Impacts of Multiple Extreme Disturbance Events On Landscape Cover: Wildfire and Flooding in Canyon Riparian Ecosystems of the Jemez Mountains, New Mexico, USA

Item

Title
Impacts of Multiple Extreme Disturbance Events On Landscape Cover: Wildfire and Flooding in Canyon Riparian Ecosystems of the Jemez Mountains, New Mexico, USA
Creator
Alfieri, Samuel James
Date
2020
extracted text
IMPACTS OF MULTIPLE EXTREME DISTURBANCE EVENTS
ON LANDSCAPE COVER:
WILDFIRE AND FLOODING IN CANYON RIPARIAN ECOSYSTEMS
OF THE JEMEZ MOUNTAINS, NEW MEXICO, USA

by
Samuel James Alfieri

A Thesis
Submitted in partial fulfillment
of the requirements of the degree
Master of Environmental Studies
The Evergreen State College
June 2020

© 2020 by Samuel Alfieri. All rights reserved.

This Thesis for the Master of Environmental Studies Degree
by Samuel Alfieri
has been approved for
The Evergreen State College
by

_______________________________
John Withey, PhD
Member of the Faculty

__________________________
Date

ABSTRACT
Impact of Multiple Extreme Disturbance Events on Landscape Cover:
Wildfire and Flooding in Canyon Riparian Ecosystems
of the Jemez Mountains, New Mexico, USA

Changes in regional climate and human land use in the Southwest United States have contributed
to widespread tree mortality and provided the conditions for large-extent high-intensity fires to
proliferate across the landscape. Increasingly extreme disturbance regimes initiated by
widespread fires and high-volume flooding may cause irreversible changes to ecosystems, and
lead to unsuitable conditions for long-standing forests to re-establish. On the land occupied by
the Bandelier National Monument on the Pajarito Plateau of the Jemez Mountains in New
Mexico, United States, three large-scale fires have occurred since 1996, resulting in intense
flooding and sediment transport, and eliminating propagules and habitat connectivity of longstanding riparian forests. This study interpreted aerial photographs of the area of interest from
1991-1992, 2008, 2014, and 2018 to quantify and describe changes to the landscape and cover
types prior to and following known disturbance events. In each canyon, four distinct segments
were classified according to the types and degrees of disturbance and effects observed. Overall,
dramatic decreases in forest cover following fires were observed and quantified. Forest cover in
Frijoles Canyon decreased from over 70% in the early 1990s to under 20% in 2018, and in
Capulin Canyon, from 57% to under 5% over the same time period. Understanding patterns of
disturbance and vegetative re-establishment helps reveal opportunities and challenges for
management directions to support the intrinsic resilience of the system, or intervene where
necessary. If management goals seek to maintain critical ecosystem services such as water
filtration, flood control, and carbon sequestration, in addition to preserving cultural and natural
resources, comprehensive assessment of the changing landscape will be critical for determining
current impacts and future trajectories following recent disturbance events.

Table of Contents
ABSTRACT ................................................................................................................................... iv
List of Figures ................................................................................................................................. v
List of Tables .................................................................................................................................. v
Chapter One: Literature Review ..................................................................................................... 1
Riparian ecosystems.................................................................................................................... 1
Disturbance in riparian systems .................................................................................................. 2
Fire and flood regimes and climate change ................................................................................ 5
Adaptations of riparian vegetation .............................................................................................. 6
Implications for viability of forest recovery and management ................................................... 9
Stream segment classification ................................................................................................... 11
History of land use and fires in Bandelier National Monument, New Mexico ........................ 15
Chapter Two: Research Manuscript.............................................................................................. 22
Introduction ............................................................................................................................... 22
History of fire on Pajarito Plateau ........................................................................................ 23
Direct effects of recent fires on Capulin and Frijoles Canyons ............................................ 25
Methods..................................................................................................................................... 27
Study area.............................................................................................................................. 27
Sample design ....................................................................................................................... 28
Satellite interpretation ........................................................................................................... 30
Data analysis ......................................................................................................................... 33
Results ....................................................................................................................................... 34
Frijoles Canyon ..................................................................................................................... 34
Capulin Canyon .................................................................................................................... 39
Discussion ................................................................................................................................. 43
Conclusion .................................................................................................................................... 46
Appendix ....................................................................................................................................... 59

iv

List of Figures
Figure 1. Study area map showing Frijoles and Capulin Canyons, and three large recent fire
extents. .......................................................................................................................................... 28
Figure 2. Buffered centers of grid squares in the Frijoles Canyon bottom alongside the Bandelier
National Monument Visitor Center. ............................................................................................. 30
Figure 3. NAIP Imagery from 2011 and 2018 showing examples of sample point classified as
forest (F), other vegetation (O), wetted channel (W), bare ground (B), and shadow (S). ............ 31
Figure 4. Proportions of vegetative cover observed in Frijoles Canyon overall. ......................... 36
Figure 5. Proportions of vegetative cover type observed by 1 km reaches in Frijoles Canyon. ... 37
Figure 6. Proportions of cover types observed by disturbance segment and year in Frijoles
Canyon. ......................................................................................................................................... 39
Figure 7. Proportions of vegetative cover observed in Capulin Canyon overall by year. ............ 40
Figure 8. Proportions of vegetative cover observed by 1 km reaches in Capulin Canyon. .......... 41
Figure 9. Proportions of vegetative cover observed by disturbance segment and year in Capulin
Canyon. ......................................................................................................................................... 43
Figure A1. . Proportions of vegetative cover observed in Frijoles Canyon overall by year ......... 60
Figure A2. Proportions of vegetative cover observed in Capulin Canyon overall by year. ......... 62

List of Tables
Table 1. Criteria used to designate different cover types for NAIP imagery. .............................. 32
Table 2. Proportions of vegetative cover type observed in Frijoles Canyon overall. ................... 36
Table 3. Classification and descriptions of segments by disturbance in Frijoles Canyon. ........... 38
Table 4. Proportions of vegetative cover type observed in Capulin Canyon overall by year....... 40
Table 5. Classification and descriptions of segments by disturbance in Capulin Canyon. .......... 42
Table A1. Frijoles Canyon raw cover proportions without channel designated. .......................... 59
Table A2. Counts of sample points by cover type, disturbance segment, and year for Frijoles
Canyon. Data represented graphically as proportions in Figure 5. ............................................... 60
Table A3. Capulin Canyon raw proportions without channel designated. ................................... 61
Table A4. Counts of sample points by cover type, disturbance segment, and year for Capulin
Canyon. Data represented graphically as proportions in Figure 9. ............................................... 62

v

Acknowledgements
The land that is the focus of this study is the ancestral home of the Cohiti, San Felipe, San
Ildefonso, Santa Clara, Santo Domingo, and Zuni Pueblo people. Several Indigenous
communities and nations, including the San Juan Pueblo, Zia Pueblo, Hopi, and Navajo also
assert traditional relationships with the area. My deepest thanks to the Indigenous people and
their ancestors who have been the traditional stewards and inhabitants of the area that is now
known as New Mexico in the American Southwest.
My gratitude goes out to many who have helped and supported me throughout the MES
program and this thesis.
To my reader John Withey, who has guided me through statistics and R, and taught me
the fundamentals of landscape ecology. To Mike Ruth, for his eternal passion for GIS, as well as
the rest of the faculty for their commitment to the program and teaching. To my peers for their
constant engagement and support, especially my peer reviewers, Marisa Pushee and Christine
Davis, who read countless drafts and sections of this thesis, and whose suggestions were
invaluable in crafting this product. To all the human, plant, and animal friends that I have made
in Washington over the last two years, and the home that I have found in Olympia.
Thank you to Tim Assal for his help on the early stages of interpretation and his
generation of the bottomland boundary, and to those at the National Park Service whose input
and perspectives informed my thinking and this work.
And my deepest gratitude to Pat Shafroth at USGS in Fort Collins, Colorado, who took a
chance following up on a random email from a grad student in Washington, for the unending
support and encouragement throughout this endeavor, and the opportunity to work on this and
other projects.

vi

Chapter One: Literature Review
Riparian ecosystems are characterized by regular disturbances, unique ecological
processes, and heterogeneous composition, providing habitat to diverse aquatic and terrestrial
plants and animals (Naiman, Decamps, et al., 2005). Increasingly intense and frequent
disturbance events including wildfires and flooding are occurring in many riparian systems, and
long-standing patterns of disturbance and recovery are derailed as some systems experience
unprecedented stressors and permanently altered trajectories (Harmon et al., 1986). Riparian
plants possess a range of specialized adaptations that allow them to be resistant and resilient to
regular disturbances; however extraordinary disturbance events may stretch systems beyond their
ability to resume functionality congruent to their pre-disturbance status (Allen, 2014).
Categorizing stream segments based on disturbance can serve as an important first step to
understanding local dynamics, and later serve as an aid to defining management goals and
strategies for distinct areas (Quinn et al., 2001)

Riparian ecosystems
Riparian zones extend from the edges of water bodies into upland communities, and are
characterized by the regular influence of fresh water, strong energy regimes, and biological and
physical diversity (Naiman, Decamps, et al., 2005). They are home to unique environmental
processes and communities of animal and plant assemblages, and connect ecological systems and
physical processes from upland areas to the oceans (Naiman & Decamps, 1997). Due to their
positioning as hydraulic conduits of biotic and abiotic elements and energy from their
headwaters to mouths, rivers serve as “critical transition zones” connecting ecosystems together,

1

with disturbances resulting in consequences at every level of interaction (Richardson et al.,
2007).
Riparian ecosystems are often unique in their surroundings, integrating aquatic and
terrestrial components of the landscape, and providing substantial habitat heterogeneity for
native plants and animals (Naiman, Bechtold, et al., 2005). Riparian ecosystems are connected in
three dimensions: laterally, longitudinally, and vertically, with hydrology influenced by both
surface and groundwater, as well as local geology and topography (Naiman, Decamps, et al.,
2005). Considering the fourth dimension of time adds complexity to understanding riparian
ecological processes, as considerable changes in composition and structure from one time period
to another can fundamentally alter the functionality and trajectory of the system (Reiners, 2005).

