Habitat alteration and habitat fragmentation differentially affect beta diversity of stream fish communities

Context

The cumulative impact of broad scale environmental change includes altered land-cover and fragmentation. Both altered land-cover and fragmentation have a negative effect on species diversity, but the scale they act on may differ because land-cover alters environmental characteristics, whereas fragmentation alters movement among sites.

Objectives

We evaluated the scale specific effects of land-cover, fragmentation, and habitat size on alpha and beta diversity (total, turnover, and nestedness).

Methods

Stream fish communities were sampled across five urbanizing watersheds. Generalized mixed linear models were used to test how diversity (alpha and beta) is affected by land-cover, connectivity, and habitat size. Indices of land-cover were calculated from correspondence analyses on land-cover data, fragmentation was estimated with the dendritic connectivity index, and habitat size was calculated as the length of the stream segment (alpha diversity) or the length of the stream network (beta diversity).

Results

Alpha diversity was most strongly related to land-cover variables associated with urban development and agriculture (negative relationship with urbanization). Whereas, beta diversity was most strongly influenced by habitat size (positive relationship) and fragmentation (positive relationship). Turnover was positively correlated with fragmentation and habitat size, whereas species loss was negatively correlated with habitat size.

Conclusions

Land-cover has a larger effect on alpha diversity because it alters the environmental conditions at a site, whereas fragmentation has a larger effect on beta diversity because it affects the movement of individuals among sites. Assessing the cumulative impact of environmental change requires a multiscale approach that simultaneously considers alpha and beta diversity.

     

A geography of ecosystem vulnerability

Land-cover change and the subsequent potential loss of natural resources due to conversion to anthropogenic use is regarded as one of the more pervasive environmental threats. Population and road data were used to generate interpolated surfaces of land demand across a large region, the mid-Atlantic states of Pennsylvania, Delaware, Maryland, Virginia, and West Virginia. The land demand surfaces were evaluated against land-cover change, as estimated using temporal decline in Normalized Difference Vegetation Index (NDVI). In general, the interpolated surfaces exhibited a plateau along the eastern seaboard that sank to a valley in the center of the study area, and then rose again to a plateau in the west that was of overall lower height than the plateau on the eastern seaboard. The spatial pattern of land-cover change showed the same general pattern as the interpolated surfaces of land demand. Correlations were significant regardless of variations used to generate the interpolated surfaces. The results suggest that human activity is the principal agent of land-cover change at regional scales in this region, and that natural resources that change as land cover changes (e.g., water, habitat) are exposed to a gradient of vulnerability that increases from west to east.     

Temporal change in fragmentation of continental US forests

Changes in forest ecosystem function and condition arise from changes in forest fragmentation. Previous studies estimated forest fragmentation for the continental United States (US). In this study, new temporal land-cover data from the National Land Cover Database (NLCD) were used to estimate changes in forest fragmentation at multiple scales for the continental US. Early and late dates for the land-cover change data were ca. 1992 and ca. 2001. Forest density was used as a multi-scale index of fragmentation by measuring the proportion of forest in neighborhoods ranging in size from 2.25 to 5314.41 ha. The multi-scale forest density maps were classified using thresholds of 40% (patch), 60% (dominant), and 90% (interior) to analyze temporal change of fragmentation. The loss of dominant and interior forest showed distinct scale effects, whereas loss of patch forest was much less scale-dependent. Dominant forest loss doubled from the smallest to the largest spatial scale, while interior forest loss increased by approximately 80% from the smallest to the second largest spatial scale, then decreased somewhat. At the largest spatial scale, losses of dominant and interior forest were 5 and 10%, respectively, of their ca. 1992 amounts. In contrast, patch forest loss increased by only 25% from the smallest to largest spatial scale. These results indicate that continental US forests were sensitive to forest loss because of their already fragmented state. Forest loss would have had to occur in an unlikely spatial pattern in order to avoid the proportionately greater impact on dominant and interior forest at larger spatial scales.     

