ECOHYDROLOGICAL CONDITIONS ASSOCIATED WITH THE DISTRIBUTION AND PHENOLOGY OF THE PIMA PINEAPPLE CACTUS (Coryphantha scheeri var. robustispina)
by
Amí L. Kidder
______
A Thesis Submitted to the Faculty of the
SCHOOL OF NATURAL RESOURCES AND THE ENVIRONMENT
In Partial Fulfillment of the Requirements
For the Degree of
MASTER OF SCIENCE
In the Graduate College
THE UNIVERSITY OF ARIZONA
2015
1 | P a g e
STATEMENT BY AUTHOR
This thesis has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.
Brief quotations from this thesis are allowable without special permission, provided that an accurate acknowledgement of the source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate
College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.
SIGNED: Amí Lynne Kidder
APPROVAL BY THESIS DIRECTOR
This thesis has been approved on the date shown below:
______01/13/2015 ______Shirley Anne Papuga Date Professor, Watershed Management and Ecohydrology
______01/13/2015______David D. Breshears Date Professor, Watershed Management and Ecohydrology
2 | P a g e
ACKNOWLEDGEMENTS
This research was possible due to project funding provided by the U.S. Air Force
Life Cycle Management Center environmental compliance team supporting Air Force
Plant 44. Additional support was provided by Raytheon Missile Systems Environmental,
Health, Safety, and Sustainability Department, as well as NSF Critical Zone Observatory
(NSF-EAR-1331408).
I would like to thank my advisors Dr. Shirley Papuga and Dr. David Breshears, both of whom guided me through this project and were patient while I learned. I am also thankful for my committee members Dr. Mitchel McClaran and Dr. Darin Law who each provided me with unique guidance and were always willing to share their insights with me. I would also like to thank Dr. D. Phillip Guertin and Dr. Prakash Dhakal for their help when I needed to expand my programming and lab skills. Additional thanks to
Katherine Belzowski and Toni Flora for their review and advice on the final manuscript.
I owe special thanks to the members of the Papuga and Breshears labs for all their help through the years in learning new skills and discussing new ideas: Cianna, Jason,
Juan, Mallory, Zulia, MaPilar, Zack, Bhaskar, Jessica, Daphne, Adam, Daniel, Matt,
Rachel, and Vanessa.
I would also like to thank the Raytheon Environmental and Industrial Hygiene team for all their support through this project: Heather, Kate, David, Dianna, Katie,
Richard, and Christine.
3 | P a g e
DEDICATION
This thesis is dedicated to the many people who have helped me along the way, and without whom I would not be who I am today. To my parents and siblings who love me: I am ever thankful for your support, encouragement, and Jeep loans. To Tom, who supported me through the rough patches and is the best field-companion I could ask for: here’s to finding C. scheeri where it’s supposed to be and to bringing Irish Folk music to the desert. And finally, to really good hiking boots: I could not have endured my many crossings of the “Land of Cholla” without you.
4 | P a g e
Table of Contents List of Figures ...... 6 List of Tables ...... 7 Abstract ...... 8 Chapter 1: Context ...... 10 1.1 Climate Change ...... 10 1.2 Precipitation Pulse Dynamics ...... 11 1.3 Phenology ...... 13 1.4 Endangered Species ...... 13 1.5 Pima Pineapple Cactus ...... 14 1.6 Thesis Format ...... 17 Chapter 2: Present Study ...... 19 2.1 Summary of Results ...... 20 2.2 Summary of Conclusions ...... 21 2.3 Future Research Opportunities...... 22 2.3.1 Soil moisture as an indicator of spatial distribution ...... 22 2.3.2 Shifting phenology as an additional threat to endangered species ...... 23 2.3.3 Shifting phenology and resource distribution in alternate ecosystem types ...... 23 References ...... 25 APPENDX A: CLASSIFICATION TREE ESTIMATION OF THE CLIMATE CHANGE VELOCITY GAP FOR LITTLE-STUDIED, SLOW-DISPERSING ENDANGERED SPECIES ...... 34 APPENDIX B: ECOHYDROLOGICAL CONTROLS ON CACTUS FLOWERING PHENOLOGY AND PLANT REPRODUCTIVE CHARACTERISTICS: A TWO YEAR OBSERVATIONAL STUDY OF AN ENDANGERED SPECIES ...... 72 APPENDIX C: SUPPLEMENTAL DATA FROM SANTA RITA EXPERIMENTAL RANGE ON PLANT PHYISCAL CHARACTERISTICS AND ASSOCIATED ABIOTIC CHARACTERISTICS ...... 106
5 | P a g e
List of Figures
Figure 1: Map of Coryphantha scheeri var. robustispina distribution, reproduced from:
Baker, Marc, 2004 USGS Report. Within this report, Coryphantha scheeri var. robustispina is named Coryphantha robustispina var. robustispina…………………….14
Figure 2: Photo of C. scheeri flowering…………………………………………………15
6 | P a g e
List of Tables
Table 1: Journal publications and theses on C. scheeri and other species in the genus; also included non-comprehensively are some relevant reports……………………………….33
7 | P a g e
Abstract
Climate changes in temperature and precipitation are already occurring and are projected to further exhibit increasing temperature and precipitation extremes and increasing variation. Such increased temperature variation and decreased precipitation are likely to have a profound impact on vegetation communities, particularly in regions that are dominated by extreme temperatures and strongly seasonal precipitation events. Both temperature and precipitation are tightly linked to vegetation growth and distribution, and in regions such as the U.S. desert southwest, there are a number of rare and endangered species that have a particularly tight knit relationship with their environment. Here, I examine the relationship between these ecohydrological drivers and a specific, little- researched cactus: the Pima Pineapple Cactus (Coryphantha scheeri var. robustispina).
C. scheeri is a small, hemispherical cactus that resides in the Santa Cruz and Altar
Valleys of Southern Arizona, and very little is known about the conditions that promote
C. scheeri distribution and growth. To provide information that may aide in managing this species, I investigate aspects of the distribution and the phenology of this species.
With respect to distribution, I hypothesize that (H1) C. scheeri locations are associated with spatial physical and climatic data within its geographic limits. A framework describing the climatic associations of C. scheeri would enable species managers to take advantage of suitable habitat when opportunities arise. With respect to phenology, within established C. scheeri habitat we lack a clear understanding of the impact ecohydrological factors can have on reproduction and size. Therefore, I also hypothesize
(H2) that C. scheeri flowering phenology is triggered by available moisture, which may be in the form of precipitation, humidity, or soil moisture. My results indicate that
8 | P a g e through the use of the classification tree, C. scheeri habitat is strongly associated with climatic and physical variables at a state-wide scale; these associations indicate large losses of suitable habitat under future projected climate scenarios. Additionally, I find that C. scheeri flowering phenology appears to be associated with precipitation and the resulting increase of soil moisture; the data are also suggestive that bud formation might be associated with water-year growing degree day. Because the results indicate a tight coupling with climatic variables, with most suitable habitat within the current range in
Arizona projected to be lost under future climate, I suggest managers may be inclined to increase monitoring C. scheeri in an ecohydrological context relative to the variables identified here and to consider conditions and locations where supplemental watering or microclimate amelioration could be beneficial for the species.
