Suburban frog ponds: mapping local land cover yields different estimates of landscape composition than the National Land Cover Database
- Published
- Accepted
- Subject Areas
- Conservation Biology, Ecology, Environmental Sciences
- Keywords
- exurban, houses, GIS, land use, landscape ecology, lawn, periurban, remote sensing, residential neighborhoods, urban ecology
- Copyright
- © 2014 Lambert
- Licence
- This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ PrePrints) and either DOI or URL of the article must be cited.
- Cite this article
- 2014. Suburban frog ponds: mapping local land cover yields different estimates of landscape composition than the National Land Cover Database. PeerJ PrePrints 2:e570v1 https://doi.org/10.7287/peerj.preprints.570v1
Abstract
Suburban neighborhoods are rapidly spreading globally. As such, there is an increasing need to study the environmental and ecological effects of suburbanization. At large spatial extents, from county-level to global, remote sensing-derived land cover data, such as the National Land Cover Dataset (NLCD), have yielded insight into patterns of urbanization and concomitant large-scale ecological patterns in response. However, the components of suburban land cover (houses, yards, etc.) are dispersed throughout the landscape at a finer scale than the relatively coarse grain size (30m pixels) of NLCD may be able to detect. Our understanding of ecological processes in heterogeneous landscapes is reliant upon the accuracy and resolution of our measurements as well as the scale at which we measure the landscape. Analyses of ecological processes along suburban gradients are restricted by the currently available data. As ecologists are becoming increasingly interested in describing phenomena at spatial extents as small as individual households, we need higher-resolution landscape measurements. Here, I describe a simple method of translating the components of suburban landscapes into finer-grain, local land cover (LLC) data in GIS. Using both LLC and NLCD, I compare the suburban matrix surrounding ponds occupied by two different frog species. I illustrate large discrepancies in Forest, Yard, and Developed land cover estimates between LLC and NLCD, leading to markedly different interpretations of suburban landscape composition. NLCD, relative to LLC, estimates lower proportions of forest cover and higher proportions of anthropogenic land covers in general. These two land cover datasets provide surprisingly different descriptions of the suburban landscapes, potentially affecting our understanding of how organisms respond to an increasingly suburban world. LLC provides a free and detailed fine-grain depiction of the components of suburban neighborhoods and will allow ecologists to better explore heterogeneous suburban landscapes at multiple spatial scales.
Author Comment
This will be a submission to PeerJ for formal review
Supplemental Information
Fig. A
Percent cover of Forest, Yard, and Developed covers using LLC (a,b) and NLCD (c,d) within a 50m buffer surrounding wood frog ponds (a,c) and green frog ponds (b,d). LLC and NLCD data can be compared for a given. For wood frogs, LLC describes Developed land cover that is not described by NLCD and NLCD estimates more Yard cover than LLC. For green frogs, LLC describes Forest cover where NLCD does not and NLCD more Yard and Developed cover than does LLC. White bars are Forest cover, gray bars are Yard cover, and black bars are Developed land cover. X-axis number labels represent the same pond for top and bottom panels.
Fig. A2
Relationships between Forest (a,b), Yard (c,d), and Developed (e,f) land covers between LLC (x-axis) and NLCD (y-axis) estimates. Land cover data are estimated both within a 50m buffer (a,c,e) and a 200m buffer (b,d,f) surrounding ponds. Both x- and y-axes are plotted on the same scale for a given panel. Green triangles are wood frog ponds and red circles are green frog ponds. When significant, coefficients of determination are presented. At the 50m scale, LLC estimated Forest cover for green frogs (a) and estimated Developed cover for wood frogs (b) that were entirely absent in NLCD data.
Suburban Pond Land Cover Raw Data _ LLC and NLCD
Land cover data surrounding eight suburban wood frog and eight suburban green frog ponds in Connecticut. Pond area is presented in square meters and all land cover values are presented as percentages. All columns with “NLCD” in the header are National Land Cover database information; 2011 indicates that the data are from the NLCD dataset for the year 2011. All columns with “LLC” in the header are Local Land Cover data information. If a column has “50” in the header it indicates that data are percent cover within a 50m radius of the pond and “200” indicates that data are percent cover within a 200m radius. For NLCD, Forest and Developed covers are self-explanatory; “Open Space” cover indicates yard cover. For LLC, lawn, building, water, and forest are self-explanatory. Garden represents all non-lawn yard cover in LLC. Also for LLC, “YardSum” is garden and lawn categories summed and “DevSum” is building and paved categories summed.