Impact of data integration technique on historical land-use/land-cover change: Comparing historical maps with remote sensing data in the Belgian Ardennes

Historical reconstructions of land-use/cover change often require comparing maps derived from different sources. The objective of this study was to measure land-use/cover changes over the last 225 years at the scale of a Belgian landscape, Lierneux in Ardennes, on the basis of a heterogeneous time series of land cover data. The comparability between the land-cover maps was increased following a method of data integration by map generalisation. Two types of time series were built by integrating the maps either by reference to the initial map of the time series or by pair of successive maps. Land-cover change detection was performed on the initial time series without data integration and on the two types of integrated time series. Results reveal that land cover and landscape structure have been subject to profound changes in Lierneux since 1775, with an annual rate of change at the landscape level of up to 1.40%. The major land-cover change processes observed are expansion of grasslands-croplands and reforestation with coniferous species, leading to amore fragmented landscape structure. The annual rates of land-cover change estimated from integrated data are significantly different from the annual rates of change estimated without a prior integration of the data. There is a trade-off between going as far back in time as possibleversus performing change detection as accurately as possible. This revised version was published online in July 2006 with corrections to the Cover Date.     

Forest cover changes in the Prignitz region (NE Germany) between 1790 and 1960 in relation to soils and other driving forces

Decadal to centennial land-cover changes are important drivers of many environmental issues, including biodiversity, biogeochemical cycles and, especially, the global carbon balance. In general, changes are well documented over only a few decades. Studies of land-cover changes and its drivers over centuries are rare. Therefore, the main objectives of this study are (1) to trace the development of the actual pattern of forest-open land over 170 years, and (2) to associate land-cover classes with site conditions (soils) as well as with other driving forces during three periods (1790–1838, 1838–1870, and 1870–1960). For these purposes, we used a combined approach of GIS-techniques and historical reconstructions from archives. The shifts of percentages for established forests, afforestation, clearings and open land on different soils were checked using a chi square test. From the archives, we obtained information on demographic, political/institutional and economic/technological factors, which are assumed to be drivers for past land-cover changes. Percentages of most land-cover classes hardly differed between the periods. However, established forests remained mainly on sandy soils and, to a large extend, afforestation was realised on sandy soils. Clearings reached high percentages on fluvial sands and organic sediments in the early period. A complex of demographic, political/institutional and economic/technological factors also had a considerable impact on land use/cover change in the Prignitz region. Thus, in addition to the strong association of land-cover classes with soils, our study demonstrated that other driving forces, i.e. political and economic factors, played an important role in the full understanding of land use from the past to the present.     

Response of forest soil properties to urbanization gradients in three metropolitan areas

We investigated the effects of urban environments on the chemical properties of forest soils in the metropolitan areas of Baltimore, New York, and Budapest. We hypothesized that soils in forest patches in each city will exhibit changes in chemistry corresponding to urbanization gradients, but more strongly with various urban metrics than distance to the urban core. Moreover, differences in parent material and development patterns would differentially affect the soil chemical response in each metropolitan area. Results showed that soil chemical properties varied with measures of urban land use in all three cities, including distance to the urban core, which was an unexpected result. Moreover, the results showed that the spatial extent and amount of change was greater in New York than in Baltimore and Budapest for those elements that showed a relationship to the urbanization gradient (Pb, Cu, and to a lesser extent Ca). The spatial relationship of the soil chemical properties to distance varied from city to city. In New York, concentrations of Pb, Cu, and Ca decreased to approximately background concentrations at 75 km from the urban core. By contrast, concentrations of these elements decreased closer to the urban core in Baltimore and Budapest. Moreover, a threshold was reached at about 75% urban land use above which concentrations of Pb and Cu increased by more than twofold relative to concentrations below this threshold. Results of this study suggest that forest soils are responding to urbanization gradients in all three cities, though characteristics of each city (spatial pattern of development, parent material, and pollution sources) influenced the soil chemical response.     

