More

Overlaying Line (Vector) Layer on Raster Zones?


I am using ArcGIS 10

I have a raster layer with 3 values that make up zones. For the purpose of this example, lets say these zones are forest, grass, and open water. My whole raster is made up of these three values. I also have a line layer (it is currently a vector file but I'm fine converting to raster if I need to). This line layer is streams. I want to attach this stream layer to the raster layer based on zones. so I will end up with 6 values (Forest with stream, forest without stream, grass with, grass without, open water with, open water without). I basically want to do an overlay, but apply the stream value to the whole zone and not just a pixel. If a stream falls into a forest pixel, I want all neighboring forest pixels to also have that stream.

I really want to avoid converting the raster to vector. I looked into the zonal function tools but they seem to only give me statistics and values based on zones rather than overlay two layers based on zones.


Biological integrity in urban streams: Toward resolving multiple dimensions of urbanization

Most studies of urban streams have relied on single variables to characterize the degree of urbanization, which may not reflect interactions among features of urban landscapes adequately. We report on an approach to the characterization of urbanization effects on streams that used principal components analysis and multiple regression to explore the combined, interactive effects of land use/land cover, human population demography, and stream habitat quality on an index of biological integrity (IBI) of fish communities. Applied to a substantially urbanized region in northeast OH, USA, the analysis demonstrated the interactive nature of urbanization effects. Urban land use and stream habitat quality were significant predictors of IBI, but were no better than and, in some cases, poorer predictors than other gradients and interactions among gradients. High integrity sites were characterized by low forest cover and high grassland cover at sub-catchment scale, but high forest cover within a 500 m radius local zone of the sample point, conditions often found in protected parklands in the region. The analysis also indicated that variability in stream habitat quality was unrelated to landscape or demographic features, a result we attribute to the interaction between the geological and urbanization histories of the region.


Materials and Methods

Data Generation

We collected and vouchered animals from across the putative hybrid zone over a 10-km-wide transect in May 2017 and “pure” individuals, for species delimitation, from allopatric populations of each species across Puerto Rico between 2012 and 2017. We identified putative hybrids via intermediate, and anomalous, color and patterning schemes across the dorsum of the animals (N = 18). Our sampling reflected the geographic extremes where we could definitively categorize individuals as “pure” S. nicholsi or S. townsendi on either side of the putative hybrid zone. A single hybrid individual (BJP016 – locality H) was sequenced from an autotomized tail (animal not captured), and thus was unidentifiable in the field (adjusted sample sizes: hybrids N = 19, S. nicholsi N = 19, S. townsendi N = 25). Uneven sampling across the putative hybrid zone is due to biologically relevant factors (uneven habitat) and land-use factors specifically, the Military Training Center (Fort Allen) that lies in the middle of the sampling area. We extracted genomic DNA from tail or liver tissue using the Qiagen ® DNeasy Blood and Tissue Kit. We generated sequence data for all 63 individuals, using PCR amplification and subsequent, single-pass Sanger sequencing in both directions (Genewiz ® ). All sequences were accessioned to GenBank ( Supplementary Table 1 ), and for taxonomic assignment of hybrid animals submitted to GenBank, we standardized by using the animals’ mitochondrial assignment. We amplified fragments of 7 genetic markers, 1 mitochondrial gene: NADH dehydrogenase subunit 2 (ND2) and 6 nuclear markers: Oocyte Maturation Factor mos (CMOS), Death Inducer-Obliterator 1 (DIDO), Microtubule-Actin Crosslinking Factor 1 (MACF), Microtubule Associated Protein 2 (MAP2), Protein Tyrosine Phosphatase, Non-receptor type 12 (PTPN12), and Recombination Activating Gene 1 (RAG1) primers and associated references for each marker are provided ( Supplementary Table 2 ). Raw sequence reads were assembled using Geneious ® (v10.2.2) ( Kearse et al. 2012). We identified heterozygous sites using the “Find Heterozygotes” function in Geneious ® and verified putative heterozygous sites by examining trace chromatograms for dual base-called sites that were equal in height and shorter than neighboring sites. We aligned DNA sequences using MUSCLE software ( Edgar 2004) and refined alignments by eye. We phased allelic variants for each nuclear gene using PHASE software ( Stephens et al. 2001), with default settings, implemented in DNAsp (v5.10.1) ( Librado and Rozas 2009).