Disturbance in riparian systems
Understanding disturbance regimes in riparian ecosystems is of critical importance for
maintaining ecosystem services and cultural resources in flood- and fire-prone zones (Naiman et
al., 1998). While a wide range disturbances may occur naturally in many riparian systems,
increased frequency of high-intensity and large-scale disruptions in areas not typically prone to
such disturbances are presently producing historically unobserved and lasting consequences,
where ecosystems may fail to recover to locally recognizable states (Odum et al., 1979; Rapport
& Whitford, 1999; Sparks et al., 1990). Conceptual models of disturbance and recovery
trajectories distinguish between different scales, intensity, and frequency of disturbance, with
differing degrees of influence that shape the landscape and vegetation dynamics on seasonal,
decadal, or larger scales (Brinson, 1990; Richardson et al., 2007). Such disturbances can include
landslide, flood, wind, fire, drought, disease, litter accumulation, herbivory, drought, or other

2

physical influences, and interact to exert short-term and lasting impacts on riparian landscapes
and ecosystems (Naiman et al., 1998).
While rivers are subject to frequent and regular disturbances at multiple scales, especially
extreme and successive disturbances can cause ecosystems and cover types to convert to
historically unprecedented states (Didham et al., 2005). In most parts of the world, riparian
vegetation is dominated by woody species. However, in places where climate is cold,
hydrogeomorphology is waterlogged, or disturbances are too frequent for trees to establish, grass
and shrublands dominate (Richardson et al., 2007). Shifts from forested ecosystems to grass- and
shrub-dominated systems can have long-term consequences altering hydrologic regimes,
biodiversity, and carbon sequestration, among other ecosystem and cultural services, as the
landscape ceases to function as it did before (Allen, 2014; Hicke et al., 2012; Richardson et al.,
2007).
Flood disturbance and fluvial have been shown to exert considerable influence on
vegetation, and act as drivers of habitat patch dynamics in riparian systems, producing a mosaic
of shifting habitat that is maintained in quasi-equilibrium by simultaneous cut and fill processes
throughout the river channel (Mouw et al., 2009). However, in riparian systems that are also
influenced by fire, fire has been shown to exert the strongest total effect on variability of
floodplain habitat patch composition compared to stream power and geomorphic position
(Kleindl et al., 2015). Regular seasonal or intermittent flooding and fire patterns may not be
considered disturbances, since their occurrence facilitates habitat heterogeneity and species
diversity; however when the extent and intensity of disturbances increase drastically, they can
result in irreversible and fundamental shifts to community compositions and ecosystem
functionality (Rapport & Whitford, 1999; Sparks et al., 1990).

3

In places such as the southwestern United States, climate and drought regimes that are
unprecedented in at least the last 1000 years, in tandem with changing human land use and fire
suppression, have provided the conditions for the most intense and large-scale fires of the
historical era to occur (Allen, 2014). Because changing disturbance regimes have been relatively
recent and extreme on the scale of forest development, the impacts and existence of legacies,
persistent effects of past events, are poorly understood (Johnstone et al., 2016). Additionally,
stabilizing interactions and resistance of mature vegetation to minor disturbances in longstanding intact forests can obscure reduced resilience of a system until more intense disturbance
occurs (Ghazoul et al., 2015). A mature forest may not display signs of acute distress until a
single fire or pest outbreak causes uncharacteristically major damage from which the ecosystem
is unable to recover.
In many cases, a single disturbance may not be enough to initiate regime shift, but linked
disturbances, such as the effects of wildfire after increased tree mortality following a bark-beetle
outbreak, can be exponentially severe (Simard et al., 2011). Disturbances at various time scales
and intensities select for specific life-history traits that are important to consider when assessing
current ecosystem health and future trajectories. While regular flooding and fire regimes are
normal in most river systems, more frequent larger-scale and higher-intensity disturbances that
cause high mortality can reduce or eliminate the possibility that long-established communities
are able to repopulate their historic ranges, and altered structural and functional states may
emerge (Harmon et al., 1986).
In many cases, long-standing mature, diverse forest can be burned and repopulated by a
more limited range of trees, grasses, and shrubs, with distinct ecosystem functions and services
for animals and humans. This shift from one ecosystem type to another can be characterized as a

4

regime shift or “landscape trap,” alluding to the less-desired or compromised functions offered
by the new ecosystem (Lindenmayer et al., 2011). However, the question of ascribing value to
specific landscapes and evaluating best management practices is one that is strongly sitedependent and defined in accordance the interests of decision-making stakeholders, and beyond
the scope of the current study (Nassauer, 1997; Richardson et al., 2007). Assessing riparian
ecosystem trajectories, recovery, and resilience following recent wildfires will be critical for
addressing management needs and defining reasonable restoration goals and trajectories.
Minimal tree regeneration, or regeneration of drastically altered communities, following recent
disturbances from wildfires in the southwestern United States currently indicates that many
forests may have already reached tipping points of regional-scale forest ecosystem change
(Allen, 2014).

Fire and flood regimes and climate change
A prominent effect of the changing global climate has been wildfires of increased
frequency, intensity, and extent (Sun et al., 2019). Changes in fire season lengths and wildfire
frequencies are expected to increase in magnitude substantially over 74% of land worldwide,
especially in the United States, Canada, Brazil, China, and Australia (Sun et al., 2019).
Immediate effects of such fires include the destruction of mature established vegetation, animal
habitats, release of CO2 into the atmosphere, and threats to agriculture, human homes and air
quality (Zybach et al., 2009). Although many forests are well-adapted to regular surface burns,
intense large fires can fundamentally alter landscape cover composition when they occur at
unprecedented scales (Harmon et al., 1986).
In riparian areas, additional impacts of extreme fire disturbance may include drastically
increased stormflow, sediment transport, and flooding in areas where vegetation historically

5

absorbed or slowed runoff of much of the water following seasonal rains (Bock & Bock, 2014;
Dwire & Kauffman, 2003). The changed dynamics following the extirpation of existing
vegetation contribute to changes in the geomorphology of river channels and floodplains,
altering physical structure of established river systems, and fundamental characteristics of
habitats that have been historically adapted to smaller-scale, intermittent disturbances.
Depending on a river’s natural geomorphology, native species, and disturbance legacies, river
banks may be largely held together by vegetation that is rapidly established in the aftermath of
intense disturbance. Subsequent destabilization of river banks during vegetative dormancy can
then produce positive feedback loops of intensifying flood and fire disturbance regimes that
prevent re-establishment of previously existing vegetation (Tickner et al., 2001).

Adaptations of riparian vegetation
Riparian ecosystems experience flooding, sediment deposition, and other physical
disturbances requiring colonizing species to possess specialized adaptations that allow for their
establishment and survival (Naiman, Decamps, et al., 2005). Specific adaptations include stem
flexibility, root suckering, ability to reproduce from plant fragments, water resistance, rapid root
growth, resprouting from hardy below-ground rhizomes, and seed dispersal methods that are
aided by seasonal hydrology and intermittent floods (Allen, 2014; Naiman, Decamps, et al.,
2005). Abiotic and biotic legacies left by disturbances such as avalanche, flood, wind, fire,
drought, disease, litter accumulation, herbivory, and other physical influences, have enduring
impacts on vegetation communities and cover types (Naiman et al., 1998).
As colonizers of naturally dynamic systems, riparian species and vegetation communities
exhibit high resistance and resilience to disturbance. Holling (1973) describes resilience as “a
measure of persistence of systems and of their ability to absorb change and disturbance and still

6

maintain the same relationships between populations or state variables,” contrasted with
stability, or resistance, the “ability of a system to return to an equilibrium state after a temporary
disturbance” (emphasis added). Specific areas of the riparian system may favor different
colonizing vegetation more than others, and a more limited set of sites will be suitable for longestablishment of plant species (Harper, 1977).
In places where erosional processes dominate, plant communities tend to be resilient to
disturbance; breakage and loss of biomass are likely, but relationships between species will
persist and the community will re-establish over time. In contrast, where sediment deposition is
the dominant process, plants may be resistant to anoxia and burial, but less likely to suffer
breakage or other disturbance from high-velocity water flow that significantly alters community
dynamics and relationships (Bornette et al., 2008). A resistant system, subject to unusual
disturbance, may not possess the intrinsic resilience to return to its previous state.
Understanding what factors determine the structure and composition of riparian
vegetation communities, and how they will respond to regular and extreme disturbances, will be
critical to managing riparian zones for long-term resilience and maintenance of desirable
ecosystem services. Merritt et al. (2010) propose a framework of organizing riparian plants into
non-phylogenetic groupings, or guilds that share traits related to their role in the hydrologic
regime: life history, reproductive strategy, morphology, adaptations to fluvial disturbance, and
adaptations to water availability. Organizing vegetation types by their functions in the ecosystem
requires a complex understanding of interactions and relationships between traits and
environmental gradients (Merritt et al., 2010). However, such a framework may make clear
functional possibilities that would not be apparent if managing to maintain or restore species
composition.

7

Healthy coniferous forests generally exhibit some degree of fire resistance, and are able
to re-establish following disturbance by dispersing seed from undisturbed patches to adjacent
zones (Haffey et al., 2018). Forest understories in surface fire-prone landscapes often display
high resilience, with previously existing communities returning to the understory after lowintensity burns (Naiman, Decamps, et al., 2005). However, greatly increased patch size of highseverity burns can reduce or eliminate the possibility of seed dispersal to open areas by
previously prolific species (Johnstone et al., 2016). In contrast to surface fires, high-severity fires
reach the forest crowns and cause mortality of trees that are well-adapted to surface burns.
In these situations, species known as “ruderal” or ‘r’ strategists tend to dominate.
Characteristics of ‘r’ strategists include: small size and limited lateral spread, short life spans,
high relative growth rates where nutrients are available, early maturation, frequent flowering,
production of a large number of small seeds with wide dispersal potential, establishment of a
large seed bank, and persistence of dormant seeds or zygotes (Bornette et al., 2008; Grime, 2006;
Kautsky, 1988; Southwood, 1988). Both native and non-native plant species can exhibit ‘r’
strategies and and fill important ecological niches. Less disturbed areas are likely to be recolonized by seeds and propagules from adjacent patches, and may continue to resemble the
vegetation communities in their immediate surroundings; whereas more highly disturbed areas
over larger extents are more likely to be repopulated by extremely resourceful ‘r’ strategists
colonizing across great distance to the detriment of slower-growing species (Connell & Slatyer,
1977). Where ‘r’ strategists are early colonizers, their high relative growth rates and prolific seed
production in low-competition environments can then quickly exclude slower-growing
vegetation, establish positive feedbacks mechanisms, and fundamentally alter the plant
community in the medium- or long-term (Richardson et al., 2007). Various models of succession

8

distinguish between different vegetation types and their propensity to colonize and establish in
riparian zones, with competition and disturbance dynamics, and interactions between species,
heavily influencing the trajectory of vegetative community development (Naiman et al., 1998;
Oliver et al., 1996).

Implications for viability of forest recovery and management
Riparian vegetation communities respond to disturbances such that specific features of
the system measured individually may be dynamic, however over scales of decades or centuries,
ecosystem organizations tend to exhibit quasi-equilibrium in the absence of extraordinary
disturbance (Naiman, Decamps, et al., 2005). Regular flood and fire pressures on riparian
floodplains contribute to a shifting habitat mosaic where similar but heterogeneous habitats
coexist and convert to different types in the context of disturbance (Kleindl et al., 2015; Mouw et
al., 2009). Extreme disturbance events can contribute to widespread mortality of established trees
and other species, fundamentally altering community composition and microclimates, and
preventing the re-establishment of similar ecosystems (Benda et al., 1998, 2004; Harmon et al.,
1986).
Many species are well-adapted to regenerating after fire by a variety of means; however,
others are unable to re-establish, especially when burns are of larger extents and intensities than
populations have historically experienced (Grime, 2006). When those opportunistic species
rapidly re-establish following disturbance, they can exclude the conditions for populations of
other slower-growing species to return, even if propagules or seeds are available (Richardson et
al., 2007). Drastic and immediate shifts in species composition can confer changes in capacity of
the system to filter stormwater and mitigate flooding, sequester carbon, and provide habitat to
native animals (Sweeney et al., 2004).

