Spring phenological variability promoted by topography and vegetation assembly processes in a temperate forest landscape
Introduction
Vegetation phenology significantly influences diverse ecosystem processes across multiple ecological scales, from local organismal interactions (Halupka and Halupka, 2017; Heberling et al., 2019; Royo and Stanovick, 2019; Singer and Parmesan, 2010; Visser et al., 2004) to global biogeochemical cycles (Morisette et al., 2009; Richardson et al., 2013). Vegetation phenology is also among the most sensitive indicators of global environmental change, supporting improved understanding of climate change impacts on terrestrial ecosystems (Fitter and Fitter, 2002; Morisette et al., 2009; Root et al., 2003). As a result, vegetation phenology is increasingly studied, particularly in light of advancing remote sensing technologies that repetitively resolve phenological patterns, aka, Land Surface Phenology (LSP), across multiple spatial and temporal scales (Cleland et al., 2007; Morisette et al., 2009; Nagai et al., 2016). However, despite increased research capacity, we continue to lack adequate mechanistic understandings of vegetation phenology and its variability and drivers (Chmura et al., 2019), limiting the development of predictive models in the face of continued environmental change (Basler, 2016; Richardson et al., 2013).
The timing of spring leaf development among temperate forest ecosystems has been shown to be particularly sensitive to temperature (Linkosalo et al., 2006; Polgar et al., 2014; Polgar and Primack, 2011; Vitasse et al., 2009). At large spatial scales, phenological patterns remain largely consistent with climatic variation across broad latitudinal and elevational gradients (Fitzjarrald et al., 2001; Hopkins, 1918). However, spatial patterns often reveal significant fine-scale variation as well (Elmore et al., 2012; Fisher et al., 2006; Liang et al., 2011; Melaas et al., 2013). Thus, phenological patterns likely also incorporate a variety of possibly interacting local processes (Chmura et al., 2019). For example, local topography influences microclimatic variation, such as processes contributing to cold air drainage into small valleys, which has been shown to delay phenological timing of forest vegetation in comparison to surrounding uplands (Fisher et al., 2006; Schuster et al., 2014). Individual tree species, which also frequently sort according to local topography (Desta et al., 2004; Hix, 1988; Hix and Pearcy, 1997; Martin et al., 2011), can display large interspecific variation in phenological timing as well (Delpierre et al., 2017; Denéchère et al., 2019; Lechowicz, 1984; Murray et al., 1989; Richardson et al., 2006). There also exists some variability in leaf phenology that can be explained by differences in tree height (Osada and Hiura, 2019; Seiwa, 1999), likely serving as a proxy for both individual tree response to site conditions (Bassow and Bazzaz, 1998) as well as differences between juvenile and adult age classes (Augspurger and Bartlett, 2003; de Souza and da Costa, 2020). Because spatial variability likely incorporates a greater range of direct and indirect processes at more local scales, as opposed to more direct climatic forcing across large regions, models at these scales should account for the combined effects of local topography on the composition of forest vegetation and how these factors interact to influence phenological behavior.
Significant efforts have identified the dual importance of topography and forest stand characteristics in explaining variation in LSP behavior (Isaacson et al., 2012; Kraus et al., 2016; Liang et al., 2011; Misra et al., 2018; Reaves et al., 2018; Xie et al., 2015). However, few studies have interpreted spatial variation in light of interactions between topography and forest vegetation composition (Reaves et al., 2018; Xie et al., 2015). First, many LSP studies continue to incorporate generalized models of broad forest types in lieu of detailed species-level information resolving the ecophysiological variation of forest stands (Isaacson et al., 2012; Xie et al., 2015). Second, modeling efforts should also ideally employ statistical procedures that account for the often causative and correlative factors among spatial drivers (Bassow and Bazzaz, 1998; Schemske and Horvitz, 1988). For example, oak (Quercus spp.) dominated forests among dissected landscapes of the Central Hardwood Region of the United States often display strong topographically-mediated vegetation gradients that underlie variation in the fundamental ecological strategies of component tree species (Fralish, 2003). Here, drought-tolerant oaks dominate drier site conditions on southwest-facing hillslopes and ridgetops and transition to mixed mesophytic species assemblages, adapted to withstand increased competition, on opposing mesic northeast-facing hillslopes and bottomlands (Adams et al., 2019; Desta et al., 2004; Martin et al., 2011). Understanding how these environmental and organismal processes (and interactions) contribute to phenological patterns could help improve forest management practices in supporting plant diversity in light-structured environments (Heberling et al., 2019; Royo and Stanovick, 2019), especially in predicting the potential consequences of anticipated environmental change (Iverson et al., 2019b; McEwan et al., 2011).