Key words: Classification tree, endangered species, phenology, spatial distribution
9 | P a g e
Chapter 1: Context
1.1 Climate Change
Changes in climate are already underway and projected to increase in the future
(Garfin et al., 2013; IPCC, 2013). Projected changes include atmospheric composition
(Isaksen et al., 2009), temperature (Crowley, 2000; Hansen et al., 2006; Karl et al., 1991;
Meehl et al., 2000), precipitation (Christensen & Christensen, 2002; Gorman &
Schneider, 2009; Meehl et al., 2000), and species composition and distribution (Hansen et al., 2001). Increases in greenhouse gas concentrations are primarily evident in the changes observed in temperature (Crowley, 2000; Hansen et al., 2006; Karl et al., 1991) and precipitation (Christensen & Christensen, 2002; Gorman & Schneider, 2009).
Projected temperature changes include increasingly frequent and extreme heat and cold events (Mearns et al., 1984; Meehl et al., 2000). Additionally, average temperature ranges are observed to be increasing toward the poles, and are projected to continue to do so
(Seager et al., 2007; Walther et al., 2002). Precipitation changes are projected to exhibit more extreme flooding and drought events (Meehl et al., 2000), altering the hydrologic cycle and moisture availability globally.
These changes in temperature and precipitation are impacting species distributions (IPCC, 2014; Kotwicki & Lauth, 2013; Stevenson & Lauth, 2012) and the timing of life cycle events—phenology (Kotwicki & Lauth, 2013; Ma et al., 2013;
Stevenson & Lauth, 2012). The magnitude of response to these changes varies across species and ecosystems, leading to a variety of challenging consequences for natural resource managers. Additionally, climate conditions are projected to occur so quickly, even under the least extreme scenarios, that biome shifts are predicted to out-pace the
10 | P a g e species that exist therein (Sandel et al., 2011; Schloss et al., 2012). Species will require dispersal and establishment rates to track climate-driven changes in suitable habitat if they are to offset such losses (IPCC 2014).
Projected climate changes are expected to be particularly pronounced in the
Southwestern U.S. (Garfin et al., 2013). Both hotter and drier conditions are projected for the Southwestern U.S. and associated deserts (Albright et al., 2010; Feng et al., 2014).
Additionally, greater variability in temperature and precipitation is projected (Albright et al., 2010; Feng et al., 2014), as well a greater occurrence of extreme events including heat waves and drought (Kharin et al., 2007; Welbergen et al., 2008). These climate changes are expected to impact not only individual species, but also ecosystems, with respect to factors such as nutrient availability and fire disturbance regimes (Hurteau et al., 2014;
Nitschke & Innes, 2007). Consequently, improved regional planning and management of natural resources is needed to address this challenge.
1.2 Precipitation Pulse Dynamics
Of particular importance to Southwestern ecosystems and other drylands worldwide is the spatiotemporal variability in available water resources (Cable &
Huxman, 2004; Huxman et al., 2004a; Huxman et al., 2004b), because in these locations water is the dominant limiting factor for much of the year (Reynolds et al., 2004;
Rosenstock et al., 2005). The seasonal distribution of precipitation is important in the
Southwestern U.S., and varies along a gradient from the Mojave to Sonoran to
Chihuahuan deserts (Brutsaert, 2012; Cañón et al., 2011). The Sonoran desert, intermediate between the Mojave and Chihuahuan, is characterized by precipitation limited to a heavy monsoon season, variable winter precipitation, and long dry periods in
11 | P a g e between (Cable & Huxman, 2004). Much of biological activity in the Sonoran desert and other dryland ecosystems has been characterized as pulsed responses to precipitation
“pulses” (Cable & Huxman, 2004; Huxman et al., 2004a; Huxman et al., 2004b;
Schwinning et al., 2004; Schwinning & Sala, 2004).
Diverse biological responses to precipitation pulses span microbial responses, biological soil crusts, and vegetation physiology, growth, phenology and distribution. For example, microbial responses to precipitation pulses in biological soil crusts exhibit carbon ecosystem exchange and nutrient cycling pulses (Huxman et al., 2004b).
Precipitation pulses trigger responses even when they are small in magnitude. For instance, biological soil crusts contribute to both carbon and nitrogen cycling (Belnap,
2006; Cable & Huxman, 2004) as a result of precipitation pulses that are too small to activate shallow plant roots (Cable & Huxman, 2004). Biological soil crust response to precipitation pulses also enables small pulses to infiltrate down into the shallow root later
(Belnap, 2006). Similarly, vegetative responses to precipitation pulses are reflected in observed patterns in distribution (Schmidt et al., 2013), transpiration (Huxman et al.,
2004b; Nagler et al., 2007; Yepez et al., 2005), and phenological events such as growth
(Wu et al., 2009) and frequency and duration of flowering events (Crimmins et al., 2014,
2011, 2013).
In addition to the amount of precipitation occurring in an ecosystem and the specific pulse dynamics of that precipitation, the seasonal distribution of the precipitation in drylands is important (DeBano et al., 1995). For example, shifts in the timing of precipitation can alter timing of phenological events (Lesica & Kittelson, 2010; Neil et al., 2010). Consequently, shifts in the seasonal patterns of precipitation pulses in
12 | P a g e combination with other climatic variables (e.g. growing degree day) may impose challenges for biota.
1.3 Phenology
Phenology examines the timing and magnitude of life-cycle events (Rathcke &
Lacey, 1985) and how they are triggered (Gunarathne & Perera, 2014; Ovaskainen et al.,
2013; Peñuelas et al., 2004). In the water-limited Southwestern U.S., phenological events are often triggered by precipitation pulses (Angert et al., 2010; Miranda et al., 2014,
2009). Where precipitation pulses are seasonal, so then are the phenological responses
(Crimmins et al., 2014, 2013), which are proportionally greater with the size of the precipitation pulse (Schwinning & Sala, 2004). The timing of phenological events is often critical to the reproductive strategy of a species to reduce competition or take advantage of prime reproductive conditions (Clements et al., 2010; Kaspari et al., 2001).
Understanding the phenology of dominant species is an important characterization of landscapes (Kurc & Benton, 2010), and is particularly important for providing insight into risks for threatened or endangered species, as well as identifying opportunities for natural resource managers to take action (Neveu, 2009; Schiller et al., 2000; Ye et al.,
2006).