Species Phylogeny and Species Delimitation

To generate a species tree in a coalescent framework, we utilized a subset of our aforementioned dataset (ND2, CMOS, PTPN12, RAG1), and generated additional sequence data for mitochondrial ribosomal subunit 16S and 3 additional closely related taxa from within this clade, S. monensis, S. levinsi, and S. klauberi for all loci ( Supplementary Table 2 ). Sphaerodactylus klauberi was used as an outgroup for all phylogenetic analyses. The best-partitioning scheme for each gene was determined using PartitionFinder2 (v2.1.1) ( Guindon et al. 2010 Lanfear et al. 2012, 2017) on the CIPRES cluster ( Miller et al. 2010) ( Supplementary Table 3 ). Each locus consisted of a single data partition except ND2, where each codon had its own partition. Both mitochondrial loci, 16S and ND2, were combined into a single tree partition. Models of sequence evolution for each locus and partition are listed in Supplementary Table 3 . We generated a species tree using the StarBEAST2 package (v0.13.5) ( Ogilvie et al. 2017) in BEAST2 (v2.4.6) ( Bouckaert et al. 2014), also on the CIPRES cluster ( Miller et al. 2010). Each locus utilized an uncorrelated lognormal clock and a birth-death model with other priors set as default. We ran 3 independent chains of 10 × 10 8 mcmc iterations, with a 10% burnin, and examined likelihood values for convergence using Tracer (v1.6) ( Rambaut et al. 2018). Tree files were compiled using LogCombiner and the final tree was generated in TreeAnnotator.

To further resolve the species-level relationships within this clade and perform an initial assessment of the validity of S. townsendi and S. nicholsi as distinct species, we analyzed our postburnin species trees to identify the frequency of a S. townsendi + S. nicholsi clade. We calculated the posterior probability of this hypothesis by filtering postburnin species trees that were consistent with a topology that constrained an S. townsendi + S. nicholsi clade, exclusive of S. levinsi and S. monensis, using Phylogenetic Analysis Using Parsimony (PAUP*) (v 4.0a157) ( Swofford 2002).

To identify whether S. nicholsi and S. townsendi are genetically distinct populations relative to each other, we conducted statistical species delimitation under the multispecies coalescent to test for species-level divergence in this clade using STACEY (v1.2.4) ( Jones 2017). We conducted STACEY analysis using the previously described StarBEAST2 dataset, with all taxa and partitions conserved in both analyses ( Supplementary Table 3 ). All priors were left default unless specifically stated below. In accordance with program documentation and additional specifications outlined by Barley et al. (2018), we provided an exponential distribution with a mean of 0.1 for the “popPriorScale” parameter, a lognormal distribution with a mean of 5 and a standard deviation of 2 to the “bdcGrowthRate” prior, and the “collapseWeight” was provided a uniform distribution with the lower and upper bounds set at 0 and 1, respectively ( Barley et al. 2018). In addition, each partition was provided an independent strict molecular clock, with rate priors calculated from a log-normal distribution that were given a mean of 0 and standard deviation of 1 ( Barley et al. 2018). We ran 3 independent chains of 5.0 × 10 7 mcmc repetitions, sampling every 5000 trees, and compared likelihood values from trace files using Tracer (v1.6) ( Rambaut et al. 2018). We combined tree files using LogCombiner and analyzed the resulting 30,000 trees using the SpeciesDelimitationAnalyzer (SpeciesDA) (v1.8). We used a burnin of 5000 and a collapse-height of 0.0001 to calculate our final species delimitation result. To corroborate the results generated using STACEY and SpeciesDA, we analyzed a dataset of nuclear-only loci using BPP software (v4.0) ( Flouri et al. 2018). We analyzed CMOS, PTPN12, and RAG1 genes for S. nicholsi, S. townsendi, S. levinsi, S. monensis, and S. klauberi using a specified guide tree from our StarBeast2 analysis. To examine an array of biological scenarios, we used 3 different configurations of population size (inverse-gamma θ = a, b) and divergence time (inverse-gamma τ = a, b) priors to begin our parameter estimation ( Leaché and Fujita 2010). In configuration 1, we assumed “medium” Ne (θ = 3, 0.002) with “medium” divergence time (τ = 3, 0.03) in configuration 2, we assumed “small” Ne (θ = 3, 0.0002) with “recent” divergence time (τ = 3, 0.003) in configuration 3, we assumed “large” Ne (θ = 3, 0.02) with “long” divergence time (τ = 3, 0.3). This allowed us to interrogate the effects of various biological changes within the system on species delimitation analysis. We ran each mcmc chain for 5 × 10 5 , sampling every 5, with a 10% burnin. We ran 2 independent mcmc chains for each configuration and checked log files’ likelihood values for convergence using Tracer (v1.6) ( Rambaut et al. 2018). In addition to statistical species delimitation methods, we generated table of uncorrected P-distances between lineages of the mitochondrial gene ND2 using MEGA7 ( Kumar et al. 2016) to compare across other recognized gecko species ( Supplementary Table 4 ).