We designed a study to determine how variation in the timing of spring greenup is explained by topography and vegetation assembly processes across a topographically complex and diverse forest landscape in southeastern Ohio, USA. We used LSP methods, combining multi-temporal Landsat observations from the years 1984 to 2020 (including an 8-day temporal resolution for a given sensor among overlapping scenes), to retrieve long-term average dates of spring greenup across a series of field inventories. Phenological curves fitted to multidecadal Landsat observations were used to resolve long-term climatological average spring greenup dates, insensitive to annual climate anomalies (Fisher et al., 2006; Melaas et al., 2013) and at spatial resolutions (30 m) complementary to local management activities (Adams et al., 2019; Iverson et al., 2017). Field inventories elucidated variation in tree species richness and community mean wood density (WD), the ratio of oven-dry mass to total green volume. We used WD to collectively represent variation in assemblage and functional composition of individual tree stands (Chave et al., 2009; Stahl et al., 2013). Finally, data from digital elevation models and LiDAR were used to quantify key topographic variables (elevation, slope percent, and aspect) and mean vertical canopy height.
Our layered approach used regression modeling and variance partitioning procedures to determine the extent to which site-to-site variation in spring greenup is explained by topography, vegetation, and topographically-structured vegetation assembly processes. Next, we incorporated path analysis to precisely determine how topographic features influence phenological variation by mediating changes in forest stand vegetation. Finally, we isolated the relative importance of individual species in explaining forest stand phenological variation, based on relative dominance profiles and contributions towards community-level functional composition. Together, we illustrate how spring leaf phenology is shaped by direct and often complex interacting processes across tightly integrated forest ecosystems.
Section snippets
Study area
We targeted neighboring forest sites (Vinton Furnace State Experimental Forest and Zaleski State Forest) within a “sidelap” region (overlapping zones between Landsat scenes P19/R33 and P18/R33, which improves the temporal frequency from 16 to 8 days for a single instrument and earlier depending on whether more than one instrument is operating at a time and at 30 m resolution) in the Western Hocking Plateau ecological subsection of southeastern Ohio for this study (Bailey et al., 1994) (Fig. 1).
Results
LSP revealed high spatial and temporal variability in long-term climatological average phenology dates across the study area (Fig. 1b) and field data. The timing of spring greenup occurred over an ∼18-day window, ranging from day 119 (April 30th) to day 137 (May 17th). The simple regression model, including topography, tree species richness, community-mean WD, and canopy height terms, explained nearly 52% of the variation in greenup across the field data. Although slope was the only
Landscape controls over spring phenology
This study combined LSP techniques and field inventories to examine the relationships of topography and vegetation characteristics on phenological patterns at landscape scales. Our analyses reiterated the dual importance of local topography and vegetation characteristics on the timing of spring greenup, following prior studies over a variety of forest types (Klosterman et al., 2018; Liang et al., 2011; Misra et al., 2018; Xie et al., 2015). However, our analyses extended the results of previous
Conclusion
LSP revealed spatial and temporal variation in the phenological timing of forests across the dissected study area. This variation was attributed directly to variation in topography and vegetation features, including vertical canopy height, diversity, and species and functional composition. Elevation, related to relief above valley bottomlands, explained particularly large portions of this variance. However, this relationship was nonlinear, and implied varying rates of canopy development across
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This research was funded by the State Wildlife Grants Program, administered jointly by the United States Fish and Wildlife Service and the Ohio Division of Wildlife, through the Ohio Biodiversity Conservation Partnership and by United States Department of Agriculture Forest Service Northern Research Station agreement 15-CS-11242302-122 (to SM). The authors wish to thank several individuals for their assistance in data collection: Kaley Donovan, James Hanks, Garrett Evans, Sara Zaleski, Alex
References (89)
- et al.
Mapping floristic gradients of forest composition using an ordination-regression approach with Landsat OLI and terrain data in the Central Hardwoods region
For. Ecol. Manag.
(2019) Evaluating phenological models for the prediction of leaf-out dates in six temperate tree species across central Europe
Agric. For. Meteorol.
(2016)- et al.
Assessing spring phenology of a temperate woodland: a multiscale comparison of ground, unmanned aerial vehicle and Landsat satellite observations
Remote Sens. Environ.
(2019) - et al.
Continental-scale land surface phenology from harmonized Landsat 8 and Sentinel-2 imagery
Remote Sens. Environ.
(2020) - et al.
Shifting plant phenology in response to global change
Trends Ecol. Evol.
(2007) - et al.
Tree phenological ranks repeat from year to year and correlate with growth in temperate deciduous forests
Agric. For. Meteorol.
(2017) - et al.
Cross-scalar satellite phenology from ground, Landsat, and MODIS data
Remote Sens. Environ.
(2007) - et al.
Green leaf phenology at Landsat resolution: scaling from the field to the satellite
Remote Sens. Environ.
(2006) - et al.
Google earth engine: planetary-scale geospatial analysis for everyone
Remote Sens. Environ.
(2017) - et al.
Oaks were the historical foundation genus of the east-central United States
Quat. Sci. Rev.
(2016)
An analysis of Landsat 7 and Landsat 8 underflight data and the implications for time series investigations
Remote Sens. Environ.
Overview of the radiometric and biophysical performance of the MODIS vegetation indices
Remote Sens. Environ.