1.4 Endangered Species
Projected shifts in temperature and precipitation and associated shifts in suitable habitat range can disproportionately affect threatened and endangered species (Hannah et al., 2002; Sodhi et al., 2004; Wernberg et al., 2011). These species may already be more susceptible to fluctuations in climatic conditions because their populations are typically smaller, and may be less resilient to typical variations in climate and environment that
13 | P a g e induce fluctuation in healthy and robust species populations (Betini et al., 2014; Morita et al., 2014). There are numerous threatened and endangered species in the Southwestern
U.S. (Dobson et al., 1997). Species in this region are often inherently sensitive to precipitation and temperature changes, and therefore may be particularly impacted by projected climate changes (Erickson & Waring, 2014).
1.5 Pima Pineapple Cactus
Of particular concern to natural resource managers are potential climate change impacts on endangered species. An additional challenge is that many endangered species are little studied, so key information on basic biological characteristics such as dispersal rates are lacking.
Vulnerability of an endangered species to climate change can be particularly great if dispersal rates Figure 1: Map of Coryphantha scheeri var. robustispina distribution, reproduced from: Baker, Marc, 2004 USGS Report. Within this report, Coryphantha scheeri var. robustispina is are slow, thereby limiting a species’ named Coryphantha robustispina var. robustispina. ability to track changes in in climate if suitable habitat is projected to be lost.
An example of a little-studied and likely slow-dispersing endangered species is the Pima Pineapple Cactus (Coryphantha scheeri var. robustispina, here C. scheeri) of the Sonoran desert. This species has been federally protected since 1993 (Roller, 1996;
Schmalzel et al., 2004; U.S. Fish and Wildlife Service, 2007) and within the U.S. is located only in a limited region of the Southwest (Roller, 1996; Schmalzel et al., 2004;
14 | P a g e
U.S. Depertment of Interior, 2001; U.S. Fish and Wildlife Service, 2007). Clusters of
Coryphantha scheeri subspecies are found within limited areas of Arizona, New Mexico, and Texas, and small regions within Chihuahua and Sonora, Mexico (Figure 1; Baker,
2004; McDonald & McPherson, 2005; Roller,
1996; Schmalzel et al., 2004). There are only a few
published journal articles and theses on this species
and others in the genus, in addition to a limited
number of additional relevant reports (Table 1);
most of the studies focus on only a single aspect of
the species. C. scheeri are small (growing up to 46
cm), hemispherical cacti, with nodes referred to as
Figure 2: Photo of C. scheeri flowering. “tubercles” rather than the typical “ribs” characteristic of many cactus species in the southwestern U.S. (McDonald, 2007; Roller,
1996; Sayre, 2005; Schmalzel et al., 2004; U.S. Fish and Wildlife Service, 2007). Other closely related species are often mistaken for C. scheeri var. robustispina but a phenetic analysis indicated it is distinct from variations valida, uncinata, and scheeri (Baker,
2004). Another characteristic of the C. scheeri var. robustispina is that the pups that grow around the base of the parent, produced asexually, and are speculated to serve as an opportunity to continue reproducing if the parent dies (McDonald, 2007; Roller, 1996). A
C. scheeri adult can have as many as 100 pups in a cluster, though very rare, but typically
have fewer than 10 (McDonald, 2007).
Current observations of C. scheeri within the U.S. are limited to small regions in the Southwest between 700 and 1,500 m on slopes of 10% or less (McDonald &
15 | P a g e
McPherson, 2005; Roller, 1996; Schmalzel et al., 2004; U.S. Fish and Wildlife Service,
2007). Distribution is sparse, with existing literature indicating only a single individual per ~10 ha (McDonald, 2007; McDonald & McPherson, 2005; Sayre, 2005), though they can be found in much denser clusters (up to 8 plants per ~0.4 ha) (U.S. Depertment of
Interior, 2001; U.S. Fish and Wildlife Service, 2007). Habitat where C. scheeri is found is characterized as gently sloping alluvial fans within desert scrubland and desert grassland
(McDonald, 2007; McDonald & McPherson, 2005; Roller, 1996; U.S. Fish and Wildlife
Service, 2007).
C. scheeri var. robustispina exhibits varied growth rates, declining in the fall and winter months and in the summer monsoon season, and increasing during the spring
(Roller, 1996). C. scheeri flowers following rain events and can produce multiple flowers in each flowering event (McDonald & McPherson, 2005; Roller, 1996; U.S. Depertment of Interior, 2001). Flower abortion can occur, and can increase in frequency through the flowering season for unknown reasons (McDonald & McPherson, 2005). Each C. scheeri flowers multiple times each year, producing large (6 - 10 cm in diameter) flowers that range in color from almost white to bright, vivid yellow (McDonald, 2007; Roller, 1996).
C. scheeri tends to flower synchronously across similar sites, and when successfully pollinated will form a fruit approximately 7.5 cm long that turns from dark to pale green when ripe (McDonald, 2007; Roller, 1996). The species is an obligate outcrosser with a strong relationship with its primary pollinator, the Diadasia rinconis, a medium sized bee whose preferred habitat coincides with C. scheeri habitat (McDonald, 2007; Roller,
1996). Although less well documented, other bee species and ants also visit its flowers
(McDonald & McPherson, 2005). Because C. scheeri flower after many other desert
16 | P a g e cactus, they offer pollen to Diadasia later in the year and are likely an important resource to provide pollen to juvenile Diadasia (McDonald, 2007). Other potential competitors for pollinator attention include prickly pear (Opuntia) and barrel cactus (Ferocactus), which tend to flower in the spring and a few weeks after the monsoon season begins, respectively (McDonald & McPherson, 2005). Distribution of seeds from developed fruit is not well documented, though small mammals (e.g. rabbits, rodents) are thought to fill this ecological role (Roller, 1996).
One of the many challenges facing C. scheeri is its proximity to the U.S./Mexico border, and the resulting human impacts from border patrol and illegal immigration activities (Marris, 2006). Much C. scheeri habitat is located within range of heavy traffic from immigration and drug trafficking, which also increases the challenge in studying the species safely (Marris, 2006). Another challenge to the success of C. scheeri is the introduction of Lehmann Lovegrass (Eragrostis lehmanniana), which can subject C. scheeri to increased resource competition and risk from fire damage (Roller, 1996; Sayre,
2005; U.S. Depertment of Interior, 2001; U.S. Fish and Wildlife Service, 2007).
Additional challenges include illegal collection, grazing, habitat fragmentation and loss from anthropogenic activities such as mining, and development (U.S. Depertment of
Interior, 2001; U.S. Fish and Wildlife Service, 2007).
1.6 Thesis Format
In the following thesis, I present a summary of the key results and conclusions of my research on ecohydrological associations for the distribution and phenology of C. scheeri var. robustispina, followed by discussion of possibilities for future work.
Following the summary are two appendices which present manuscripts of papers
17 | P a g e intended for publication, followed by a third appendix with supplemental data. The manuscripts present the data collected and analyzed in response to hypotheses on ecohydrological controls on C. scheeri var. robustispina distribution and on phenology.