Genetic Cline Analyses

To assign alleles in hybrids to their respective parental species, we constructed maximum-likelihood (ML) gene trees of phased alleles using RAxML-HPC BlackBox (v8.2.10) ( Stamatakis 2014) on the CIPRES cluster ( Miller et al. 2010). These gene trees were rooted using S. klauberi, with bootstrap support necessary for one species-specific clade (≥70) to the exclusion of the other (i.e., support for all individuals within a group excludes all members of other group). All species-specific allelic variants were condensed into binary allelic assignment, where all individuals were either “pure” S. townsendi (0), “pure S. nicholsi” (1), and/or, for nuclear markers, heterozygous (0.5) assignments.

To calculate genetic clines, we calculated the mean value of the individuals at each locality to produce a per-locality allele frequency for each locus. We plotted these values along a collapsed, 1-dimensional (longitudinal) transect using the hzar package (v0.2–5) ( Derryberry et al. 2014) in R ( R Core Team 2016). To find the best-fit cline model, we conducted a series of model-testing analyses in hzar. First, to identify the best-fit model for each locus we tested 4 different potential models using corrected Akaike information criterion (AICc). Model I maintained that each interval was fixed at either 0 or 1 without an exponential tail model II maintained that each interval was free to fluctuate based on the data (was not fixed at 0 or 1), also without an exponential tail. Between these models, model I was favored, thus, models III and IV maintained fixed intervals. To test for the presence of asymmetric introgression, we model III tested for the presence of a left-side tail and model IV a right-side tail ( Table 1). Further, to test for discordance between mitochondrial and nuclear markers (cline movement), we generated models V and VI, which constrained the cline center and width, respectively, of our mitochondrial cline to the average values of our nuclear data for each. For all model-testing analyses, a model with an AICc difference of >2 was considered a significantly better fit to the data than the alternate model ( Leaché et al. 2017).

(A) Corrected AICc values for each cline-fitting model tested at each locus. Score differences of >2 indicate a significantly greater fit to the model, and the best-fit model for each locus is bolded (* indicates 3 best-fit models to the data). (B) Cline center and width ranges for each locus. Cline center is respective of the longitudinal cline (i.e., along a 0–10 km axis). Cline widths are kilometer ranges irrespective of their location along the cline (i.e., distance wide)