Temporal change in forest fragmentation at multiple scales

Previous studies of temporal changes in fragmentation have focused almost exclusively on patch and edge statistics, which might not detect changes in the spatial scale at which forest occurs in or dominates the landscape. We used temporal land-cover data for the Chesapeake Bay region and the state of New Jersey to compare patch-based and area–density scaling measures of fragmentation for detecting changes in the spatial scale of forest that may result from forest loss. For the patch-based analysis, we examined changes in the cumulative distribution of patch sizes. For area–density scaling, we used moving windows to examine changes in dominant forest. We defined dominant forest as a forest parcel (pixel) surrounded by a neighborhood in which forest occupied the majority of pixels. We used >50% and ≥60% as thresholds to define majority. Moving window sizes ranged from 2.25 to 5,314.41 hectares (ha). Patch size cumulative distributions changed very little over time, providing no indication that forest loss was changing the spatial scale of forest. Area–density scaling showed that dominant forest was sensitive to forest loss, and the sensitivity increased nonlinearly as the spatial scale increased. The ratio of dominant forest loss to forest loss increased nonlinearly from 1.4 to 1.8 at the smallest spatial scale to 8.3 to 11.5 at the largest spatial scale. The nonlinear relationship between dominant forest loss and forest loss in these regions suggests that continued forest loss will cause abrupt transitions in the scale at which forest dominates the landscape. In comparison to the Chesapeake Bay region, dominant forest loss in New Jersey was less sensitive to forest loss, which may be attributable the protected status of the New Jersey Pine Barrens.     

Forest conservation and land development in Puerto Rico

In the Caribbean island of Puerto Rico, rapid land-use changes over the past century have included recent land-cover conversion to urban/built-up lands. Observations of this land development adjacent to reserves or replacing dense forest call into question how the changes relate to forests or reserved lands. Using existing maps, this study first summarizes island-wide land-cover change between 1977-78 and 1991-92. Then, using binomial logit modeling, it seeks evidence that simple forest cover attributes, reserve locations, or existing land cover influence land development locations. Finally, this study quantifies land development, reserve protection and forest cover by ecological zone. Results indicate that 1) pasture is more likely to undergo land development than shrubland plus forest with low canopy density, 2) forest condition and conservation status appear unimportant in that development locations neither distinguish between classes of forest canopy development nor relate to forest patch size or reserve proximity, and 3) most land development occurs in the least-protected ecological zones. Outside the boundaries of strictly protected forest and other reserves, accessibility, proximity to existing urban areas, and perhaps desirable natural settings, serve to increase land development. Over the coming century, opportunities to address ecological zone gaps in the islands forest reserve system could be lost more rapidly in lowland ecological zones, which are relatively unprotected.This revised version was published online in May 2005 with corrections to the Cover Date.     

Accessibility as a determinant of landscape transformation in western Honduras: linking pattern and process

This study evaluates the relationship between landscape accessibility and land cover change in Western Honduras, and demonstrates how these relationships are influenced by social and economic processes of land use change in the region. The study area presents a complex mosaic of land cover change processes that involve approximately equal amounts of reforestation and deforestation. Landsat Thematic Mapper (TM) satellite imagery of 1987, 1991 and 1996 was used to create three single date classifications and a land cover change image depicting the sequence of changes in land cover between 1987–1991–1996. An accessibility analysis examined land cover change and landscape fragmentation relative to elevation and distance from roads. Between 1987 and 1991, results follow ‘expected’ trends, with more accessible areas experiencing greater deforestation and fragmentation. Between 1991 and 1996 this trend reverses. Increased deforestation is found in areas distant from roads, and at higher elevations; a result of government policies promoting expansion of mountain coffee production for export. A ban on logging, and abandonment of marginally productive agricultural fields due to agricultural intensification in other parts of the landscape, has led to increased regrowth in accessible regions of the landscape. Roads and elevation also present different obstacles in terms of their accessibility, with the smallest patches of cyclical clearing and regrowth, relating mostly to the agricultural fallow cycle, found at the highest elevations but located close to roads. This research highlights the need to locate analyses of land cover change within the context of local socio-economic policies and land use processes. This revised version was published online in August 2006 with corrections to the Cover Date.     