Fine-scale perspectives on landscape phenology from unmanned aerial vehicle (UAV) photography
Agric. For. Meteorol.
Validating satellite phenology through intensive ground observation and landscape scaling in a mixed seasonal forest
Remote Sens. Environ.
Phenology and wood density of plants growing in the semi-arid region of northeastern Brazil
J. Arid Environ.
Coupling of vegetation layers and environmental influences in a mature, second-growth Central Hardwood forest landscape
For. Ecol. Manag.
Detecting interannual variation in deciduous broadleaf forest phenology using Landsat TM/ETM+ data
Remote Sens. Environ.
LiDAR derived topography and forest stand characteristics largely explain the spatial variability observed in MODIS land surface phenology
Remote Sens. Environ.
Structural and compositional shifts in forests undergoing mesophication in the Wayne National Forest, southeastern Ohio
For. Ecol. Manag.
Drivers of spatial variability in greendown within an oak-hickory forest landscape
Remote Sens. Environ.
Climate change, phenology, and phenological control of vegetation feedbacks to the climate system
Agric. For. Meteorol.
Characterization of Landsat-7 to Landsat-8 reflective wavelength and normalized difference vegetation index continuity
Remote Sens. Environ.
Deer browsing overwhelms extended leaf phenology benefits: a test case with Rubus allegheniensis and a recalcitrant hay-scented fern layer
For. Ecol. Manag.
Changes in leaf phenology are dependent on tree height in Acer mono, a deciduous broad-leaved tree
Ann. Bot.
Preliminary analysis of the performance of the Landsat 8/OLI land surface reflectance product
Remote Sens. Environ.
Global climate change leads to mistimed avian reproduction
Adv. Ecol. Res.
Detecting change-point, trend, and seasonality in satellite time series data to track abrupt changes and nonlinear dynamics: a Bayesian ensemble algorithm
Remote Sens. Environ.
Object-based cloud and cloud shadow detection in Landsat imagery
Remote Sens. Environ.
Diverse temperate forest bird assemblages demonstrate closer correspondence to plant species composition than vegetation structure
Ecography
Differences in leaf phenology between juvenile and adult trees in a temperate deciduous forest
Tree Physiol.
Ecoregions and Subregions of the United States (map)
Contrasting distribution and seasonal dynamics of carbohydrate reserves in stem wood of adult ring-porous sessile oak and diffuse-porous beech trees
Tree Physiol.
How environmental conditions affect canopy leaf-level photosynthesis in four deciduous tree species
Ecology
Aspect transformation in site productivity research
J. For.
Partialling out the spatial component of ecological variation
Ecology
A meta-analysis of the fire-oak hypothesis: does prescribed burning promote oak reproduction in Eastern North America?
For. Sci.
The effect of environmental factors on the abundance of oak regeneration beneath mature oak stands
For. Sci.
Towards a worldwide wood economics spectrum
Ecol. Lett.
The mechanisms of phenology: the patterns and processes of phenological shifts
Ecol. Monogr.
The shifting phenological landscape: within- and between-species variation in leaf emergence in a mixed-deciduous woodland
Ecol. Evol.
Differences in leaf phenology between juvenile and adult individuals of two tree species in a seasonally dry tropical woodland
Aust. Ecol.
The within-population variability of leaf spring and autumn phenology is influenced by temperature in temperate deciduous trees
Int. J. Biometeorol.
Aspect induced differences in vegetation, soil, and microclimatic characteristics of an Appalachian watershed
Castanea
Loss of foundation species: consequences for the structure and dynamics of forested ecosystems
Front. Ecol. Environ.
Cited by (6)
Evaluating fine-scale phenology from PlanetScope satellites with ground observations across temperate forests in eastern North America
2022, Remote Sensing of EnvironmentCitation Excerpt :The fine-scale LSP characterizations over large vegetated landscapes using PlanetScope data (Figs. 11 and 12) generate new opportunities to help assess the phenology variability across various levels from individual trees/species to forest ecosystems. As both biotic and environmental factors are important when explaining the LSP variability across various levels (Figs. 11 g-h and 12 g-h; Adams et al., 2021; Fisher et al., 2006), our results further suggest that the PlanetScope-derived fine-scale LSP data, together with other data, such as maps of abiotic (e.g., topography, soil type) and biotic (e.g., species, plant functional traits) factors, would be useful to advance the understanding of mechanisms that drive leaf phenology variability at different levels. Moreover, the fine-scale LSP characterizations over multiple years using PlanetScope data (e.g., Moon et al., 2022) also has the potential to help us assess and understand the phenological responses to climate variability and change.
Topography modulates climate sensitivity of multidecadal trends of holm oak decline
2023, Global Change BiologyFunctional vegetation community responses to soil and topographic factors in the Loess Plateau of China
2023, Land Degradation and DevelopmentImpacts of terrain on land surface phenology derived from Harmonized Landsat 8 and Sentinel-2 in the Tianshan Mountains, China
2023, GIScience and Remote Sensing