9

The quaking aspen (Populus tremuloides), a tree species native to the American
Southwest that resprouts from clonal root systems and long-distance seed dispersal, is facing
regional decline linked to drought stress from warming temperatures, and declining precipitation
that does not provide suitable habitat for the species in its historic range (Rehfeldt et al., 2009;
Worrall et al., 2013). Historically, trees would resprout from below-ground tissue following
surface fires, however the combination of a warming climate and drought do not allow for young
trees to regrow. A better understanding of how to manage for resilient landscapes in the context
of new climate conditions and disturbance regimes may be necessary to maintain biodiversity
and healthy ecosystems.
With patterns of increased tree mortality from intensifying fire disturbances, disease, and
pest outbreaks being seen across western North America, land managers need cohesive direction
and robust information to revise goals and projected outcomes at local and regional levels
(Meddens et al., 2012; Raffa et al., 2008; Westerling et al., 2006). As vegetation communities
change in response to novel disturbance and climate, ecosystem services and functions, such as
habitat, nutrient cycling, water filtration and distribution, and carbon sequestration, will be
altered as well. Riparian systems consist of several vegetation and cover types, with no single
site performing all desirable functions at once (Findlay et al., 2002). Restored habitat must
similarly consist of a patchwork of ecosystem, vegetation, and cover types to perform a suite of
functions. However, choosing which ecosystem services to invest in and ascribe value to are
management decisions that must be made in consultation with and consideration of local
communities and stakeholders (Nassauer, 1997; Richardson et al., 2007). While there may not be
consensus on how and what to manage for, beginning to address these questions in the context of

10

shifting regimes can help motivate effective long-term projects possessing social and political
will for success.

Stream segment classification
Categorization of distinct stream segments that share characteristics such as geomorphic
condition, recovery potential, and levels of management priority, may be one of the most
important first steps in formulating robust network-based maps that assist in advancing
management goals (Buffington & Montgomery, 2013; Naiman, Decamps, et al., 2005; O’Brien
et al., 2017). Understanding local processes and linkages that connect rivers laterally,
longitudinally, and vertically, can help managers identify refugia, healthy nodes of riparian
vegetation that may act as seed and propagule sources, and legacies that support resilience in the
context of subsequent disturbance, which may become apparent at finer scales of reaches,
habitats, or microhabitats (Johnstone et al., 2016; Naiman, Decamps, et al., 2005; Sedell et al.,
1990). Once specific segments can be identified as particularly susceptible or resilient to
disturbance, opportunities and limitations of restoration efforts can be evaluated, and
management can be maximized for successful enhancement of desirable services and functions
(Naiman, Decamps, et al., 2005).
The loss of mature trees across widely disturbed areas, in tandem with present and
projected increases in forest drought-stress index (FDSI), suggests that regeneration needs of
specific species are less likely to be met. This includes factors such as availability of seed, which
must coincide with favorable climatic conditions for successful germination. Management with
attention to appropriate scales and intervals that can support regeneration of productive
landscapes are needed (Jackson et al., 2009; Johnstone et al., 2016; A. P. Williams et al., 2013).

11

Bornette et al. (2008) suggest a model for predicting organization of plant communities in river
floodplains that considers:
i. the nature of the physical constraints that affect plant communities (the scouring or
depositing character of flood disturbances)
ii. the frequency and intensity of disturbances that limit competitive interactions and create
gaps for recruitment for new individuals, and ultimately impede plant colonization
iii. the specific life-history traits that allow plant maintenance, recruitment and colonization
in the variously disturbed riparian systems.
This model can help categorize segments by vegetative communities possessing species-specific
traits that allow them to persist, given localized physical constraints and expectations of
disturbance.
Given the context of changing climate and disturbance regimes, it should be expected that
without intervention and with altered disturbance regimes and climatic conditions, forests will
convert into different ecosystems, including grasslands and shrublands, or alternative forests
with different dominant species (Jackson et al., 2009; J. W. Williams & Jackson, 2007).
Understanding how specific segments and reaches have been disturbed, and what habitat
potential they possess, will be critical to interpreting landscape change and developing effective
management plans. Even in places where legacies such as nurse structure persist, which provide
critical shade to vulnerable seedlings establishing in the semiarid landscape, changes to climate
and disturbance regimes may lead to recruitment failures at otherwise suitable sites (Haffey et
al., 2018; Johnstone et al., 2016). Broad changes in elevational distribution and dominance of
many plant species have already been documented, and suggest that novel patterns will continue
to emerge over the course of the 21st century with projected warming and drought conditions

12

(Allen, 2014). However, consequences ranging from loss of biodiversity, to reduced carbon
sequestration and depletion of water supplies due to increased uptake from novel vegetation,
must be considered when deciding where and how to intervene (Allen, 2014; Hicke et al., 2012).
Brierley and Fryirs (2013) propose an organizational framework of geomorphic
assessment consisting of four steps: (1) river classification (i.e. segment classification), (2)
geomorphic condition assessment, (3) recovery potential analysis, and (4) development of a
management plan that strategically addresses restoration and rehabilitation goals. Performing a
careful assessment in this manner can help inform effective restoration projects and avoid
expensive mistakes; managers can attempt to work with, rather than against nature, to support
resilient and diverse ecosystems that benefit people, communities, and native bioassemblages
(Haffey et al., 2018).
Well-managed river floodplains should provide functions such as sediment retention,
floodwater attenuation, nutrient absorption, erosion control, and biodiversity (Hughes et al.,
2013; Meitzen et al., 2018). Rather than aiming to reconstruct an ecosystem state that existed at a
specific point in time, management can seek to restore functional processes and provide habitat
for threatened native flora and fauna (Richardson et al., 2007). Management of invasive species
should respect the function that they now serve as part of a novel ecosystem, and avoid costly
and short-sighted attempts at total eradication that could jeopardize the stability of riparian
geomorphology and riverbanks (Tickner et al., 2001). Recognition that novel ecosystems in
drastically altered riparian ecosystems will develop and can be managed to enhance biodiversity
and functional ecosystem services will be critical for restoration success (Brooks et al., 2004;
Sarr et al., 2005). However, many invasive plant species are early seral colonizers and thrive in
low-competition environments. If native species fail to establish, positive feedback mechanisms

13

may be triggered that support the continued dominance of invasive species to the exclusion of
native vegetation (Richardson et al., 2007). In areas where forest regeneration is unlikely due to
the limitations of changing environmental conditions, planting or seeding dense native
herbaceous groundcover species in burned areas to promote the establishment of desired
functional species rather than early colonizers that confer degraded environmental conditions,
such as annual cheatgrass, which does little to prevent soil erosion or impede fire recurrence
(Haffey et al., 2018). However, there are significant challenges to engineering novel landscapes
that have no historical precedent, and where complex interactions between abiotic and biotic
factors are poorly understood. Invasive species can serve critical functions within novel
ecosystems, and management should not necessarily focus exclusively on extirpation of those
species (Hobbs et al., 2009). Despite some challenges and limitations, directly researching a
changing landscape may be the most effective means of understanding it, since suitable reference
systems are globally rare (Richardson et al., 2007).
Restoration and management decisions require the definition of priorities in consultation
with stakeholders, including traditionally-associated indigenous people, setting of multi-step
goals, and continued monitoring and re-evaluation of priorities and progress (Richardson et al.,
2007). Status quo management of fire suppression, and optimistic restoration projects involving
the re-planting of disturbed species that are unable to effectively re-establish, are exceedingly
expensive and reap limited rewards (Stephens et al., 2013). Managing for reduced flammability,
as could be achieved through establishment of drought-tolerant deciduous-dominated landscapes,
could help mitigate impacts of changing climate on fire frequency (Kelly et al., 2013).
Controlled burning, mechanical tree harvesting, and extensive ground mulching could also
reduce forest densities, competition for water, and fine fuel connectivity, in turn alleviating some

14

of the forest drought stress and risk of high-intensity and wide-ranging fires (Ager et al., 2010;
Finney et al., 2008). Additionally, in places where re-establishment or preservation of standing
forests is still possible, pursuit of effective management strategies should be prioritized to reduce
the likelihood of continued loss of threatened ecosystems, and the initiation of positive feedbacks
that degrade landscape functionality (Haffey et al., 2018). Understanding the structural
patchwork of vegetation, geomorphic conditions, and how both forest and non-forested
landscapes can provide continued ecosystem services and fulfill management goals in both broad
and hyper-local contexts will be critical as managers are forced to seek new solutions in a
changing world.

History of land use and fires in Bandelier National Monument, New Mexico
Human land use and habitation has coincided with the stabilization of regional forest
patterns over the last several thousand years on the Pajarito Plateau in North America, and
corresponded with changes to fire regimes over at least the last several hundred years (A. P.
Williams et al., 2013). Examination of fire scars around prehistoric Jemez village sites in the
uplands of the study area show that few fires occurred in and around occupied village sites prior
to European colonization, a pattern that is likely explained by continual fuel use by human
populations that prevented the unimpeded spread of fire across the landscape (Swetnam et al.,
2016). Evidence from tree-ring studies suggest that low-intensity surface fires that do not reach
the crowns of ponderosa pine and mixed conifer forests were common throughout the Holocene.
However, the size and intensity of recent stand-replacing fires is likely unprecedented since the
establishment of regional climate, vegetation and fire regime patterns around 6,000-9,000 years
before present (Allen, 2014; Anderson, Jass, et al., 2008; Fulé et al., 2014; Swetnam & Baisan,
2003).

15

Prior to contact with Spanish colonists, indigenous Puebloan people lived in and around
the Pajarito Plateau for thousands of years. It is estimated that 5000-8000 people lived in an
approximately 500 km2 area from ca. 1300-1640 C.E., on par with the present-day definition of
wildland-urban interface, where human land use interacts with and is strongly impacted by the
natural environment and ecosystem (Radeloff et al., 2005). Unlike conventional wilderness
management which seeks to eliminate human influence on the landscape, indigenous populations
and cultures were integrated directly with the natural environment and ecosystems, and
traditionally were participants in rather than disruptors of environmental regimes (Swetnam et
al., 2016). Human habitation of upland village sites prior to 1600 corresponded with continuous
clearing of fuelwood and reduced landscape fire connectivity. Weak associations between fires
and interannual climate variations within the Jemez Mountains prior to 1680 indicate that fires
were likely set by humans and kept in check by continuous land and fuel use (Swetnam et al.,
2016). Small-scale, low-intensity fires were common and frequent, disturbing patches of forest
that would leave mature trees standing, clear the understory, and open ecological niches for a
diverse and resilient forest community to thrive (Swetnam et al., 2016). Indigenous history is
corroborated by preserved pollen and tree-ring records that tell the same story of shifting land
use and occupation patterns over the last millennium.
The earliest megadrought in the pollen and tree-ring record occurred during the 1200s,
and corresponded with widespread tree mortality as well as changing land use practices by the
San Ildefonso Pueblo people (Whitman & Whitman, 1947; A. P. Williams et al., 2013).
Subsequently, several Pueblo villages were established on the Pajarito Plateau from the late 12th
century until the 16th century, when drought pressures forced plateau-dwellers into the valley of
the still-flowing Rio Grande River (Merlan & Levine, 2000; Whitman & Whitman, 1947;

16

Wilson, 2013). By the time the first Europeans arrived in the 16th century, the only permanent
habitation on the plateau was the town of Tsirege. This name, meaning “bird-place” in Tewa,
lent its meaning to the Spanish place name of the Pajarito, or “Little Bird” Plateau (Harrington,
1920; Merlan & Levine, 2000). Although the plateau might have been repopulated following the
drought in the absence of colonization, European influence permanently altered historic patterns
of habitation such that the uplands of the study area have not been home to any permanent
settlements since approximately 1700 C.E. (Swetnam et al., 2016).
Depopulation, the introduction of livestock grazing to the region, and ensuing policy of
fire suppression, contributed to the densification of mountainous forests in the region (Allen,
2014). Changes in land use that started in the dry period prior to Spanish arrival and continued
through the colonial period known as congregación provided the conditions for forests to
proliferate in and around former village sites at higher densities (Swetnam et al., 2016). Healthy
ponderosa pine forests historically consisted of an open understory where low-intensity surface
fires were common; as conditions changed, stem density increased in some places on the order of
tenfold or more, from less than 100 to over 1000 trees per acre (Allen et al., 2002). The increased
fuel load and development of a tall deciduous understory provided the conditions for several
large scale stand-replacing fires to burn across tens of thousands of acres (Allen, 2014).
Following the establishment of forests of unprecedented density, increasingly widespread and
high-intensity fires have initiated drastic shifts in ecosystem organization (Allen, 2014). In the
period from 1500 to 1680 C.E., only two widespread fires were recorded; in the following 180year period ending in 1860, nine widespread fires occurred across the Jemez Mountains
(Swetnam et al., 2016). This time period, following colonization but preceding any organized
changes in management, was already seeing significant changes in the typical fire regime.