Patterns and scale relations among urbanization measures in Stockholm,Sweden

In this study we measure urbanization based on a diverse set of 21 variables ranging from landscape indices to demographic factors such as income and land ownership using data from Stockholm, Sweden. The primary aims were to test how the variables behaved in relation to each other and if these patterns were consistent across scales. The variables were mostly identified from the literature and limited to the kind of data that was readily accessible. We used GIS to sample the variables and then principal component analyses to search for patterns among them, repeating the sampling and analysis at four different scales (250 × 250, 750 × 750, 1,250 × 1,250 and 1,750 × 1,750, all in meters). At the smallest scale most variables seemed to be roughly structured along two axes, one with landscape indices and one mainly with demographic factors but also impervious surface and coniferous forest. The other land-cover types did not align very well with these two axes. When increasing the scale this pattern was not as obvious, instead the variables separated into several smaller bundles of highly correlated variables. Some pairs or bundles of variables were correlated on all scales and thus interchangeable while other associations changed with scale. This is important to keep in mind when one chooses measures of urbanization, especially if the measures are indices based on several variables. Comparing our results with the findings from other cities, we argue that universal gradients will be difficult to find since city shape and size, as well as available information, differ greatly. We also believe that a multivariate gradient is needed if you wish not only to compare cities but also ask questions about how urbanization influences the ecological character in different parts of a city.     

Influence of biophysical factors and differences in Ojibwe reservation versus Euro-American social histories on forest landscape change in northern Wisconsin,USA

Landscape ecology studies have demonstrated that past modifications of the landscape frequently influence its structure, highlighting the utility of integrating historical perspectives from the fields of historical ecology and environmental history. Yet questions remain for historically-informed landscape ecology, especially the relative influence of social factors, compared to biophysical factors, on long-term land-cover change. Moreover, methods are needed to more effectively link history to ecology, specifically to illuminate the underlying political, economic, and cultural forces that influence heterogeneous human drivers of land-cover change. In northern Wisconsin, USA, we assess the magnitude of human historical forces, relative to biophysical factors, on land-cover change of a landscape dominated by eastern white pine (Pinus strobus L.) forest before Euro-American settlement. First, we characterize land-cover transitions of pine-dominant sites over three intervals (1860–1931; 1931–1951; 1951–1987). Transition analysis shows that white pine was replaced by secondary successional forest communities and agricultural land-covers. Second, we assess the relative influence of a socio-historical variable (“on-/off-Indian reservation”), soil texture (clay and sand), and elevation on land-cover transition. On the Lake Superior clay plain, models that combine socio-historical and biophysical variables best explain long-term land-cover change. The socio-historical variable dominates: the magnitude and rate of land-cover change differs among regions exposed to contrasting human histories. Third, we developed an integrative environmental history-landscape ecology approach, thereby facilitating linkage of observed land-cover transitions to broader political, economic, and cultural forces. These results are relevant to other landscape investigations that integrate history and ecology.     

Forest recovery in a tropical landscape: what is the relative importance of biophysical,socioeconomic, and landscape variables?