18 | P a g e
Chapter 2: Present Study
In 2013, in effort to better understand how to manage the endangered species on their property, the United States Air Force (hereafter referred to as Air Force) provided support to study a relatively dense population of C. scheeri var. robustispina at a location in southern Arizona. This endangered species has little information on many of its key biological characteristics. In addition to providing an opportunity to study this rarely examined endangered species, the Air Force study enables the study of C. scheeri var. robustispina at a site with controlled public access, potentially minimizing the effects of some factors that may be influencing the cactus associated land disturbances such as grazing and recreational pedestrian traffic. To advance our understanding of factors influencing the distribution and phenology of C. scheeri, I take a two-part ecohydrological approach that considers climate drivers and soil influences, first focusing on larger scale patterns of distribution and second focusing on specific phenological responses.
At the larger scale considering distribution, a fundamental conservation challenge is assessing potential climate change impacts and management options for endemic, specialized, and rare species (including threatened and endangered species) (Caldow et al., 2007), many of which have been little studied and for which key information on dispersal rates are lacking. Of particularly concern are species for which dispersal rates are expected (even if not robustly documented) to be slow relative to “climate velocity”, or the speed at which a species must track climate change to remain within suitable habitat (Loarie et al., 2009; Sandel et al., 2011; Schloss et al., 2012). Given the need for maximal insights based on limited available information for little-studied and likely slow-
19 | P a g e dispersing species, classification trees can be especially useful because they offer potential for high predictive ability without a priori assumptions about functional relationships between environmental variables used and species suitability (Bhattacharya,
2013; Pal & Mather, 2003). Here I develop classification trees for the little-studied endangered desert cactus species, Coryphantha scheeri var. robustispina. I examine how climatic and physical factors contribute to C. scheeri distribution across the landscape.
My hypothesis (H1) is that C. scheeri locations are associated with spatial physical and climatic data within its geographic limits. Classification trees are developed both with and without elevation, and with and without seasonal precipitation totals; these trees are then used to predict current distribution and to project future distribution under climate change scenarios. Based on functional type and limited literature, C. scheeri is likely a relatively slow dispersing species and therefore may be particularly sensitive to changes in suitable habitat associated with projected climate change.
At a smaller scale, I examine the climatic conditions associated with two key phenological responses: budding and flowering. My hypothesis (H2) is that C. scheeri flowering phenology is triggered by available moisture, which may be in the form of soil moisture, humidity, or rainfall. I examine the physical site characteristics that coincide with C. scheeri habitat including soil texture, infiltration rates, and incoming solar radiation. I examine parameters associated with C. scheeri phenology, considering attributes of precipitation, temperature, and soil moisture.
2.1 Summary of Results
For the distribution study, the classification trees correctly estimated either predicted absence or predicted presence for > 95% of present distribution cases and were
20 | P a g e most influenced by maximum temperature and average annual precipitation; average summer precipitation was also a strong influence when included. Using climate projections for 2050, a classification tree with elevation estimated > 99% loss of suitable habitat within its current Arizona distribution and only a few locations of additional suitable habitat in Arizona outside of current Arizona distribution. However, without an elevation constraint, extensive additional suitable habitat in Arizona was projected outside the current Arizona distribution. When seasonal precipitation patterns were included (and elevation was not included), the classification tree estimated only very limited additions in future suitable habitat elsewhere in Arizona outside the current
Arizona distribution. The data limitations (e.g. spatial resolution) associated with these estimates should be considered in application of the results, and field surveys will likely be required for site-specific determinations.
For the phenology study, flowering events are generally associated with precipitation events > 10 mm and a corresponding increase in soil moisture. Flowering events occur multiple times within a single monsoon season within 6 days of precipitation pulses. Soil texture type for study locations were mostly either sandy loam or sandy clay loam, with larger sized individuals associated with lower sand content.
Higher infiltration rates are associated with smaller cacti and lower node count. Direct site factor (DSF) values, a measure of the amount of near-ground incoming solar radiation over a cloudless year, were not associated with metrics of size.
2.2 Summary of Conclusions
Collectively my results indicate that (1) C. scheeri distribution is tightly associated with climate conditions, (2) flowers can occur multiple times per year,
21 | P a g e apparently in response to precipitation pulses, and (3) based on the developed classification trees, this species is estimated to be at risk of large loss of suitable Arizona habitat under future (2050) climate. Additionally, the classification trees estimate some new suitable habitat in other parts of Arizona under projected future climate but this habitat is mostly outside the species’ current range. Assuming that this species is a slow disperser, the changes in suitable habitat could necessitate assisted migration for this species to track climate change; however, any such assisted migration would have to consider numerous complexities, including pollinator movement.
With respect to research approaches, my study highlights that classification tree results for little-studied but likely slow-dispersing rare species can aide in assessing species vulnerability, identifying locations within current range where climate velocity might be tracked, and mapping candidate locations for assisted migration. My study also highlights how phenological monitoring using time-series data can reveal important aspects of species biology that are relevant for natural resource manager consideration.
2.3 Future Research Opportunities
Based on the results presented here and considering the relevant literature, I identify three key possibilities for further research on C. scheeri in the context of its surrounding environment and the phenology of the desert Southwest.
2.3.1 Soil moisture as an indicator of spatial distribution
Results from this study indicate the importance of soil moisture for C. scheeri phenology, but soil moisture data is quite limited, especially relative to C. scheeri locations. Work is currently underway to enable wide-scale, remotely-sensed spatial data for soil moisture, which would enable the inclusion of soil moisture in a distribution
22 | P a g e analysis. Because moisture is such a limiting resource in arid and semi-arid environments, the inclusion of spatially explicit soil moisture data could strengthen a future C. scheeri analysis, as well as more fully enable the examination of other species distribution in relation to soil moisture as a resource.
2.3.2 Shifting phenology as an additional threat to endangered species
Existing literature on C. scheeri from the 1990’s and early 2000’s indicates that bud formation occurs in mid-May and flowering begins in June in response to monsoon events in June and July (Roller, 1996, McDonald and McPherson, 2005). In this study, phenological responses occurred later: bud formation was observed nearly 4 weeks later in mid-June, and flowering events responded to monsoon events in July. A change of this magnitude in the timing of phenology could potentially affect the ability of C. scheeri to attract pollinators away from competitors and to attract fruit distributors if those species do not exhibit similar changes in phenology. Because climate change can affect many species across the landscape differently, phenological shifts could occur in many species that desynchronize them from available resources and alter the ecological functional relationships across the ecosystem. To better understand how these shifts may take shape, a more comprehensive study on recorded shifts in desert phenology across taxa and functional groups could be useful.