17

In the period following 1860 through the present day, cattle ranching came to dominate
the landscape, and suppressing forest fires became a high priority for local and regional forest
managers (Stephens et al., 2013; Swetnam et al., 2001; Swetnam & Baisan, 2003). Especially
wet periods between 1905-1922 and 1978-1995 further promoted the establishment of dense
woody vegetation (Allen et al., 2002). For most of the 20th century, fire suppression dominated,
and the increasing standing biomass was not interpreted as a loss of resilience. Standing biomass
likely reached unprecedented levels, with tree densities of ponderosa pine and mixed-conifer
forests increasing ten-fold or more, and greater proportions of shade-tolerant and fire-sensitive
species proliferating (Covington & Moore, 1994). This set the stage for more high-severity
stand-replacing fires to occur throughout the region, including the La Mesa (1977), Dome
(1996), Cerro Grande (2000), and Las Conchas (2011) fires that burned through the Bandelier
National Monument (Allen, 2014; Veenhuis, 2002). Evidence from preserved bog records dating
back over 15,000 years suggests that the sizes of patches recently disturbed by high-severity fires
across southwestern ponderosa pine forests is likely unprecedented in the last 6,000-9,000 years,
around the time that regional climate, vegetation, and fire regime patterns had stabilized up until
Western colonization (Anderson, Jass, et al., 2008; Fulé et al., 2014).
Contextualized regionally within the southwestern United States, the Pajarito Plateau and
Bandelier National Monument have well-documented records of forest drought-stress index
(FDSI) from comprehensive tree-ring data sets preserved across the region (A. P. Williams et al.,
2013). The reconstructed FDSI is critical for our understanding of how climatic trends influence
tree growth and mortality, and for making predictions about how future changes will influence
forest health and resilience. Examining how observed climate data from 1896-2007 is associated
with recent tree-ring-derived FDSI values, it appears that warm-season vapor-pressure

18

differential, which is largely a factor of temperature and cold-season precipitation, are the
strongest predictors of FDSI (A. P. Williams et al., 2013). This information, taken with
predictions of decreased cold-season precipitation and increased temperatures over the course of
the 21st century, suggests that FDSI is on track to reach its most severe in at least a millennia by
2050 (Breshears et al., 2005; A. P. Williams et al., 2013). Additional observed correlations
derived from climate records and satellite measurements indicate an exponential correspondence
between FDSI and areas burned by recent wildfires, compounding the effects of changing land
use on fire regimes (A. P. Williams et al., 2013).
Over the course of the 19th and 20th centuries, conflicting interests between
archaeologists, homesteaders, and livestock owners on the Pajarito Plateau called for the
establishment of more comprehensive and centralized management to address different
stakeholders’ needs (National Park Service, 2015). Following years of contention and discussion,
Bandelier National Monument was designated by Woodrow Wilson in 1916, with the stated
purpose of “reserving these relics of a vanished people,” making explicit the assumption that the
indigenous people and cultures traditionally associated with the land had been extirpated (Merlan
& Levine, 2000). In reality, Pueblo and Navajo communities and governments persist as living
cultures and autonomous nations within the borders of the United States, with their own
traditional uses and relationship with the area. Through review of literature and primary
consultation with tribes, Merlan and Levine (2000) determined that six Pueblos: the Cochiti, San
Felipe, San Ildefonso, Santa Clara, Santo Domingo, and Zuni, are traditionally associated with
the land that has been designated as Bandelier National Monument. Three other communities,
the San Juan Pueblo, Zia Pueblo, and Hopi Tribe, also assert historic or traditional relationships

19

with the area, and the Navajo Nation additionally has noted that at least four Navajo clans likely
have origins in Puebloan communities of the Rio Grande (Merlan & Levine, 2000).
Defining appropriate management goals necessarily requires consultation with
indigenous communities and understanding the cultural context and history of land use in the
area. After beginning the process of performing an ethnographic investigation of the area in the
late 20th century, anthropologists Merlan and Levine soon changed their focus from conducting
empirical research to establishing partnerships. By the second phase of their work, in
consultation with the traditionally associated tribes, they expanded their project to primarily
focus on the formation of a consultation committee, with representatives of the six
aforementioned Pueblos. The tribes “wanted to emphasize government-to-government
consultation on management issues rather than ethnographic research” (Merlan & Levine, 2000).
This trajectory is reflective of the historic relationship between colonial establishments and
indigenous culture, wherein Native American communities of practice and bodies of knowledge
are perceived as extinct and something to be studied and evaluated through an external lens,
rather than living through active cultural practice that continues to be passed from generation to
generation. Management today is under the primary jurisdiction of the National Park Service,
and is performed in consultation with indigenous governments. However, United States federal
agencies continue to serve as primary decision makers in management decisions.
Within Bandelier National Monument, stated goals by the National Park Service include
protecting and preserving ancestral Pueblo archaeological sites, as well as cultural and natural
resources for the living Pueblo cultures traditionally associated with the Monument (National
Park Service, 2015). With fundamentally altered native plant communities, disturbance regimes,
and geomorphological and hydrologic conditions within the Bandelier National Monument,

20

management must recognize that riparian systems are open and dynamic, and historically and
presently influenced by human practices and use. Today, it seems that unprecedented conditions
and ensuing drought have set the stage for several of the most severe and widespread wildfires
on record to occur throughout the region, converting conifer-dominated landscapes to other
cover types for likely the first time in the Holocene Era (Anderson, Jass, et al., 2008; Fulé et al.,
2014).

21

Chapter Two: Research Manuscript
Introduction
River systems and floodplains are resilient to seasonal and intermittent flooding and
depend on them to distribute nutrients, organic matter, sediment, and organisms. However,
dramatic changes to flood regimes in either direction can produce novel reinforcing feedback
loops and result in establishment of a degraded ecosystem that is resilient to newly intensified
disturbance regimes (Sparks et al., 1990). Significant changes to the hydrology of riparian
systems and geomorphic conditions can alter vegetative communities that are specifically
associated with distinct fluvial landforms, and prevent them from reestablishing due to lack of
recruitment and unsuitability of altered habitat (Hupp & Osterkamp, 1985). A thorough analysis
of changes to vegetative cover over time can help extrapolate the effects of fire disturbance in
analogous situations as unprecedented disturbances continue to occur worldwide.
The Bandelier National Monument in New Mexico, USA, exhibits diverse vegetation
types and habitats at different elevations, including juniper savannas, piñon-juniper woodlands,
canyon-wall shrublands, ponderosa pine (Pinus ponderosa) forests, mixed conifer and riparian
forests, and montane grasslands (National Park Service, 2015). From the onset, federal protection
of the land occupied by the Bandelier National Monument has been motivated by the presence of
rich archaeological and cultural artifacts belonging to the Pueblo people. These artifacts and
historical sites are the legacy of the living Pueblo cultures that have been displaced by drought
and colonization. Today’s management incorporates the goals both of preserving archaeological
remains and ecological functions and value within the Monument (National Park Service, 2015).

22

The present study focuses on the Frijoles and Capulin canyons of the Pajarito Plateau,
from their headwaters in the Eastern Jemez Mountains to their confluence with the Rio Grande in
Northeastern New Mexico, USA, and the effects of wildfire and flood disturbance on vegetative
cover since the early 1990s. Large, high-intensity stand replacing fires have taken place over
great extents in the American Southwest, in an area with well-preserved bog and tree ring
records indicating that present climactic and disturbance conditions are unprecedented in at least
the last 6,000-9,000 years (Anderson, Allen, et al., 2008; Anderson, Jass, et al., 2008).
Quantifying how these disturbances have affected the landscape in the immediate aftermath and
years following disturbances will be critical for addressing management needs and defining
reasonable restoration goals and trajectories.
How have recent wildfires and floods have altered the composition of riparian canyon
vegetation in Bandelier National Monument? To answer this question and quantify changes to
landscape cover, remotely-sensed imagery was manually reviewed to determine proportions of
cover type occupied by forests, smaller vegetation, and bare ground. The results were compiled
by stream segment and categorized by type and degree of disturbance. Classifying riparian zones
contributes to a framework for riparian management that recognizes the unique internal
dynamics as well as costs and outcomes associated with managing distinct segments (Quinn et
al., 2001). This inventory of riparian vegetation will be used by the National Park Service to
evaluate management practices and make determinations for how to adapt to meet evolving
goals.

History of fire on Pajarito Plateau
Prior to the arrival of Europeans to the Pajarito Plateau in the 16th century, humans
coexisted with fire in the semiarid landscape, using fire for active management of the landscape,

23

and keeping their communities and villages safe by clearing and burning fuel wood around
inhabited sites (Swetnam et al., 2016). Small-scale, low-intensity, surface fires that did not reach
the crowns of mature forests were common and frequent during pre-colonial times, and would
disturb patches of the forest, clearing the understory, and opening ecological niches for a diverse
community to establish (Swetnam et al., 2016). The estimated density of the Plateau from 13001640 C.E is on par with the present definition of wildland-urban interface, which today are
managed for their high risk of environmental consequences on human populations, however
weak associations between fires and interannual climate variations during precolonial times
indicates that fires that would be destructive to human habitations were likely kept in check by
continuous land use and effective management (Haight et al., 2004; Swetnam et al., 2016).
During periods of drought, indigenous populations would move to areas where water was more
plentiful, however regular human migrations provided for long-term large-scale management of
fire (Merlan & Levine, 2000).
After colonization, changing land use and fire suppression prevented regular surface
fires, and standing forest biomass multiplied by as much as ten-fold, with increased proportions
of fire-sensitive species proliferating (Covington & Moore, 1994). Several high-intensity standreplacing fires have recently occurred throughout the region, including the La Mesa (1977),
Dome (1996), Cerro Grande (2000), and Las Conchas (2011) fires that burned areas of Bandelier
National Monument (Allen, 2014; Veenhuis, 2002). Today, intensifying forest drought-stress
index and an exponential correspondence between FDSI and recent severe wildfires, indicates
that the conditions that have given rise to the current fire regime and challenged the reestablishment of native forest vegetation are set to continue (Haffey et al., 2018; A. P. Williams
et al., 2013). For what is probably the first time since humans inhabited the region, conifer-

24

dominated landscapes characterized by frequent surface fires across the American southwest are
being replaced by novel cover types and disturbance regimes (Anderson, Jass, et al., 2008).