Socioeconomic changes in many areas in the tropics have led to increasing urbanization, abandonment of agriculture, and forest re-growth. Although these patterns are well documented, few studies have examined the drivers leading to landscape-level forest recovery and the resulting spatial structure of secondary forests. Land cover transitions from agricultural lands to secondary forest in the island of Puerto Rico have been ongoing since the 1940s. This study is a glimpse into this landscape level trend from 1991 to 2000. First, we relied on Landsat images to characterize changes in the landscape structure for forest, urban, and agricultural land classes. We found that although forest cover has increased in this period, forest has become increasingly fragmented while the area of urban cover has spread faster and become more clustered. Second, we used logistic regression to assess the relationship between the transition to forest and 21 biophysical, socioeconomic, and landscape variables. We found that the percentage of forest cover within a 100 m radius of a point, distance to primary roads and nature reserves, slope, and aspect are the most important predictors of forest recovery. The resulting model predicts the spatial pattern of forest recovery with accuracy (AUC-ROC = 0.798). Together, our results suggest that forest recovery in Puerto Rico has slowed down and that increasing pressure from urbanization may be critical in determining future landscape level forest recovery. These results are relevant to other areas in the tropics that are undergoing rapid economic development.     

Evaluating the Relative Roles of Ecological Regions and Land-cover Composition for Guiding Establishment of Nutrient Criteria

The continuing degradation of United States surface waters by excessive nutrient loads has motivated the establishment of nutrient criteria for streams, lakes, and estuaries as a means to protect aquatic resources. Nutrient criteria have been established based on ecoregional differences, recognizing that geographic variation in climate, topography, geology, and land use require use of different criteria values for different regions of the continental United States. Several studies have demonstrated that land-cover composition also strongly influences nutrient concentrations and yields. We examined the relative importance of ecoregions and watershed land-cover composition in explaining variability in nitrogen (N) and phosphorus (P) concentrations by re-analyzing the National Eutrophication Survey (NES) data reported by Omernik (1977). The variance of N concentrations among land-cover composition classes within ecoregions was six times larger than the variance among ecoregions. For P concentrations, land-cover composition within ecoregions accounted for three times more variance than ecoregions themselves. Variance across ecoregions was only weakly significant after accounting for variance in land-cover composition within ecoregions. The results suggest that the relationship between land-cover composition and nutrient concentrations in aquatic systems should also be used to help guide establishment of nutrient criteria.     

Regional land-cover conversion in the U.S. upper Midwest: magnitude of change and limited recovery (1850–1935–1993)

Land-use legacies can persist for hundreds to thousands of years, influencing plant species composition, nutrient cycling, water flows, and climate. To understand how land use has affected regional land-cover composition in Wisconsin (USA), we assessed the magnitude and direction of change in land cover between: (1) c.1850, at the onset of Euro-American settlement; (2) c.1935, the period of maximum clearing for agriculture following widespread forest logging; and (3) 1993, which, especially in northern Wisconsin, follows farm abandonment and forest recovery. We derived land-cover maps using U.S. Public Land Survey records (c.1850), the Wisconsin Land Economic Inventory (c.1935), and Landsat TM satellite data (1993). We stratified Wisconsin (145,000 km²) into two ecological provinces and used spatial error models, multinomial logistic regression, and non-metric multi-dimensional scaling ordination to examine change. Between 1850 and 1935, forest cover in the North declined from 84% to 56%, cropland increased to 24%, and mixed/coniferous forests and savannas were replaced by deciduous forests. In the South, formerly dominant savannas (69%) and prairies (6%) were mostly converted to cropland (51%) and pasture (11%). Remnant deciduous savannas and coniferous forests and savannas were replaced by deciduous forests. Remarkably little recovery to pre-settlement land-cover classes occurred from 1935 to 1993. Less cropland was abandoned than expected, and there was little net gain in coniferous/mixed forest. Based on these general land-cover classes, current cover is significantly different from that in 1850, but not from that in 1935, and thus continues to reflect historical logging and agricultural patterns. These results provide a historical framework for measuring associated changes in ecosystem function and can be used to guide restoration where desirable and feasible.     