2.3.3 Shifting phenology and resource distribution in alternate ecosystem types
Such shifts in phenology and their consequences across the ecosystem in terms of functional group interactions and nutrient redistribution are a topic that I wish to further pursue my studies as a PhD student. I have gained much knowledge in the duration of this project that is unrelated to the specifics of C. scheeri. This research experience has
23 | P a g e given me valuable experience in designing a research project with limited means in order to best evaluate a set of hypotheses. Related to this, I’ve gained experience in examining data for a single species, and based on existing knowledge of natural processes, understanding how the data relates to the broader ecosystem. Although field data and tools used within the course of this project may be limited in scope, practice in being methodical in data collections, observations, and field campaigns are useful skills across disciplines. Specific methods that are cross-functional include the programming and GIS skills I’ve established in the course of this project which are valuable tools in many disciplines. It is with these skills that I hope to ask questions of phenological shifts and the resulting effects on resource availability in other ecosystem types, such as aquatic freshwater and marine ecosystems.
24 | P a g e
References
Albright, T. P., Pidgeon, A. M., Rittenhouse, C. D., Clayton, M. K., Wardlow, B. D., Flather, C. H., Culbert, P. D., Radeloff, V. C. 2010. Combined effects of heat waves and droughts on avian communities across the conterminous United States. Ecosphere, 1(5), 1-11. doi:10.1890/ES10-00057.1
Angert, A. L., Horst, J. L., Huxman, T. E., & Venable, D. L. 2010. Phenotypic plasticity and precipitation response in Sonoran Desert winter annuals. American Journal of Botany, 97(3), 405–11. doi:10.3732/ajb.0900242
Baker, M. 2004. Phenetic analysis of Coryphantha, section Robustispina (cactaceae), part 1: stem characters.U.S. Fish and Wildlife Service report (pp. 1–21). Tempe, Arizona.
Belnap, J. 2006. The potential roles of biological soil crusts in dryland hydrologic cycles. Hydrological Processes, 20(15), 3159–3178. doi:10.1002/hyp.6325
Betini, G. S., Griswold, C. K., Prodan, L., & Norris, D. R. 2014. Body size, carry-over effects and survival in a seasonal environment: consequences for population dynamics. The Journal of Animal Ecology, 83, 1313–1321. doi:10.1111/1365- 2656.12225
Bhattacharya, M. 2013. Bioclimatic modelling: a machine learning perspective. International Journal of Advanced Computer Science and Applications, 4(2), 1–8.
Brutsaert, W. 2012. Are the North American deserts expanding? Some climate signals from groundwater storage conditions. Ecohydrology, 5(5), 541–549. doi:10.1002/eco.263
Cable, J. M., & Huxman, T. E. 2004. Precipitation pulse size effects on Sonoran Desert soil microbial crusts. Oecologia, 141(2), 317–24. doi:10.1007/s00442-003-1461-7
Caldow, R. W. G., Stillman, R. A, le V. dit Durell, S. E. A., West, A. D., McGrorty, S., Goss-Custard, J. D., Wood, P. J., Humphreys, J. 2007. Benefits to shorebirds from invasion of a non-native shellfish. Proceedings of the Royal Society Biological Sciences, 274, 1449–1455. doi:10.1098/rspb.2007.0072
Cañón, J., Domínguez, F., & Valdes, J. B. 2011. Vegetation responses to precipitation and temperature: a spatiotemporal analysis of ecoregions in the Colorado River Basin. International Journal of Remote Sensing, 32(20), 5665–5687. doi:10.1080/01431161.2010.507259
Christensen, J. H., & Christensen, O. B. 2002. Severe summertime flooding in Europe. Nature, 421, 2002–2003.
25 | P a g e
Clements, M. N., Clutton-Brock, T. H., Albon, S. D., Pemberton, J. M., & Kruuk, L. E. B. 2010. Getting the timing right: antler growth phenology and sexual selection in a wild red deer population. Oecologia, 164(2), 357–68. doi:10.1007/s00442-010- 1656-7
Crimmins, T. M., Bertelsen, D. C., & Crimmins, M. a. 2014. Within-season flowering interruptions are common in the water-limited Sky Islands. International Journal of Biometeorology, 58(4), 419–26. doi:10.1007/s00484-013-0745-9
Crimmins, T. M., Crimmins, M. A., & Bertelsen, C. D. 2011. Onset of summer flowering in a “Sky Island” is driven by monsoon moisture. The New Phytologist, 191(2), 468–79. doi:10.1111/j.1469-8137.2011.03705.x
Crimmins, T. M., Crimmins, M. A., & Bertelsen, C. D. 2013. Spring and summer patterns in flowering onset, duration, and constancy across a water-limited gradient. American Journal of Botany, 100(6), 1137–47. doi:10.3732/ajb.1200633
Crowley, T. J. 2000. Causes of climate change over the past 1000 years. Science, 289(5477), 270–277. doi:10.1126/science.289.5477.270
DeBano, L. F., Ffolliott, P. F., Ortega-Rubio, A., Gottfried, G. J., Hamre, R. H., & Edminster, C. B. 1995. Biodiversity and management of the Madrean Archipelago: The sky islands of Southwestern Untied States and Northwestern Mexico. In Biodiversity and Management of the Madrean Archipelago (p. 669).
Dobson, A. P., Rodriguez, J. P., Roberts, W. M., & Wilcove, D. S. 1997. Geographic distribution of endangered species in the United States. Science, 275, 16–19.
Erickson, C. C., & Waring, K. M. 2014. Old Pinus ponderosa growth responses to restoration treatments, climate and drought in a southwestern US landscape. Applied Vegetation Science, 17(1), 97–108. doi:10.1111/avsc.12056
Feng, W., Hardt, B. F., Banner, J. L., Meyer, K. J., James, E. W., Musgrove, M., Edwards, R. L., Cheng, H., & Min, A. 2014. Changing amounts and sources of moisture in the U.S. southwest since the Last Glacial Maximum in response to global climate change. Earth and Planetary Science Letters, 401, 47–56. doi:10.1016/j.epsl.2014.05.046
Garfin, G., Jardine, A., Merideth, R., Black, M., & LeRoy, S. 2013. Assessment of climate change in the Southwest United States. (G. Garfin, A. Jardine, R. Merideth, M. Black, & S. LeRoy, Eds.). Washington, DC: Island Press/Center for Resource Economics. doi:10.5822/978-1-61091-484-0
Gorman, P. A. O., & Schneider, T. 2009. The physical basis for increases in precipitation extremes in simulations of 21st-century climate change. Proceedings of the National Academy of Science, 106(35), 14773–14777.
26 | P a g e
Gunarathne, R. M. U. K., & Perera, G. A. D. 2014. Climatic factors responsible for triggering phenological events in Manilkara hexandra ( Roxb .) Dubard., a canopy tree in tropical semi-deciduous forest of Sri Lanka. Tropical Ecology, 55(1), 63–73.
Hannah, L., Midgley, G. F., Lovejoy, T., Bond, W. J., Bush, M., Lovett, J. C., Scott, D., & Woodward, F. I. 2002. Conservation of biodiversity in a changing climate. Conservation Biology, 16(1), 264–268. doi:10.1046/j.1523-1739.2002.00465.x
Hansen, A. J., Neilson, R. P., Dale, V. H., Flather, C. H., Iverson, L. R., Currie, D. J., Shafer, S., Cook, R., & Bartlein, P. J. 2001. Global change in forests: responses of species, communities, and biomes. BioScience, 51(9), 765–779.