Direct effects of recent fires on Capulin and Frijoles Canyons
While the present study will serve to formally describe and quantify changes in
vegetative cover types, the immediate impacts of recent severe wildfires on the vegetative
communities and geomorphological conditions of the Capulin and Frijoles Canyons are already
clear. Mortality of mature trees was obvious in large areas of burned forest, and significant
changes to streamflow volume and sediment transport were immediately observed at gaging
stations as well (Veenhuis, 2002).
In the first season following the 1977 La Mesa and 1996 Dome fires, peak flow at most
downstream streamflow-gaging stations in the Frijoles and Capulin Canyons increased to about
160 times the maximum flow recorded before the fire (Veenhuis, 2002). As vegetation reestablished two seasons following the fires, maximum peak flow decreased to 10-15 times prefire maximums in the second year, and in the third year it decreased to 3-5 times pre-fire
maximums. In addition to maximum peak flow volumes being dramatically affected, the
frequency of large stormflow events increased markedly in the three years following the La Mesa
and Dome fires (Veenhuis, 2002). The flooding caused by high volume stormflow events had
drastic effects on vegetative cover composition even in areas that did not directly burn.
The previously existing riparian ecosystem provided habitat to native flora and fauna, as
well as ecosystem services that may not be fulfilled in the context of altered dynamics following
disturbance. Drastically altered vegetation dynamics may introduce positive feedback loops
encouraging the long-term dominance of shrub and herbaceous species, and in turn prohibiting

25

riparian forests from reestablishing and providing the same habitat and services that they did
before the disturbance (Richardson et al., 2007).
Changes in climate conditions and land use practices in the southwestern United States
may be preventing the re-establishment of some native vegetation following fires, which require
specific structural and environmental conditions, as well as availability of viable propagules to
re-establish following disturbance (Haffey et al., 2018). In a study of regeneration patterns in
ponderosa pine forest landscapes following eight fires in Arizona and New Mexico during an 18year regional drought from 1996 to 2013, researchers found evidence indicating that lowelevation, dryer areas, and areas further from conifer seed sources, were less likely to regenerate
with similar vegetation communities following disturbance (Haffey et al., 2018). Warmer
temperatures have generally been associated with increased tree mortality, and reduced success
for establishment of new seedlings, likely due to water stress from increased atmospheric vapor
pressure deficits (Allen, 2014; McDowell et al., 2011; A. P. Williams et al., 2013). Recent largeextent high-severity wildfires have caused mortality of virtually all tree seed sources across tens
of thousands of acres, with the consequence of grasses and shrubs achieving dominance without
competition (Allen, 2014). Similar patterns are likely to be observed in similar forests such as
those of the Frijoles and Capulin canyons, unless there is significantly more availability of seed
in the riparian systems, or human intervention. Quantifying changes to the canyons and
distinguishing between segment categories by disturbance will help inform effective
management that heavily weighs localized factors and characteristics.

26

Methods
Study area
The area of interest consists of the Frijoles and Capulin Canyon riparian ecosystems in
the Eastern Jemez Mountains of New Mexico, USA. It includes the areas from their headwaters
at 2825 m above sea level to their confluence with the Rio Grande at 1630 m. The study
encompasses the entirety of the riparian canyon bottoms. The Frijoles Canyon bottom consists of
154 hectares, including a main channel that runs about 19 km from the Rio Grande towards its
uplands before splitting into two: a northern fork extends four additional kilometers, and the
southern fork extends three kilometers from the junction. The Capulin Canyon bottom is 141
hectares, and consists of a main channel that splits approximately 17 km upstream of the Rio
Grande to a northern fork extending an additional 5 km and the southern fork just 2 km past the
junction (Figure 1).

27

Figure 1. Study area map showing Frijoles and Capulin Canyons, and three large recent fire
extents.

Sample design
Canyon bottom areas were designated using a bottomland boundary developed by the
United States Geological Survey (USGS). The procedure calculated the topographical position
index from a 10 foot DEM using R to identify connected raster cells that were associated with a
highly negative TPI (Assal et al., 2015). The polygons were then manually reviewed to fill in
holes and remove islands that appeared outside the canyon bottom so that each canyon bottom

28

was confined by a single contiguous border. This layer was divided into 1 km stream reaches for
exploratory purposes and was used as the focus layer for ArcGIS interpretation. A grid of
approximately 30 m x 30 m squares was created over the bottomland layer using the “Grid Index
Features” geoprocessing tool (Figure 2). We then selected only the squares that had their center
in the bottomland layer, excluding those whose centers did not fall within the area of interest. We
created points at the centers of the squares of the grid polygon layer, and buffered the points to
generate circles with a radius of 3 m. The sample design was developed to provide a thorough
representation of all stream reaches, with at least 40 buffered sample points per 1 km reach in
each river. The final sample consisted of 2094 points in Frijoles Canyon and 1911 points in
Capulin Canyon.

29

Figure 2. Buffered centers of grid squares in the Frijoles Canyon bottom alongside the Bandelier
National Monument Visitor Center.

Satellite interpretation
Aerial photographs and satellite imagery were obtained from the Los Alamos National
Lab (LANL) and United States Forest Service (USFS) for the years 1991 and 1992, respectively,
and from the National Agriculture Imagery Project (NAIP) for the years 2011, 2014, and 2018.

30

LANL and USFS imagery included 3 spectral bands that were symbolized with a standard
deviation clip to highlight extreme values for visual clarity. NAIP imagery was symbolized using
the color infrared filter and a standard deviation clip to highlight the presence or absence of
vegetation (Figure 3).

Figure 3. NAIP Imagery from 2011 and 2018 showing examples of sample point classified as forest (F),
other vegetation (O), wetted channel (W), bare ground (B), and shadow (S).

The sample array was overlaid on each set of imagery using ArcGIS Pro. Each point was
classified as one of seven cover types by visually interpreting the imagery (Table 1). The three
most abundant cover types were: large arboreal riparian vegetation (F), other, smaller woody or
herbaceous vegetation (O), and dry channels or bare floodplain sediment, which may include

31

dormant herbaceous vegetation (B). A small number of sample points landed on semi-permanent
reservoirs and were designated as wetted land (W). Additionally, some sample points were
shadowed (S), or did not have imagery available (U), and were excluded from analysis of
relative proportions of cover types.
A bare channel category (C) was designated using a spatial analysis. The locations of the
river channels in Frijoles and Capulin canyons were designated as a polygon layer using the
post-Las Conchas Fire 2014 imagery, when the channels were most disturbed. The channel was
determined by the boundary of the centrally scoured area, with no channel designated where
vegetation obscured the ground. A spatial analysis was performed on the data to designate a bare
channel category – these were sample cells that were both manually designated as bare ground,
and spatially coincided with the channel layer. Changes in this category may reveal patterns of
geomorphology affecting legacies and recovery of riparian vegetation. These cover types were
selected to accurately illustrate functional changes to the landscape at the highest possible level
of detail (Table 1).
Table 1. Criteria used to designate different cover types for NAIP imagery.

Cover Type
Forest
Other vegetation

Bare ground
Bare Channel
Wetted ground
Shadow
Unidentifiable

Criteria
Bright red spectral signature
Crown diameter >6 m
Bright red spectral signature
Scattered or uniform- may be small deciduous, coniferous, or herbaceous
Crown diameter <6 m
Infrared or green light not reflected in spectral signature
Bare ground cover that spatially coincides with the designated channel
Ground is covered by water
50% or more of cell is completely black
No imagery for sampled cell

Spectral signatures including color, texture, and shadow from satellite imagery were used
to manually distinguish each cover type. Each point was visually inspected at a scale of 1 to 1000

32

using the visible signature within each buffered circle to determine which cover type was present
at the sample. For each cell, the cover type that comprised the majority of the visible (nonshadowed) area of the buffered sample point was designated as the dominant cover type for the
sample. Points that were 50% or more shadowed were designated as shadow, however if at least
50% of the image within circle was visible, then it was interpreted according to the preceding
rule. Where the sample intersected the wetted channel, the wetted area was treated as shadow
and the rest of the circle was analyzed for interpretation. For points where the majority cover
type was not immediately obvious, the imagery was inspected at a larger scale; where
uncertainty remained, the point was flagged and revisited with a second reviewer, and a decision
was taken by consensus.

Data analysis
Data were summarized using R (R Core Team 2020) to explore the overall composition
and changes to cover type over time for each of the canyons. Initial analysis consisted of
calculating the proportional composition of each canyon for each year analyzed. Charts were
produced to visualize the cover proportions by 1 km reaches, and were used as the basis for
exploring changes to the landscape detectable through the remotely-sensed imagery. Snapshots
in time, as determined by aerial photographs (1991/92, 2011, 2014, and 2018), and disturbance
type were considered as variables that had an influence on cover proportions. The Dome (1996),
Cerro Grande (2000), and Las Conchas (2011) fires all occurred within the time period over
which the analysis spans and had direct and indirect effects on vegetative cover. Areas within
and downstream of the fires also experienced increased sediment discharge, stormflow, and
flooding in the years following burns, causing additional mortality of mature forest, and limiting
the reestablishment of vegetation.

33

Stream segment types were classified by disturbance and include (1) areas that were
directly impacted by burns, (2) areas flooded by high volume stormflow downstream of burned
areas and (3) areas that both burned and flooded. A permutation test was used to assess the
independence of land cover counts with disturbance type (i.e. burned and/or flooded areas),
stratified by year, in each canyon separately. This test was used to assess any changes to
vegetative cover that have occurred in the context of these disturbances, in contrast to the 1 km
reaches. The R package ‘coin’ was used to calculate an asymptotic general independence test
using χ2 values, which is a permutation-based calculation of the classical Cochran-MantelHaenszel test of independence (Hothorn et al., 2008).

Results
Initial analysis consisted of calculating the proportional composition of each canyon for
each year analyzed. The proportion designated as shadow ranged from 5% in Capulin 2018, to
25% in Frijoles 2014 (see Appendix). Shadowed sample points and those for which imagery was
not available were excluded from analyses to obtain weighted proportions of Forest, Other
vegetation, Bare ground, and Bare channel. Wetted ground consisted of less than 1% of sampled
points and was excluded from the analysis.

Frijoles Canyon
Parts of Frijoles Canyon burned in the 2000 Cerro Grande Fire, and 2011 Las Conchas
Fire. Less than 10% of the canyon burned in the Cerro Grande Fire, and over 75% burned in the
Las Conchas Fire, according to fire perimeters published by the Monitoring Trends in Burn
Severity (MTBS) Program. While analysis of additional snapshots in time, especially in the time
periods immediately following each fire, may further enhance our understanding of how cover

34

changes over time, looking at the points directly and indirectly affected by fires even years later
can reveal the types of lasting effects that stand-replacing fires incur.
Although a small decline in forest cover was observed between 1992 and 2011 in forest
cover, the cover change to the burned area was likely much greater in the years more
immediately following the 2000 Cerro Grande Fire, which is not reflected in the data due to the
19-year gap in measurement, and lack of data for the early 2000s. Likewise, the significant drop
in forest cover, and increases in other vegetation and bare ground cover, between 2011 and 2014
show changes that occurred as a result of the 2011 Las Conchas Fire (Table 2 and Figure 4).