Land-cover and structural changes in a western Norwegian cultural landscape since 1865, based on an old cadastral map and a field survey

Many studies of land-cover and structural changes in cultural landscapes have used historical maps as a source for information about past land-cover. All transformations of historical maps onto modern coordinate systems are however burdened with difficulties when it comes to accuracy. We show that a detailed land survey of the present landscape may enable transformation of an old cadastral map directly onto the present terrain with very high accuracy. The detailed resulting map enabled us to locate remnants of semi-natural grasslands and man-made structures with continuity from 1865 and to test hypotheses about relationships between landscape changes and landscape characteristics. The main land-cover change 1865–2002 was decrease of arable fields, and addition of three new land-cover classes: horticultural, orchard and abandoned areas. Of the 330 man-made structures present in 1865, only 58 remained in 2002, while 63 new structures had been built after 1865. We found that semi-natural grasslands with continuity since 1865 were situated on ground with significantly lower production capacity than mean 1865 production capacity. The man-made structures with continuity since 1865 were also associated with areas with significantly lower production capacity than the 1865 mean, situated in significantly steeper terrain but not further from the hamlet. Our study illustrates the potential of digitised and accurately transformed historical cadastral maps combined with detailed field surveys for analysis of land-cover and structural changes in the cultural landscape.     

Impact of error on landscape pattern analyses performed on land-cover change maps

Researchers have emphasized the value of linking observed patterns of land-cover change to the processes driving changes in land-use to explain the dynamics of a land change system. The association of pattern and process requires an accurate quantification of the spatial characteristics of land-cover change. The objective of this research is to assess the impact of error on the accuracy of landscape pattern analyses performed on maps of change. Simulation was used to develop of a series of error-free and error-perturbed change maps, which varied with respect to the pattern of change occurring between the time-1 and time-2 land-cover maps and the patterns of error associated with the time-1 and time-2 land-cover maps. A variety of change and error patterns were examined. The error-free and error-perturbed change maps were compared by calculating landscape pattern metrics, which revealed the degree to which error altered the pattern of change. The introduction of error notably changed the structure of the persistent and transitioning classes, with metrics indicating a more fragmented and variable landscape under error. Agreement between the error-free and error-perturbed maps improved when a greater amount of change occurred within the time-series, change was concentrated at the boundaries of land-cover classes and when time-2 errors were increasingly correlated to their time-1 counterparts. These results have several implications for change pattern analyses given the fundamental nature of land-cover change.     

Seasonal dynamics of soil nitrogen availability and phosphorus fractions under urban forest remnants of different vegetation communities in Southern China

Urban forest remnants are a useful tool to study forest response to global change with urbanization. Soil nutrient status in urban forests has not been well understood, especially under the pressure of rapid urbanization in developing countries. In this study, ion-exchange resin bags and a modified Hedley P fractionation procedure were used to measure seasonal dynamics of soil N forms (ammonium and nitrate) and P fractions (available, labile, slow, occlude and weathered mineral P) under urban forest remnants across a successional sequence and non-forest land in the city of Nanchang, Southern China. Results showed that soil N availability varied with season and vegetation community (P < 0.05). Soil P fractions showed minimal seasonal variation except available P, while their averages generally increased with forest development from non-forest land to coniferous forest to conifer-broadleaf mixed forest to evergreen broad-leaved forest. The ratios of fresh soil N forms to P fractions generally decreased with forest development, while N forms absorbed by resins to P fractions generally increased from non-forest land to coniferous forest, then decreased from conifer-broadleaf mixed forest to evergreen broad-leaved forest. It is suggested that urban older forest remnants could easily move to N saturation status and soil P enrichment, causing urban water pollution due to the accumulative effect of elevated atmospheric N deposition and exogenous P input with urbanization.     

