

Understanding how spatial genetic patterns respond to landscape change is crucial for advancing the emerging field of landscape genetics. We quantified the number of generations for new landscape barrier signatures to become detectable and for old signatures to disappear after barrier removal. We used spatially explicit, individual‐based simulations to examine the ability of an individual‐based statistic [Mantel’s r using the proportion of shared alleles’ statistic (Dps)] and population‐based statistic (F_ST) to detect barriers. We simulated a range of movement strategies including nearest neighbour dispersal, long‐distance dispersal and panmixia. The lag time for the signal of a new barrier to become established is short using Mantel’s r (1–15 generations). F_ST required approximately 200 generations to reach 50% of its equilibrium maximum, although G’_ST performed much like Mantel’s r. In strong contrast, F_ST and Mantel’s r perform similarly following the removal of a barrier formerly dividing a population. Also, given neighbour mating and very short‐distance dispersal strategies, historical discontinuities from more than 100 generations ago might still be detectable with either method. This suggests that historical events and landscapes could have long‐term effects that confound inferences about the impacts of current landscape features on gene flow for species with very little long‐distance dispersal. Nonetheless, populations of organisms with relatively large dispersal distances will lose the signal of a former barrier within less than 15 generations, suggesting that individual‐based landscape genetic approaches can improve our ability to measure effects of existing landscape features on genetic structure and connectivity.     

The relationship between zooplankton community structure and lake characteristics in temperate lakes (Northern Wisconsin, USA)

Zooplankton community structure can be affected by within-lakeand by watershed ecological factors, including water chemistry(related to landscape position), lake morphology and human activityin the watershed. We hypothesized that all three groups of driverswould be correlated with zooplankton species richness and speciescomposition for lakes in northern Wisconsin. Data collectedfrom 52 lakes allowed us to explore the relationship of zooplanktoncommunity structure with ecological drivers. We found that crustaceanzooplankton species richness was not significantly correlatedwith independent environmental variables derived from PCA ordination,nor with measures of community structure based on NMS ordination.However, species composition was correlated with environmentalgradients. Larger zooplankton species (Daphnia pulicaria, Epischuralacustris, Skistodiaptomus oregonensis, Mesocyclops americanus)occurred in large and deep lakes low in the landscape gradient,whereas the smaller species Ceriodaphnia dubia, Daphnia retrocurvaand Leptodiaptomus minutus tended to occur high in the landscape.This shift in species composition was correlated with increasedconductivity, primary productivity and the hypolimnetic refugescharacteristic of larger deeper lakes lower in the landscape.Riparian housing development and littoral zone habitat (measuredas building density and by abundance of logs in littoral zones)were not correlated with zooplankton community structure. Inthese relatively low-impact lakes, natural drivers are stillthe most significant determinants of zooplankton community structure.     

The influence of landscape position on lakes in northern Wisconsin

1. Using data from the North Temperate Lakes Long-Term Ecological Research site in northern Wisconsin, we present a series of examples illustrating how landscape setting can influence the static and dynamic aspects of many physical, chemical and biological properties of lakes. 2. One important landscape attribute is the hydrologic position of a lake within the regional flow regime. Lake position determines the relative importance of groundwater and precipitation input to a lake, with lakes high in the landscape receiving a greater proportion of their input waters from precipitation than lakes lower in the landscape. Landscape position is strongly correlated with the concentration of base cations such as calcium and magnesium. 3. Landscape position also influences how lakes respond to drought conditions. Lakes high in the landscape responded to a 4-year drought with decreases in calcium mass, whereas lakes low in the landscape increased in mass of calcium. During extended dry conditions, these differential responses of lakes suggest that lakes already low in calcium (i.e. in a high position in the flow system) will have further reductions in calcium concentrations. These reductions could decrease the number of lakes offering suitable habitat for organisms such as crayfish and snails whose distributions are limited by calcium. 4. Landscape position also affects silica concentrations in lakes, with lakes low in the landscape having silica concentrations up to three orders of magnitude greater than lakes high in the landscape. Differences in silica concentration affect robustness of freshwater sponge spicules which can potentially alter some aspects of the dynamics of littoral zone food webs. 5. Landscape position can influence the vertical distribution of primary production. Concentrations of dissolved organic carbon are affected by landscape setting and can influence vertical light penetration, thus affecting the depth at which primary production can occur. 6. Lake area and fish species richness are correlated with landscape position: larger, species-rich lakes are low in the landscape, whereas smaller lakes with fewer species tend to be high in the landscape. 7. By taking a landscape-scale view, in addition to the more usual lake-specific view, it is possible to reach a more robust understanding of lake dynamics and avoid some of the problems associated with extrapolating from single lake results.     

The influence of arbuscular mycorrhizae and light on Wisconsin (USA) sand savanna understories 2. Plant competition

Wisconsin (USA) oak savannas are endangered plant communities that have remarkably high plant species diversity. To investigate factors underlying this richness, we experimentally investigated the potentially interacting effects of light gradients and arbuscular mycorrhizal fungi (AMF) on plant competition in the greenhouse, using a fully randomized block design. We used four plant species, soil, and AMF from a remnant sand savanna, under two light and five AMF treatments. Plants were grown four per pot under two competition treatments (either one or four species per pot) for 20 weeks. Using ANOVA, we found that all species showed significant treatment effects on total and shoot biomass, primarily due to differences in competition and light, less to AMF. However, effects were the opposite of predictions. Putatively mycorrhizal plants showed neutral to negative responses to AMF, and a nonmycorrhizal species outcompeted AMF species in infected pots. We concluded that our experimental setup of small pots, sandy soil, and long growing period had induced parasitism by the AMF on susceptible hosts. This unexpected result is consistent with field data from the sand savanna, and may help explain how nonmycorrhizal plants can compete successfully with AMF species in established, species-rich communities.