35

Table 2. Proportions of vegetative cover type observed in Frijoles Canyon overall.

Cover type

Year1992 Year2011 Year2014 Year2018

Forest

73%

69%

17%

18%

Other vegetation

17%

21%

45%

71%

Bare ground

4%

5%

14%

4%

Bare channel

6%

6%

25%

8%

Figure 4. Proportions of vegetative cover observed in Frijoles Canyon overall.

36

Figure 5. Proportions of vegetative cover type observed by 1 km reaches in Frijoles Canyon.

Further analysis of the proportions of cover type by 1 km stream reaches shows
differences between the lower, middle, and upper segments, which have been variably affected
by both fire and flood. Labels indicate the kilometer reach represented, with “Frijoles01” being
the lowermost reach between kilometer markers zero and one, and so on (Figure 5). Frijoles

37

Canyon was not affected by the Dome Fire, and only the northernmost 1 km of Frijoles was
burned by a large fire in 2000. Any changes, apart from those located in the highest reach of the
northern branch of Frijoles, between 1992 and 2011 cannot be directly attributed to large
wildfires.
Following the Las Conchas Fire, in 2014, the effects of wildfire disturbance throughout
Frijoles Canyon are clear. A tributary at km 17.5 brought a large volume of water, sediment, and
debris to the main channel, and resulted in significant scouring between km 6 and 17.5
represented by the bare channel. From km 6 to km 12, there is some retention of forest cover
following the Las Conchas Fire, and from km 12 to 17.5 there is virtually no forest. In 2018,
much of that area had been revegetated by pioneer species, however reforestation has not yet
occurred (Table 3, Figure 6). The lowermost kilometer of Frijoles contains a wide delta that was
mostly categorized as bare ground, and anomalous compared to the rest of the system.
Vegetative cover counts in Frijoles Canyon were not independent of disturbance type while
stratified by year (χ2 = 843.6 and p < 0.0001, Figure 6).

Table 3. Classification and descriptions of segments by disturbance in Frijoles Canyon.

Frijoles km
markers
km 0-0.3

Segment Description of disturbance

km 0.3-1.5

Z

km 1.5-6

A

km 6-12

B

km 12-17.5

C

km 17.5headwaters

D

Y

Discarded from analysis; river delta, limited established
vegetation observed across time periods
Discarded from analysis due to extensive dark shadows in 2014
and 2018 imagery
Flooding downstream of fire, some disturbance of forest
vegetation with significant retention of forest cover
Direct burn and heavy flooding, significant loss of forest cover
following Las Conchas Fire
Direct burn and heavy flooding, nearly complete loss of forest
cover following Las Conchas Fire
Directly impacted by Las Conchas Fire, some retention of forest
cover

38

Figure 6. Proportions of cover types observed by disturbance segment and year in Frijoles Canyon.

Capulin Canyon
Capulin Canyon was unaffected by the 2000 Cerro Grande Fire, however experienced
nearly complete burning in the 1996 Dome Fire and 2011 Las Conchas Fire. A drastic decline in
forest cover is apparent between 1992 and 2011, and 2011 and 2014, during which these two
fires occurred. While the canyon was not as densely forested as Frijoles at the beginning of the
study, its decline from 57% to 7% cover overall is drastic (Table 4 and Figure 7).
39

Table 4. Proportions of vegetative cover type observed in Capulin Canyon overall by year.

Cover type

Year1992 Year2011 Year2014 Year2018

Forest

57%

29%

7%

3%

Other vegetation

32%

52%

61%

75%

Bare ground

6%

13%

8%

6%

Bare channel

5%

6%

25%

17%

Figure 7. Proportions of vegetative cover observed in Capulin Canyon overall by year.

40

Figure 8. Proportions of vegetative cover observed by 1 km reaches in Capulin Canyon.

The Dome Fire in 1996 burned most of the canyon north of km 4.5. The drastic change
from near-complete forest in the middle reaches to significantly less in 2014, and nearly none in
2014 and 2018 is clear. The bare channel is persistent in the lower reaches from 1992 with the
proportion varying little between years. Following the 2011 Las Conchas Fire, the channel

41

becomes apparently scoured from km 15 down in 2014, with some recovery of vegetation by
2018. The forest fails to recover to the proportion it represented in the early 1990s. Labels
indicate the kilometer reach represented, with “Capulin01” being the lowermost reach between
kilometer markers zero and one, and so on (Figure 8). Categorizing the canyon by reaches show
four distinct categories of disturbance (Table 5, Figure 9). Vegetative cover counts in Capulin
Canyon were not independent of disturbance type while stratified by year (χ2 = 2099.7 and p <
0.0001, Figure 9).
Table 5. Classification and descriptions of segments by disturbance in Capulin Canyon.

Capulin km
markers
km 0-4.5

Segment Description of disturbance

km 4.5-11

F

km 11-13.8

G

km 13.8headwaters

H

E

Multiple direct burns from km 1-4.5; area includes the river
delta, heavy flooding and regular scouring of vegetation;
absence of surface water for much of the year
Multiple direct burns, heavy flooding, nearly complete
disturbance of forest vegetation and limited reestablishment of
herbaceous vegetation; perennial flow ends upstream of km 4.5
Multiple direct burns and heavy flooding; significant
reestablishment of vegetation following disturbance
Multiple direct burns and heavy flooding, nearly complete loss
of forest cover following Las Conchas Fire, with substantial
reestablishment of smaller vegetation

42

Figure 9. Proportions of vegetative cover observed by disturbance segment and year in Capulin Canyon.

Discussion
Canyon bottoms in the Pajarito Plateau are of particular interest to land cover researchers
because they represent a portion of the semiarid landscape with the most availability of water
and propensity for propagules to disperse by wind, seasonal stream flows and floods. In the
aftermath of extreme wildfires, these areas are likely to experience the most extreme flood

43

disturbances, while also possessing some of the most opportune conditions for vegetation to reestablish. In this study, large differences were observed between vegetative composition of
canyon bottoms before and after intense fires and subsequent flood disturbances. Cover counts
within designated stream segments that had different degrees and types of disturbance occur
were shown to be independent of disturbance type.
In the context of occurrence of stand-replacing fires of unprecedented intensity and
extent, examining the way that vegetation develops in the immediate aftermath and extended
time period following disturbances will give insight to the development of novel ecosystems in
the study area, and more broadly as climate intensifies on a global scale. Anecdotal evidence has
suggested that certain types of vegetation cover have been slow or unable to reestablish.
Specifically, riparian alder (Alnus sp.) and birch (Betula sp.) may be locally extirpated from
Capulin; and yellow ladyslipper (Cypripedium sp.) and grape fern (Botrychium sp.) from Frijoles
Canyon. In the seven years following the Las Conchas Fire, which was the main disturbance
event in the temporal and geographic areas of the present study, forest cover has not recovered to
pre-fire levels in either Frijoles or Capulin canyons. In Capulin Canyon, the 1996 Dome Fire had
a large impact on forest cover proportions that is still clear in 2011, when the next set of aerial
imagery is available. Comparing the trajectories of the recovery in the aftermath of the Las
Conchas Fire between the two canyons may give insight into how single vs. multiple extreme
disturbance events variously impact these two parallel systems, and also be compared to studies
of riparian disturbance on broader regional and global scales.
The data analysis does not reflect more fine-scale observations that could be made were
more samples analyzed for the years between 1991-2011, and 2012-2013. In the time period
between 2014-2018, no large fires burned, and the most significant changes to cover type are a

44

transition from bare ground to other vegetation that may or may not progress to forest in
subsequent years. Continued monitoring of the area will help understanding and management of
forest re-establishment, and to determine and project if the “Other vegetation” that is establishing
on former bare ground will stabilize as forest cover or herbaceous vegetation.
Large stand-replacing fires and downstream removal of vegetation due to subsequent
stormflow produce directly observable impacts on the landscape. While fire is a natural part of
disturbance regimes, consistent processes of small-scale disturbance and revegetation would
result in relatively even proportions of cover types over time (Naiman, Decamps, et al., 2005).
Drastic shifts in cover type proportions thus likely indicate that the degree and extent of
disturbance is unusual. The large differences in vegetative cover following fire and flood
disturbances indicate that the system has been disturbed to a larger degree than usual, however it
is possible that over a longer period, the systems will recover to their pre-fire dynamics. To draw
consequential long-term conclusions about vegetation trajectories in the study area, repeated
measurements should be taken to track the development of vegetation on site, and comparisons
made to analogous situations regionally and worldwide.
The manual methodology employed, wherein buffered points were manually inspected, in
some cases multiple times and by multiple sets of eyes, was time-intensive and in future
research, more efficient methods may be used to quickly identify significant changes to
landscape cover in satellite imagery. Use of supervised or unsupervised raster classification, or
training artificial intelligence to interpret the landscape at a high level, as has been done in the
Sierra Nevada Mountains of California, could make interpretation of both large and small study
areas faster, more informative, and easier to implement on a larger scale (Parisa & Nova, 2020).

45

The manual methodology was selected for its simplicity, and the ability to specifically
track crown size of forest vegetation. The human eye can make decisions about textural
characteristics that a computer would not be able to if using a built-in classifier in ArcGIS Pro. In
lieu of more advanced AI algorithms, making manual decisions about each point allowed the
data to be reported with a high level of confidence and detail. However, further research could
compare the results of computer-classified results with those generated by a human, to determine
if the increased level of detail conferred by human decision-making is worth the additional
investment of time and resources. In the short-term, employing human eyes to perform
assessments such as the one in this study may be less resource-intensive because it does not
require the development of an advanced computer program and algorithm, and the associated
limitations. However, if the technology is developed and able to be deployed on a large scale
across regions, this could allow for easier replicability of the methodology, and provide deeper
insights and a wider basis for comparisons of large-scale disturbances.

Conclusion
The interpretation of remotely-sensed imagery of Frijoles and Capulin canyons within the
Bandelier National Monument quantifies the clear disturbances caused by intense large-scale
fires and floods throughout the area. While the fire regime appears to be unprecedented in the
last several millennia according to preserved bog records, continuing to study the ongoing
development and trajectory of vegetative changes will be essential to understanding how these
disturbances affect ecosystems and their functions in the short and long term (Anderson, Jass, et
al., 2008). In the absence of robust reference systems to which the disturbances in Bandelier can
be compared, directly studying the area to understand its development and changes may be the
most effective method of understanding local ecosystem dynamics.

46

Integrating analysis of vegetation dynamics with changing geomorphic conditions, and
adapting management decisions to ongoing research findings and novel developments, will help
to inform when and where intervention may be necessary, and track the effectiveness of actions
taken to mitigate or reverse damages inflicted by fire disturbance. Quantifying the effects that
disturbances have had on ecosystem services typically provided by the riparian systems, and
comparing those results to the changes to landscape cover over time, may provide further insight
for determining which types of vegetation are most important and effective for preserving the
critical functions that the site has historically provided. Comparing vegetative cover directly with
data sources such as digital elevation models and stormflow measurements may reveal additional
patterns and correlations between riparian variables.
Riparian forests in the arid and semiarid Southwestern United States serve as critical
habitat for a variety of plant and animal species, provide essential ecosystem services, and are of
immense cultural and recreational value to human populations. Changes to climate conditions
and disturbance regimes are triggering drastic shifts in vegetation and ecosystem dynamics, for
which recovery trajectories are unclear. Continuing to study and adapt management to preserve
functional ecosystems and habitats will be critical to ensure the continued existence of habitable
landscapes that preserve native ranges of local flora and fauna and manage for novel ecosystems
in the context of fundamentally or permanently altered dynamics.