A) . AICc . mtDNA . CMOS . MACF . PTPN12 . RAG1 .
NULL model60.731129.450131.900136.889130.606
Model I (fixed, no tails) 7.894*5.01019.18116.16513.883
Model II (free, no tails) 13.149 11.480 24.401 22.003 20.807
Model III (fixed, left tail) 12.326 9.129 23.419 20.294 18.166
Model IV (fixed, right tail) 12.526 9.159 23.385 20.469 18.167
Model V (fixed, no tails constrained center) 7.737*N/A. N/A. N/A. N/A.
Model VI (fixed, no tails constrained width) 7.734*N/A. N/A. N/A. N/A.
B) Cline Ranges
Cline centers 3.759–4.432 4.284–4.496 4.306–4.597 4.246–4.456 4.212–4.485
Cline widths 0.5969–2.794 0.4402–1.044 0.6806–1.591 0.4652–1.272 0.5311–1.505
A) . AICc . mtDNA . CMOS . MACF . PTPN12 . RAG1 .
NULL model60.731129.450131.900136.889130.606
Model I (fixed, no tails) 7.894*5.01019.18116.16513.883
Model II (free, no tails) 13.149 11.480 24.401 22.003 20.807
Model III (fixed, left tail) 12.326 9.129 23.419 20.294 18.166
Model IV (fixed, right tail) 12.526 9.159 23.385 20.469 18.167
Model V (fixed, no tails constrained center) 7.737*N/A. N/A. N/A. N/A.
Model VI (fixed, no tails constrained width) 7.734*N/A. N/A. N/A. N/A.
B) Cline Ranges
Cline centers 3.759–4.432 4.284–4.496 4.306–4.597 4.246–4.456 4.212–4.485
Cline widths 0.5969–2.794 0.4402–1.044 0.6806–1.591 0.4652–1.272 0.5311–1.505

(A) Corrected AICc values for each cline-fitting model tested at each locus. Score differences of >2 indicate a significantly greater fit to the model, and the best-fit model for each locus is bolded (* indicates 3 best-fit models to the data). (B) Cline center and width ranges for each locus. Cline center is respective of the longitudinal cline (i.e., along a 0–10 km axis). Cline widths are kilometer ranges irrespective of their location along the cline (i.e., distance wide)

A) . AICc . mtDNA . CMOS . MACF . PTPN12 . RAG1 .
NULL model60.731129.450131.900136.889130.606
Model I (fixed, no tails) 7.894*5.01019.18116.16513.883
Model II (free, no tails) 13.149 11.480 24.401 22.003 20.807
Model III (fixed, left tail) 12.326 9.129 23.419 20.294 18.166
Model IV (fixed, right tail) 12.526 9.159 23.385 20.469 18.167
Model V (fixed, no tails constrained center) 7.737*N/A. N/A. N/A. N/A.
Model VI (fixed, no tails constrained width) 7.734*N/A. N/A. N/A. N/A.
B) Cline Ranges
Cline centers 3.759–4.432 4.284–4.496 4.306–4.597 4.246–4.456 4.212–4.485
Cline widths 0.5969–2.794 0.4402–1.044 0.6806–1.591 0.4652–1.272 0.5311–1.505
A) . AICc . mtDNA . CMOS . MACF . PTPN12 . RAG1 .
NULL model60.731129.450131.900136.889130.606
Model I (fixed, no tails) 7.894*5.01019.18116.16513.883
Model II (free, no tails) 13.149 11.480 24.401 22.003 20.807
Model III (fixed, left tail) 12.326 9.129 23.419 20.294 18.166
Model IV (fixed, right tail) 12.526 9.159 23.385 20.469 18.167
Model V (fixed, no tails constrained center) 7.737*N/A. N/A. N/A. N/A.
Model VI (fixed, no tails constrained width) 7.734*N/A. N/A. N/A. N/A.
B) Cline Ranges
Cline centers 3.759–4.432 4.284–4.496 4.306–4.597 4.246–4.456 4.212–4.485
Cline widths 0.5969–2.794 0.4402–1.044 0.6806–1.591 0.4652–1.272 0.5311–1.505