Assessing the effects of subtropical forest fragmentation on leaf nitrogen distribution using remote sensing data

Subtropical forest loss resulting from conversion of forest to other land-cover types such as grassland, secondary forest, subsistence crop farms and small forest patches affects leaf nitrogen (N) stocks in the landscape. This study explores the utility of new remote sensing tools to model the spatial distribution of leaf N concentration in a forested landscape undergoing deforestation in KwaZulu-Natal, South Africa. Leaf N was mapped using models developed from RapidEye imagery; a relatively new space-borne multispectral sensor. RapidEye consists of five spectral bands in the visible to near infra-red (NIR) and has a spatial resolution of 5 m. MERIS terrestrial chlorophyll index derived from the RapidEye explained 50 % of the variance in leaf N across different land-cover types with a model standard error of prediction of 29 % (i.e. of the observed mean leaf N) when assessed on an independent test data. The results showed that indigenous forest fragmentation leads to significant losses in leaf N as most of the land-cover types (e.g. grasslands and subsistence farmlands) resulting from forest degradation showed lower leaf N when compared to the original indigenous forest. Further analysis of the spatial variation of leaf N revealed an autocorrelation distance of about 50 m for leaf N in the fragmented landscape, a scale corresponding to the average dimension of subsistence fields (2,781 m²) in the region. The availability of new multispectral sensors such as RapidEye thus, moves remote sensing closer to widespread monitoring of the effect of tropical forest degradation on leaf N distribution.     

Forest fragmentation in Loudoun County,Virginia, USA evaluated with multitemporal Landsat imagery

In order to study forest fragmentation in the Virginia, USA Piedmont, a series of Landsat images from 1973, 1987, and 1999 covering a rapidly developing area (Loudoun County) was used to classify forest from non-forest. The classified images were analyzed using a geographic information system (GIS) to determine the spatial and temporal patterns of fragmentation, and to relate these patterns to infrared radiance provided by Landsat Enhanced Thematic Mapper Plus (ETM+) band 6. The analysis was concentrated on eleven major watersheds of Loudoun County. The relationship between urbanized area per watershed and mean fragment size showed a strong negative decay form (r ²=0.757, p<0.0001). Analysis of four landscape metrics showed increasing fragmentation of forest cover, particularly from 1987 to 1999, as well as an increase in forest edge and shape complexity. Of the landscape metrics used, the perimeter-to-area (P/A) ratio showed the strongest relationship with mean radiance of forest patches. In addition, there was a negative, linear relationship between distance from major roads and band 6 radiance of forested pixels. Overall, the study shows that landscape metrics can convey meaningful information on biophysical changes associated with forest fragmentation at broad scales. These changes suggest that ambient temperature increases associated with urban sprawl may have important, long-term implications for ecophysiological processes.     

The influence of patch-delineation mismatches on multi-temporal landscape pattern analysis

Investigations of land-cover change often employ metrics designed to quantify changes in landscape structure through time, using analyses of land cover maps derived from the classification of remote sensing images from two or more time periods. Unfortunately, the validity of these landscape pattern analyses (LPA) can be compromised by the presence of spurious change, i.e., differences between map products caused by classification error rather than real changes on the ground. To reduce this problem, multi-temporal time series of land-cover maps can be constructed by updating (projecting forward in time) and backdating (projecting backward in time) an existing reference map, wherein regions of change are delineated through bi-temporal change analysis and overlaid onto the reference map. However, this procedure itself creates challenges, because sliver patches can occur in cases where the boundaries of the change regions do not exactly match the land-cover patches in the reference map. In this paper, we describe how sliver patches can inadvertently be created through the backdating and updating of land-cover maps, and document their impact on the magnitude and trajectory of four popular landscape metrics: number of patches (NP), edge density (ED), mean patch size (MPS), and mean shape index (MSI). In our findings, sliver patches led to significant distortions in both the value and temporal behaviour of metrics. In backdated maps, these distortions caused metric trajectories to appear more conservative, suggesting lower rates of change for ED and inverse trajectories for NP, MPS and MSI. In updated maps, slivers caused metric trajectories to appear more extreme and exaggerated, suggesting higher rates of change for all four metrics. Our research underscores the need to eliminate sliver patches from any study dealing with multi-temporal LPA.