47

References
Ager, A. A., Vaillant, N. M., & Finney, M. A. (2010). A comparison of landscape fuel treatment
strategies to mitigate wildland fire risk in the urban interface and preserve old forest
structure. Forest Ecology and Management, 259(8), 1556–1570.
Allen, C. D. (2014). Forest ecosystem re-organization underway in the southwestern United
States: A preview of widespread forest changes in the Anthropocene? Proceedings.
RMRS-P-71. Fort Collins, CO: US Department of Agriculture, Forest Service. Rocky
Mountain Research Station. p. 103-123., 71, 123.
Allen, C. D., Savage, M., Falk, D. A., Suckling, K. F., Swetnam, T. W., Schulke, T., Stacey, P.
B., Morgan, P., Hoffman, M., & Klingel, J. T. (2002). Ecological restoration of
southwestern ponderosa pine ecosystems: A broad perspective. Ecological Applications,
12(5), 1418–1433.
Anderson, R. S., Allen, C. D., Toney, J. L., Jass, R. B., & Bair, A. N. (2008). Holocene
vegetation and fire regimes in subalpine and mixed conifer forests, southern Rocky
Mountains, USA. International Journal of Wildland Fire, 17(1), 96–114.
Anderson, R. S., Jass, R. B., Toney, J. L., Allen, C. D., Cisneros-Dozal, L. M., Hess, M.,
Heikoop, J., & Fessenden, J. (2008). Development of the mixed conifer forest in northern
New Mexico and its relationship to Holocene environmental change. Quaternary
Research, 69(2), 263–275.
Assal, T. J., Anderson, P. J., & Sibold, J. (2015). Mapping forest functional type in a forestshrubland ecotone using SPOT imagery and predictive habitat distribution modelling.
Remote Sensing Letters, 6(10), 755–764.

48

Benda, L. E., Miller, D. J., Dunne, T., Reeves, G. H., & Agee, J. K. (1998). Dynamic landscape
systems. River Ecology and Management: Lessons from the Pacific Coastal Ecoregion.
Springer-Verlag, New York, 261–288.
Benda, L. E., Poff, N. L., Miller, D., Dunne, T., Reeves, G., Pess, G., & Pollock, M. (2004). The
network dynamics hypothesis: How channel networks structure riverine habitats.
BioScience, 54(5), 413–427.
Bock, C. E., & Bock, J. H. (2014). Effects of wildfire on riparian trees in southeastern Arizona.
The Southwestern Naturalist, 59(4), 570–576.
Bornette, G., Tabacchi, E., Hupp, C., Puijalon, S., & Rostan, J.-C. (2008). A model of plant
strategies in fluvial hydrosystems. Freshwater Biology, 53(8), 1692–1705.
Breshears, D. D., Cobb, N. S., Rich, P. M., Price, K. P., Allen, C. D., Balice, R. G., Romme, W.
H., Kastens, J. H., Floyd, M. L., & Belnap, J. (2005). Regional vegetation die-off in
response to global-change-type drought. Proceedings of the National Academy of
Sciences, 102(42), 15144–15148.
Brierley, G. J., & Fryirs, K. A. (2013). Geomorphology and river management: Applications of
the river styles framework. John Wiley & Sons.
Brinson, M. M. (1990). Riverine forests In: AE Luge, M. Brinson, and S. Brown (eds) Forested
Wetlands. Ecosystems of the World, 15.
Brooks, M. L., D’antonio, C. M., Richardson, D. M., Grace, J. B., Keeley, J. E., DiTomaso, J.
M., Hobbs, R. J., Pellant, M., & Pyke, D. (2004). Effects of invasive alien plants on fire
regimes. BioScience, 54(7), 677–688.

49

Buffington, J. M., & Montgomery, D. R. (2013). Geomorphic classification of rivers. In:
Shroder, J.; Wohl, E., Ed. Treatise on Geomorphology; Fluvial Geomorphology, Vol. 9.
San Diego, CA: Academic Press. p. 730-767., 730–767.
Connell, J. H., & Slatyer, R. O. (1977). Mechanisms of succession in natural communities and
their role in community stability and organization. The American Naturalist, 111(982),
1119–1144.
Covington, W. W., & Moore, M. M. (1994). Southwestern ponderosa forest structure: Changes
since Euro-American settlement. Journal of Forestry, 92(1), 39–47.
Didham, R. K., Watts, C. H., & Norton, D. A. (2005). Are systems with strong underlying
abiotic regimes more likely to exhibit alternative stable states? Oikos, 110(2), 409–416.
Dwire, K. A., & Kauffman, J. B. (2003). Fire and riparian ecosystems in landscapes of the
western USA. Forest Ecology and Management, 178(1–2), 61–74.
Findlay, S. E., Kiviat, E., Nieder, W. C., & Blair, E. A. (2002). Functional assessment of a
reference wetland set as a tool for science, management and restoration. Aquatic
Sciences, 64(2), 107–117.
Finney, M. A., Seli, R. C., McHugh, C. W., Ager, A. A., Bahro, B., & Agee, J. K. (2008).
Simulation of long-term landscape-level fuel treatment effects on large wildfires.
International Journal of Wildland Fire, 16(6), 712–727.
Fulé, P. Z., Swetnam, T. W., Brown, P. M., Falk, D. A., Peterson, D. L., Allen, C. D., Aplet, G.
H., Battaglia, M. A., Binkley, D., & Farris, C. (2014). Unsupported inferences of highseverity fire in historical dry forests of the western United States: Response to Williams
and Baker. Global Ecology and Biogeography, 23(7), 825–830.

50

Ghazoul, J., Burivalova, Z., Garcia-Ulloa, J., & King, L. A. (2015). Conceptualizing forest
degradation. Trends in Ecology & Evolution, 30(10), 622–632.
Grime, J. P. (2006). Plant strategies, vegetation processes, and ecosystem properties. John
Wiley & Sons.
Haffey, C., Sisk, T. D., Allen, C. D., Thode, A. E., & Margolis, E. Q. (2018). Limits to
ponderosa pine regeneration following large high-severity forest fires in the United States
Southwest. Fire Ecology, 14(1), 143.
Haight, R. G., Cleland, D. T., Hammer, R. B., Radeloff, V. C., & Rupp, T. S. (2004). Assessing
fire risk in the wildland-urban interface. Journal of Forestry, 102(7), 41–48.
Harmon, M. E., Franklin, J. F., Swanson, F. J., Sollins, P., Gregory, S. V., Lattin, J. D.,
Anderson, N. H., Cline, S. P., Aumen, N. G., & Sedell, J. R. (1986). Ecology of coarse
woody debris in temperate ecosystems. In Advances in ecological research (Vol. 15, pp.
133–302). Elsevier.
Harper, J. L. (1977). Population biology of plants. Population Biology of Plants.
Harrington, J. P. (1920). Old Indian geographical names around Santa Fe, New Mexico.
American Anthropologist, 22(4), 341–359.
Hicke, J. A., Allen, C. D., Desai, A. R., Dietze, M. C., Hall, R. J., Hogg, E. H., Kashian, D. M.,
Moore, D., Raffa, K. F., & Sturrock, R. N. (2012). Effects of biotic disturbances on forest
carbon cycling in the United States and Canada. Global Change Biology, 18(1), 7–34.
Hobbs, R. J., Higgs, E., & Harris, J. A. (2009). Novel ecosystems: Implications for conservation
and restoration. Trends in Ecology & Evolution, 24(11), 599–605.
Holling, C. S. (1973). Resilience and stability of ecological systems. Annual Review of Ecology
and Systematics, 4(1), 1–23.

51

Hothorn, T., Hornik, K., van de Wiel, M. A., & Zeileis, A. (2008). Implementing a class of
permutation tests: The coin package. Journal of Statistical Software, 28(8), 1–23.
Hughes, T. P., Carpenter, S., Rockström, J., Scheffer, M., & Walker, B. (2013). Multiscale
regime shifts and planetary boundaries. Trends in Ecology & Evolution, 28(7), 389–395.
Hupp, C. R., & Osterkamp, W. R. (1985). Bottomland vegetation distribution along Passage
Creek, Virginia, in relation to fluvial landforms. Ecology, 66(3), 670–681.
Jackson, S. T., Betancourt, J. L., Booth, R. K., & Gray, S. T. (2009). Ecology and the ratchet of
events: Climate variability, niche dimensions, and species distributions. Proceedings of
the National Academy of Sciences, 106(Supplement 2), 19685–19692.
Johnstone, J. F., Allen, C. D., Franklin, J. F., Frelich, L. E., Harvey, B. J., Higuera, P. E., Mack,
M. C., Meentemeyer, R. K., Metz, M. R., & Perry, G. L. (2016). Changing disturbance
regimes, ecological memory, and forest resilience. Frontiers in Ecology and the
Environment, 14(7), 369–378.
Kautsky, L. (1988). Life strategies of aquatic soft bottom macrophytes. Oikos, 126–135.
Kelly, R., Chipman, M. L., Higuera, P. E., Stefanova, I., Brubaker, L. B., & Hu, F. S. (2013).
Recent burning of boreal forests exceeds fire regime limits of the past 10,000 years.
Proceedings of the National Academy of Sciences, 110(32), 13055–13060.
Kleindl, W. J., Rains, M. C., Marshall, L. A., & Hauer, F. R. (2015). Fire and flood expand the
floodplain shifting habitat mosaic concept. Freshwater Science, 34(4), 1366–1382.
https://doi.org/10.1086/684016
Lindenmayer, D. B., Hobbs, R. J., Likens, G. E., Krebs, C. J., & Banks, S. C. (2011). Newly
discovered landscape traps produce regime shifts in wet forests. Proceedings of the
National Academy of Sciences, 108(38), 15887–15891.

52

McDowell, N. G., Beerling, D. J., Breshears, D. D., Fisher, R. A., Raffa, K. F., & Stitt, M.
(2011). The interdependence of mechanisms underlying climate-driven vegetation
mortality. Trends in Ecology & Evolution, 26(10), 523–532.
Meddens, A. J., Hicke, J. A., & Ferguson, C. A. (2012). Spatiotemporal patterns of observed
bark beetle-caused tree mortality in British Columbia and the western United States.
Ecological Applications, 22(7), 1876–1891.
Meitzen, K. M., Phillips, J. N., Perkins, T., Manning, A., & Julian, J. P. (2018). Catastrophic
flood disturbance and a community’s response to plant resilience in the heart of the Texas
Hill Country. Geomorphology, 305, 20–32.
https://doi.org/10.1016/j.geomorph.2017.09.009
Merlan, T., & Levine, F. (2000). Study of traditionally associated Native American communities
ethnographic overview and assessment. Bandelier National Monument.
Merritt, D. M., Scott, M. L., LeRoy Poff, N., Auble, G. T., & Lytle, D. A. (2010). Theory,
methods and tools for determining environmental flows for riparian vegetation: Riparian
vegetation-flow response guilds. Freshwater Biology, 55(1), 206–225.
Mouw, J. E. B., Stanford, J. A., & Alaback, P. B. (2009). Influences of flooding and hyporheic
exchange on floodplain plant richness and productivity. River Research and Applications,
25(8), 929–945. https://doi.org/10.1002/rra.1196
Naiman, R. J., Bechtold, J. S., Drake, D. C., Latterell, J. J., O’keefe, T. C., & Balian, E. V.
(2005). Origins, patterns, and importance of heterogeneity in riparian systems. In
Ecosystem function in heterogeneous landscapes (pp. 279–309). Springer.
Naiman, R. J., & Decamps, H. (1997). The ecology of interfaces: Riparian zones. Annual Review
of Ecology and Systematics, 28(1), 621–658.