Hansen, J., Sato, M., Ruedy, R., Lo, K., Lea, D. W., & Medina-Elizade, M. 2006. Global temperature change. Proceedings of the National Academy of Sciences of the United States of America, 103(39), 14288–93. doi:10.1073/pnas.0606291103
Hurteau, M. D., Bradford, J. B., Fulé, P. Z., Taylor, A. H., & Martin, K. L. 2014. Climate change, fire management, and ecological services in the southwestern US. Forest Ecology and Management, 327, 280–289. doi:10.1016/j.foreco.2013.08.007
Huxman, T. E., Cable, J. M., Ignace, D. D., Eilts, J. A., English, N. B., Weltzin, J., & Williams, D. G. 2004. Response of net ecosystem gas exchange to a simulated precipitation pulse in a semi-arid grassland: the role of native versus non-native grasses and soil texture. Oecologia, 141(2), 295–305. doi:10.1007/s00442-003- 1389-y
Huxman, T. E., Snyder, K. A, Tissue, D., Leffler, A. J., Ogle, K., Pockman, W. T., Sandquist, D. R., Potts, D. L., & Schwinning, S. 2004. Precipitation pulses and carbon fluxes in semiarid and arid ecosystems. Oecologia, 141(2), 254–68. doi:10.1007/s00442-004-1682-4
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp, doi:10.1017/CBO9781107415324
IPCC, 2014: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1132 pp.
27 | P a g e
Isaksen, I. S. a., Granier, C., Myhre, G., Berntsen, T. K., Dalsøren, S. B., Gauss, M., Klimont, Z., Benestad, R., Bousquet, P., Collins, W., Cox, T., Eyring, V., Fowler, D., Fuzzi, S., Jockel, P., Laj, P., Lohmann, U., Maione, M., Monks, P., Prevot, A. S. H., Raes, F., Richter, A., Rognerud, B., Schulz, M., Shindell, D., Stevenson, D. S., Storelvmo, T., Wang, W.-C., van Weele, M., & Wuebbles, D. 2009. Atmospheric composition change: climate–chemistry interactions. Atmospheric Environment, 43(33), 5138–5192. doi:10.1016/j.atmosenv.2009.08.003
Karl, R., Change, J., & Ouayle, G. 1991. Global warming: evidence for asymmetric diurnal temperature change. Geophysical Research Letters, 18(12), 2253–2256.
Kaspari, M., Pickering, J., Longino, J., & Windsor, D. 2001. The phenology of a neotropical ant assemblage: evidence for continuous and overlapping reproduction. Behavioral Ecology and Sociobiology, 50(4), 382–390. doi:10.1007/s002650100378
Kharin, V. V., Zwiers, F. W., Zhang, X., & Hegerl, G. C. 2007. Changes in temperature and precipitation extremes in the ipcc ensemble of global coupled model simulations. Journal of Climate, 20(8), 1419–1444. doi:10.1175/JCLI4066.1
Kotwicki, S., & Lauth, R. R. 2013. Detecting temporal trends and environmentally-driven changes in the spatial distribution of bottom fishes and crabs on the eastern Bering Sea shelf. Deep Sea Research Part II: Topical Studies in Oceanography, 94, 231– 243. doi:10.1016/j.dsr2.2013.03.017
Kurc, S. A., & Benton, L. M. 2010. Digital image-derived greenness links deep soil moisture to carbon uptake in a creosotebush-dominated shrubland. Journal of Arid Environments, 74(5), 585–594. doi:10.1016/j.jaridenv.2009.10.003
Lesica, P., & Kittelson, P. M. 2010. Precipitation and temperature are associated with advanced flowering phenology in a semi-arid grassland. Journal of Arid Environments, 74(9), 1013–1017. doi:10.1016/j.jaridenv.2010.02.002
Loarie, S. R., Duffy, P. B., Hamilton, H., Asner, G. P., Field, C. B., & Ackerly, D. D. 2009. The velocity of climate change. Nature, 462(7276), 1052–5. doi:10.1038/nature08649
Ma, T., Zhou, C., Pei, T., & Xie, Y. 2013. A comparative analysis of changes in the phasing of temperature and satellite-derived greenness at northern latitudes. Journal of Geographical Sciences, 23(1), 57–66. doi:10.1007/s11442-013-0993-y
Marris, E. 2006. Wildlife caught in crossfire of US immigration battle. Nature, 442(July).
McDonald, C. J. 2007, April. Pima pineapple cactus: a unique cactus hiding in plain sight. The Plant Press, 1–4.
28 | P a g e
McDonald, C., & McPherson, G. 2005. Pollination of pima pineapple cactus (Coryphantha scheeri var. robustispina): does pollen flow limit abundance of this endangered species. USDA Forest Service Proceedings, 529–532.
Mearns, L., Katz, R. W., & Schneider, S. H. 1984. Extreme high-temperature events: changes in their probabilities with changes in mean temperature. American Meteorological Society, 23, 1601–1613.
Meehl, G. A., Zwiers, F., Evans, J., Knutson, T., Mearns, L., & Whetton, P. 2000. Trends in extreme weather and climate events : issues related to modeling extremes in projections of future climate change*. Bulletin of the American Meteorological Society, 81(3), 427–436.
Miranda, J. de D., Jorquera, M. J., & Pugnaire, F. I. 2014. Phenological and reproductive responses of a semiarid shrub to pulsed watering. Plant Ecology, 215(7), 769–777. doi:10.1007/s11258-014-0354-7
Miranda, J. de D., Padilla, F. M., Lazaro, R., & Pugnaire, F. I. 2009. Do changes in rainfall patterns affect semiarid annual plant communities? Journal of Vegetation Sciencce, 20, 269–276.