Species Distribution Models and Simulations

To estimate the climatic factors involved in the generation and maintenance of this hybrid zone, we constructed independent species distribution models for S. nicholsi and S. townsendi using Maxent (v3.41) ( Phillips et al. 2006, 2017 Phillips and Dudík 2008). Locality information was collected by the authors (BJP, JT, JDD, and TG, personal observation), Murphy et al. (1984), and VertNet (accessed 30 January 2018) ( Supplementary Table 5 ). GIS layers were assembled in QGIS (v3.2) ( QGIS Development Team 2009). We procured the base vector layer (“Puerto Rico administrative area”) from the Global Administrative Database (https://gadm.org). All 19 BioClim layers were acquired from WorldClim 2 ( Fick and Hijmans 2017), current 1970–2000, at 30 arc-seconds (

1 km) ( Supplementary Table 6 ). Landsat tree cover was parsed and stitched together to create a vegetative continuous field in QGIS (v3.2). Resolution was set to 30 m 2 per pixel. Vegetation was estimated as tree cover of horizontal wooded cover greater than 5 m in height. Map data were derived from Landsat-5 Thematic Mapper and Landsat-7 Enhanced Thematic Mapper Plus (ETM+) ( Sexton et al. 2013). Finally, topographic raster group was derived from Shuttle Radar Topography Mission – February 2000 (SRTM) ( Farr et al. 2007). Elevation layer was taken from SRTM at 30 m 2 resolution. Roughness, slope (using Horn (1981) algorithm [ Fleming and Hoffer 1979 Ritter 1987]), aspect, hillshade, terrain position index, and terrain ruggedness index (TRI) were derived from SRTM using R package raster 2.6–7 ( Hijmans and van Etten 2012). WorldClim 2 layers were trimmed and bound to Puerto Rico administrative boundaries in QGIS, then transformed to a raster in ASCII format for Maxent. Since locality cover for each species is high (locality data reasonably represent the ranges of these species) and broadly established within the putative range, no assumption was made for limiting presence data by arbitrary omission. Our niche model analysis was conducted with presence only data ( Fielding 2002 Veloz 2009) to predict regions where species distributions could extend without extraneous factors ( Supplementary Table 6 ). One hundred replicates with a random seed were conducted on S. nicholsi and S. townsendi using cross-validation to estimate robustness of prediction. A maximum number of background points were set to 10,000. Environmental variable importance used the Jackknife method in Maxent, and output format was written in the Cloglog. Once written, we determined whether or not the machine-learning “receiver operating curve” (ROC) fit the data by calculating the “area under the curve” (AUC) ( Supplementary Table 6 ).

To understand the ecological dynamics of the hybrid zone between S. nicholsi and S. townsendi, we conducted a model-based analysis of the niche space overlap in NicheA ( Qiao et al. 2016). This tests fundamental niche similarity between S. nicholsi and S. townsendi to determine how the dynamics of each niche influence the distribution of each species. By characterizing the fundamental niches between species, we can postulate the stability of a species range by the conserved factors within the environmental and geographic overlap ( Stigall 2014). NicheA allows for visualization of Hutchinsonian duality ( Pulliam 2000 Colwell and Rangel 2009) between environmental (E) and geographic (G) spaces across species distribution, where the niche is a multidimensional volume with a set of defined characters (E-space) influenced by the physical conditions that dictate those conditions (G-space) ( Colwell and Rangel 2009). In our case, we calculated the overlap between parental species to elucidate how a hybrid zone may form based on the shared suitability of niche space between S. nicholsi and S. townsendi. Thus, we used these niche simulations to estimate 1) fundamental niche space of S. nicholsi and S. townsendi and 2) to calculate the niche overlap between the 2 species. The niche simulations employed in NicheA software use a background cloud (BC) to constrain the environmental space within the species range (Puerto Rico’s geography in this study) within 3-dimensional space. Each species’ niche is then simulated into a hypervolume, a set of defined characters that influence the niche of the species, within the constraints set by the BC. A minimum volume ellipsoid (MVE) is the simulated fundamental niche of a species within the constrained set of characters of the BC defined by the specific characters inherent to the species itself.


Watch the video: Πολλαπλασιασμός Αριθμού με Διάνυσμα Ασκήσεις 22 (October 2021).