53

Naiman, R. J., Decamps, H., & McClain, M. E. (2005). Riparia: Ecology, conservation, and
management of streamside communities. Elsevier.
Naiman, R. J., Fetherston, K. L., McKay, S. J., & Chen, J. (1998). Riparian forests. River
Ecology and Management: Lessons from the Pacific Coastal Ecoregion, 289–323.
Nassauer, J. I. (1997). Cultural sustainability: Aligning aesthetics and ecology. Island Press.
National Park Service. (2015). Bandelier National Monument Foundation Document. National
Park Service, Department of the Interior.
O’Brien, G. R., Wheaton, J., Fryirs, K., McHugh, P., Bouwes, N., Brierley, G., & Jordan, C.
(2017). A geomorphic assessment to inform strategic stream restoration planning in the
Middle Fork John Day Watershed, Oregon, USA. Journal of Maps, 13(2), 369–381.
Odum, E. P., Finn, J. T., & Franz, E. H. (1979). Perturbation theory and the subsidy-stress
gradient. Bioscience, 29(6), 349–352.
Oliver, C. D., Larson, B. C., & Oliver, C. D. (1996). Forest stand dynamics. Wiley New York.
Parisa, Z., & Nova, M. (2020). This AI can see the forest and the trees. IEEE Spectrum, 57(8),
32–37. https://doi.org/10.1109/MSPEC.2020.9150554
Quinn, J. M., Brown, P. M., Boyce, W., Mackay, S., Taylor, A., & Fenton, T. (2001). Riparian
Zone Classification for Management of Stream Water Quality and Ecosystem Health1.
JAWRA Journal of the American Water Resources Association, 37(6), 1509–1515.
https://doi.org/10.1111/j.1752-1688.2001.tb03656.x
Radeloff, V. C., Hammer, R. B., Stewart, S. I., Fried, J. S., Holcomb, S. S., & McKeefry, J. F.
(2005). The wildland–urban interface in the United States. Ecological Applications,
15(3), 799–805.

54

Raffa, K. F., Aukema, B. H., Bentz, B. J., Carroll, A. L., Hicke, J. A., Turner, M. G., & Romme,
W. H. (2008). Cross-scale drivers of natural disturbances prone to anthropogenic
amplification: The dynamics of bark beetle eruptions. Bioscience, 58(6), 501–517.
Rapport, D. J., & Whitford, W. G. (1999). How ecosystems respond to stress: Common
properties of arid and aquatic systems. BioScience, 49(3), 193–203.
R Core Team (2020). R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. URL https://www.R-project.org/
Rehfeldt, G. E., Ferguson, D. E., & Crookston, N. L. (2009). Aspen, climate, and sudden decline
in western USA. Forest Ecology and Management, 258(11), 2353–2364.
Reiners, W. A. (2005). Reciprocal cause and effect between environmental heterogeneity and
transport processes. In Ecosystem function in heterogeneous landscapes (pp. 67–89).
Springer.
Richardson, D. M., Holmes, P. M., Esler, K. J., Galatowitsch, S. M., Stromberg, J. C., Kirkman,
S. P., Pyšek, P., & Hobbs, R. J. (2007). Riparian vegetation: Degradation, alien plant
invasions, and restoration prospects. Diversity and Distributions, 13(1), 126–139.
https://doi.org/10.1111/j.1366-9516.2006.00314.x
Sarr, D. A., Hibbs, D. E., & Huston, M. A. (2005). A hierarchical perspective of plant diversity.
The Quarterly Review of Biology, 80(2), 187–212.
Sedell, J. R., Reeves, G. H., Hauer, F. R., Stanford, J. A., & Hawkins, C. P. (1990). Role of
refugia in recovery from disturbances: Modern fragmented and disconnected river
systems. Environmental Management, 14(5), 711–724.

55

Simard, M., Romme, W. H., Griffin, J. M., & Turner, M. G. (2011). Do mountain pine beetle
outbreaks change the probability of active crown fire in lodgepole pine forests?
Ecological Monographs, 81(1), 3–24.
Southwood, T. R. E. (1988). Tactics, strategies and templets. Oikos, 3–18.
Sparks, R. E., Bayley, P. B., Kohler, S. L., & Osborne, L. L. (1990). Disturbance and recovery of
large floodplain rivers. Environmental Management, 14(5), 699–709.
Stephens, S. L., Agee, J. K., Fule, P. Z., North, M. P., Romme, W. H., Swetnam, T. W., &
Turner, M. G. (2013). Managing forests and fire in changing climates. Science,
342(6154), 41–42.
Sun, Q., Miao, C., Hanel, M., Borthwick, A. G. L., Duan, Q., Ji, D., & Li, H. (2019). Global heat
stress on health, wildfires, and agricultural crops under different levels of climate
warming. Environment International, 128, 125–136.
https://doi.org/10.1016/j.envint.2019.04.025
Sweeney, B. W., Bott, T. L., Jackson, J. K., Kaplan, L. A., Newbold, J. D., Standley, L. J.,
Hession, W. C., & Horwitz, R. J. (2004). Riparian deforestation, stream narrowing, and
loss of stream ecosystem services. Proceedings of the National Academy of Sciences,
101(39), 14132–14137.
Swetnam, T. W., & Baisan, C. H. (2003). Tree-ring reconstructions of fire and climate history in
the Sierra Nevada and southwestern United States. In Fire and climatic change in
temperate ecosystems of the western Americas (pp. 158–195). Springer.
Swetnam, T. W., Baisan, C. H., & Kaib, J. M. (2001). Forest fire histories of the sky islands of
La Frontera. Changing Plant Life of La Frontera: Observations on Vegetation in the

56

United States/Mexico Borderlands. University of New Mexico Press, Albuquerque, 95–
119.
Swetnam, T. W., Farella, J., Roos, C. I., Liebmann, M. J., Falk, D. A., & Allen, C. D. (2016).
Multiscale perspectives of fire, climate and humans in western North America and the
Jemez Mountains, USA. Philosophical Transactions of the Royal Society B: Biological
Sciences, 371(1696), 20150168. https://doi.org/10.1098/rstb.2015.0168
Tickner, D. P., Angold, P. G., Gurnell, A. M., & Mountford, J. O. (2001). Riparian plant
invasions: Hydrogeomorphological control and ecological impacts. Progress in Physical
Geography, 25(1), 22–52.
Veenhuis, J. E. (2002). Effects of wildfire on the hydrology of Capulin and Rito de los Frijoles
Canyons, Bandelier National Monument, New Mexico. U.S. Department of the Interior,
U.S. Geological Survey.
Westerling, A. L., Hidalgo, H. G., Cayan, D. R., & Swetnam, T. W. (2006). Warming and earlier
spring increase western US forest wildfire activity. Science, 313(5789), 940–943.
Whitman, W., & Whitman, M. W. (1947). The Pueblo Indians of San Ildefonso: A changing
culture. Columbia University Press.
Williams, A. P., Allen, C. D., Macalady, A. K., Griffin, D., Woodhouse, C. A., Meko, D. M.,
Swetnam, T. W., Rauscher, S. A., Seager, R., & Grissino-Mayer, H. D. (2013).
Temperature as a potent driver of regional forest drought stress and tree mortality. Nature
Climate Change, 3(3), 292.
Williams, J. W., & Jackson, S. T. (2007). Novel climates, no-analog communities, and
ecological surprises. Frontiers in Ecology and the Environment, 5(9), 475–482.

57

Wilson, C. D. (2013). The gradual development of systems of pottery production and distribution
across Northern Rio Grande landscapes. From Mountaintop to Valley Bottom.
Worrall, J. J., Rehfeldt, G. E., Hamann, A., Hogg, E. H., Marchetti, S. B., Michaelian, M., &
Gray, L. K. (2013). Recent declines of Populus tremuloides in North America linked to
climate. Forest Ecology and Management, 299, 35–51.
Zybach, B., Dubrasich, M., Brenner, G., & Marker, J. (2009). US wildfire cost-plus-loss
economics project: The "one-pager "checklist. Advances in Fire Practices’,
Fall.(Wildland Fire Lessons Learned Center) Available at Https://Www. Wildfirelessons.
Net/HigherLogic/System/DownloadDocumentFile. Ashx.

58

Appendix
This appendix includes data that was excluded from the general analysis. Wetted ground,
shadow, and no data cells were excluded from analysis, as well as the lowermost reaches of
Frijoles Canyon, which were anomalous due to the large proportions of bare ground in the delta,
and deep shadows up to km 1.5 of the river. Tables 6-9 exhibit the raw cover proportions of
Frijoles and Capulin canyons. Figures 10-11 present the proportions of all cover types, in Frijoles
and Capulin canyons, by year.

Table A1. Frijoles Canyon raw cover proportions without channel designated.

Cover type

Year1992 Year2011 Year2014 Year2018

Forest

60%

56%

13%

14%

Other

14%

17%

34%

58%

Bare

8%

9%

29%

10%

Wetted

2%

0%

0%

0%

Shadow

17%

18%

25%

18%

Unidentifiable

1%

0%

0%

0%

59

Figure A1. Proportions of vegetative cover observed in Frijoles Canyon overall by year

Table A2. Counts of sample points by cover type, disturbance segment, and year for Frijoles
Canyon. Data represented graphically as proportions in Figure 5.

Year
1992

2011

2014

2018

Disturbance
segment
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D

Bare
10
25
1
18
26
19
4
13
21
65
46
71
8
23
8
18

Bare
channel
1
3
0
0
1
2
0
0
14
89
200
2
2
15
40
1

Forest
209
254
419
346
205
313
364
269
144
50
0
66
154
35
13
96

Other
vegetation
18
19
17
191
39
65
8
218
68
145
226
246
94
269
430
392

60

Table A3. Capulin Canyon raw proportions without channel designated.

Cover Type

Year1992 Year2011 Year2014 Year2018

Forest

47%

26%

6%

3%

Other

26%

47%

56%

71%

Bare

9%

17%

30%

21%

Wetted

0%

0%

0%

0%

Shadow

17%

11%

8%

5%

Unidentifiable

1%

0%

0%

0%

61

Figure A2. Proportions of vegetative cover observed in Capulin Canyon overall by year.
Table A4. Counts of sample points by cover type, disturbance segment, and year for Capulin
Canyon. Data represented graphically as proportions in Figure 9.

Year
1992

2011

2014

2018

Disturbance
segment
E
F
G
H
E
F
G
H
E
F
G
H
E
F
G
H

Bare
66
1
5
33
126
67
9
30
47
42
22
51
40
21
9
19

Bare
channel
104
2
1
2
128
14
3
0
194
189
62
66
154
137
16
8

Forest
8
377
176
332
12
221
99
158
15
5
1
5
26
14
3
9

Other
vegetation
196
20
7
239
119
177
60
477
117
245
122
575
143
320
190
706

62