Morita, K., Tamate, T., Kuroki, M., & Nagasawa, T. 2014. Temperature-dependent variation in alternative migratory tactics and its implications for fitness and population dynamics in a salmonid fish. The Journal of Animal Ecology, 83, 1268– 1278. doi:10.1111/1365-2656.12240
Nagler, P. L., Glenn, E. P., Kim, H., Emmerich, W., Scott, R. L., Huxman, T. E., & Huete, A. R. 2007. Relationship between evapotranspiration and precipitation pulses in a semiarid rangeland estimated by moisture flux towers and MODIS vegetation indices. Journal of Arid Environments, 70(3), 443–462. doi:10.1016/j.jaridenv.2006.12.026
Neil, K. L., Landrum, L., & Wu, J. 2010. Effects of urbanization on flowering phenology in the metropolitan phoenix region of USA: findings from herbarium records. Journal of Arid Environments, 74(4), 440–444. doi:10.1016/j.jaridenv.2009.10.010
Neveu, A. 2009. Incidence of climate on common frog breeding: long-term and short- term changes. Acta Oecologica, 35(5), 671–678. doi:10.1016/j.actao.2009.06.012
Nitschke, C. R., & Innes, J. L. 2007. Climatic change and fire potential in South-Central British Columbia, Canada. Global Change Biology, 14(4), 841–855. doi:10.1111/j.1365-2486.2007.01517.x
Ovaskainen, O., Skorokhodova, S., Yakovleva, M., Sukhov, A., Kutenkov, A., Kutenkova, N., Shcherbakov, A., Meyke, E., & Delgado, M. del M. 2013. Community-level phenological response to climate change. Proceedings of the
29 | P a g e
National Academy of Sciences of the United States of America, 110(33), 13434– 13439. doi:10.1073/pnas.1305533110/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1305533110
Pal, M., & Mather, P. M. 2003. An assessment of the effectiveness of decision tree methods for land cover classification. Remote Sensing of Environment, 86(4), 554– 565. doi:10.1016/S0034-4257(03)00132-9
Peñuelas, J., Filella, I., Zhang, X., Llorens, L., Ogaya, R., Lloret, F., Comas, P., Estiarte, M., & Terradas, J. 2004. Complex spatiotemporal phenological shifts as a response to rainfall changes. New Phytologist, 161, 837–846. doi:10.1111/j.1469- 8137.2003.01003.x
Rathcke, B., & Lacey, E. P. 1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics, 16(1), 179–214. doi:10.1146/annurev.es.16.110185.001143
Reynolds, J. F., Kemp, P. R., Ogle, K., & Fernández, R. J. 2004. Modifying the “pulse- reserve” paradigm for deserts of North America: precipitation pulses, soil water, and plant responses. Oecologia, 141(2), 194–210. doi:10.1007/s00442-004-1524-4
Roller, P. 1996. Distribution, growth, and reproduction of Pima pineapple cactus (Coryphantha scheeri Kuntz var robustispina Schott). University of Arizona.
Rosenstock, S. S., Bleich, V. C., Rabe, M. J., & Reggiardo, C. 2005. Water quality at wildlife water sources in the Sonoran Desert, United States. Rangeland Ecology Management, 58(6), 623–627.
Sandel, B., Arge, L., Dalsgaard, B., Davies, R. G., Gaston, K. J., Sutherland, W. J., & Svenning, J.-C. 2011. The influence of late quaternary climate-change velocity on species endemism. Science, 334(6056), 660–4. doi:10.1126/science.1210173
Sayre, N. F. 2005. Interacting effects of landownership, land use, and endangered species on conservation of Southwestern U. S. Rangelands. Conservation Biology, 19(3), 783–792.
Schiller, J. R., Zedler, P. H., & Black, C. H. 2000. The effect of density-dependent insect visits, flowering phenology, and plant size on seed set of the endangered vernal pool plant Pogogyne abramsii (Lamiaceae) in natural compared to created vernal pools. Wetlands, 20(2), 386–396. doi:10.1672/0277- 5212(2000)020[0386:TEODDI]2.0.CO;2
Schloss, C. A., Nuñez, T. A., & Lawler, J. J. 2012. Dispersal will limit ability of mammals to track climate change in the Western Hemisphere. Proceedings of the National Academy of Sciences of the United States of America, 109(22), 8606–
30 | P a g e
8611. doi:10.1073/pnas.1116791109/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1116791109
Schmalzel, R. J., Nixon, R. T., Best, A. L., & Jr, J. A. T. 2004. Morphometric Variation in Coryphantha robustispina (Cactaceae). Systematic Botany, 29(3), 553–568.
Schmidt, M., Traoré, S., Ouédraogo, A., Mbayngone, E., Ouédraogo, O., Zizka, A., Kirchmair, I., Kabore, E., Thiombiano, A., Hahn, K., & Zizka, G. 2013. Geographical patterns of woody plants ’ functional traits in Burkina Faso. Candollea, 68(2), 197–207.
Schwinning, S., & Sala, O. E. 2004. Hierarchy of responses to resource pulses in arid and semi-arid ecosystems. Oecologia, 141(2), 211–20. doi:10.1007/s00442-004-1520-8
Schwinning, S., Sala, O. E., Loik, M. E., & Ehleringer, J. R. 2004. Thresholds, memory, and seasonality: understanding pulse dynamics in arid/semi-arid ecosystems. Oecologia, 141(2), 191–3. doi:10.1007/s00442-004-1683-3
Seager, R., Ting, M., Held, I., Kushnir, Y., Lu, J., Vecchi, G., Huang, H.-P., Hamik, N., Leetmaa, A., Lau, N.-C., Li, C., Velez, J., & Naik, N. 2007. Model projections of an imminent transition to a more arid climate in southwestern North America. Science, 316(5828), 1181–4. doi:10.1126/science.1139601
Sodhi, N. S., Koh, L. P., Brook, B. W., & Ng, P. K. L. 2004. Southeast Asian biodiversity: an impending disaster. Trends in Ecology & Evolution, 19(12), 654– 60. doi:10.1016/j.tree.2004.09.006
Stevenson, D. E., & Lauth, R. R. 2012. Latitudinal trends and temporal shifts in the catch composition of bottom trawls conducted on the eastern Bering Sea shelf. Deep Sea Research Part II: Topical Studies in Oceanography, 65-70, 251–259. doi:10.1016/j.dsr2.2012.02.021
U.S. Depertment of Interior. 2001. United States Department of the Interior, U. S. Fish and Wildlife Service AESO/SE 2-21-01-F-271 (pp. 1–16). Phoenix, Arizona.
U.S. Fish and Wildlife Service. 2007. U . S . Fish & Wildlife Service 5-Year Review Determination Pima Pineapple Cactus. 70 FR 5460. Region 2 Regional Office (pp. 1–17).
Walther, G., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T. J. C., Fromentin, J.-M., hoegh-Guldberg, O., & Bairlein, F. 2002. Ecological responses to recent climate change. Nature, 416, 389–395.
Welbergen, J. A, Klose, S. M., Markus, N., & Eby, P. 2008. Climate change and the effects of temperature extremes on Australian flying-foxes. Proceedings of the Royal Society Biological Sciences 275, 419–425. doi:10.1098/rspb.2007.1385
31 | P a g e
Wernberg, T., Russell, B. D., Moore, P. J., Ling, S. D., Smale, D. A., Campbell, A., Coleman, M., Steinberg, P. D., kendrick, G. A., & Connell, S. D. 2011. Impacts of climate change in a global hotspot for temperate marine biodiversity and ocean warming. Journal of Experimental Marine Biology and Ecology, 400(1-2), 7–16. doi:10.1016/j.jembe.2011.02.021
Wu, J. B., Xiao, X. M., Guan, D. X., Shi, T. T., Jin, C. J., & Han, S. J. 2009. Estimation of the gross primary production of an old‐ growth temperate mixed forest using eddy covariance and remote sensing. International Journal of Remote Sensing, 30(2), 463–479. doi:10.1080/01431160802372143
Ye, Q.-G., Yao, X.-H., Zhang, S.-J., Kang, M., & Huang, H.-W. 2006. Potential risk of hybridization in ex situ collections of two endangered species of Sinojackia Hu (Styracaceae). Journal of Integrative Plant Biology, 48(7), 867–872. doi:10.1111/j.1744-7909.2006.00260.x
Yepez, E. A., Huxman, T. E., Ignace, D. D., English, N. B., Weltzin, J. F., Castellanos, A. E., & Williams, D. G. 2005. Dynamics of transpiration and evaporation following a moisture pulse in semiarid grassland: A chamber-based isotope method for partitioning flux components. Agricultural and Forest Meteorology, 132(3-4), 359– 376. doi:10.1016/j.agrformet.2005.09.006
32 | P a g e
Yearsley, 2004 2005 Thomas, 2011 Martorell, Portilla-Alonso and Nobel, 1981 Valverde, 2008 & Martinez-Berdja al,Guadalupe et 2005 al, et Mandujano 2002 EmmaMarris, Sayre, 2005 2001 Consultation, US DOI Fox & 20006 Nino-Murcia, 2007 McDonald, Schmalzel al, et 2004 USFWS review, 5-year 2007 2005 & McPherson, McDonald al, et Phillips 1981 Roller, 1996 1991 Mills, Baker, 2004 Author
comprehensivelyare some relevantreports. Table1: Journal publications and theses on scheeri C. and other species in genus; the also incl structure assumption. structure rate and segregate remove model that need age removal components corrects for species for the into and that age-a of the component structure dynamics, long-term. as compared to expand of model population Results include scope growth transient to to short-term near-term sensitivity examining for study species and effectiveness dynamic robustness modelling of transient the as study robbinsorum presented "Coryphantha on density. decreaseobserved to observed spp. breeding some previously restore though of the re-establishment to appeared population, vivipara of resiliance included "Coryphantha in fire study to damage. have C. vivipara found to 3.7x was higher rate and fire mortality damage disturbance-tolerant. relate chronic to disturbance, positivley seems whereas to growth seems negatively related, indicating C. werdermannii that is relatively werdermannii for effects of chronic studied "Coryphantha locations. anthropogenic Establishment and fecundity disturbance different at study cellular to prior affectindividuals. uptake damage. CO2 to appear Cold night temps vivipara for freezing studied below -15C kill nocturnal effects. Found "Coryphantha temps below some, -20C while kill temps all observed different levels All to included (if species of a soil any) nurse plant). moisture, of nosolar to strong response radiation. responded werdermannii between of cactusrelationship included"Coryphantha growth, soil moisture, solar rediation in contributions noted study (as the and determinediversity where conservation actions will effective. be most radians, salinensis, sulcata, and wohlschlageri of various evaluate locations to species included of study inventory species in represented leuchtenbergia,"Coryphantha georgii, glanduligera, guerkeana, macro robust. var. not macromeris, statistically found, but Relationship vaupeliana, pallida nurse on association includedplants. ocatcantha, with pulleineana,"Coryphantha in study delicata, nickelsiae, is additional^There danger in field Southernillegal within due Arizona to studies imigrant crossings and drug trafficking. andand activitiesfederal other on state lands more challenging. grazing lands^Private and temporary lands with exclusions are easier after fire; restore to threatened and endangered make species restoration developer to report dynamics within and population known threats, habitat, ^Overview of PPC increases demand the for cconservation banking. threatened and endangered to ^Impacts drivers is one of conservation strongest species of the banking and strong enforcement of regulations *Undergone 9 name changes var @ time (robus robus of article), extensive of description physical * over Schmalzel al, et 2004 based on comments of reviewers. *Determination * time) the and information (at habitat, known on *Extensiveflowering as well of typical description, as description physical and reproduction, *Extensive of description physical includes and on causethreats(e.g. speculations fire,surveys, of low density weevles, loss) habitat of *Description variations, considers of taxonomic robustispina *Analysis between coryphantha relationships Focus(* C. scheeri C. scheeri C.scheeri species distinctions are questionable, does not think distinctions between subspecies are are between distinctions subspecies significant. species questionable, distinctions think does not more be likely limited pollinated ifpollinator to near not another, but C. scheeri C. scheeri is focus, ^ is should remain listed as endangered, thorough overview of existing Baker, delineation literature.2004 species Accepts survey, description of habitat, physical description, and notes on specific populations through various biological on specific and populations notes description, physical of habitat, description survey, C.scheeri
C. scheeri
.
is mentioned, is "Genus related only) , germination and growth rates, flowering process, C. scheeri var. robustispinaC. scheeri var. C. scheeri
. udednon
separate. -
33 | P a g e
APPENDX A: CLASSIFICATION TREE ESTIMATION OF THE CLIMATE
CHANGE VELOCITY GAP FOR LITTLE-STUDIED, SLOW-DISPERSING
ENDANGERED SPECIES
Amí L. Kidder1,2, Shirley A. Papuga1, David D. Breshears1,3 and Darin J. Law1
Targeted for publication in Conservation Biology
1School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721 USA 3Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ 85721 USA 1Present address: School of Natural Resources and the Environment, University of Arizona, Tucson, AZ [email protected]
34 | P a g e
Abstract
A fundamental conservation challenge is assessing potential climate change impacts and management options for endemic, specialized, and rare species (including threatened and endangered species), many of which have been little studied and for which key information on dispersal rates is lacking. Of particularly concern are species for which dispersal rates may be undocumented but are expected to be slow relative to
“climate velocity”, or speed at which a species must track climate change to remain within suitable habitat. Given the need for maximal insights based on limited available information for little-studied and likely slow-dispersing species, classification trees can be especially useful because they offer potential for high predictive ability without a priori assuming functional relationships between environmental variables used and species suitability. Here we develop classification trees for a little-studied endangered desert cactus species, Pima Pineapple Cactus (Coryphantha scheeri var. robustispina) that, based on limited literature and functional type, is likely a relatively slow-dispersing species. The classification trees correctly estimated either predicted absence or predicted presence for >95% of present distribution cases and were most influenced by maximum temperature and average precipitation. Using 2050 climate projections, a classification tree with elevation estimated > 99% loss of current suitable habitat with little suitable habitat added. Classification tree results for little-studied but likely slow-dispersing rare species can highlight species vulnerability, identify locations within current range where climate velocity might be tracked, and map candidate locations for assisted migration.
Key Words: Coryphantha scheeri var. robustispina, environmental modeling
35 | P a g e
1.0 Introduction
A fundamental conservation challenge is assessing potential climate change impacts and management options for endemic, specialized, and rare species (including threatened and endangered species), many of which have been little studied and for which key information on reproduction and dispersal rates are often lacking. Particularly important is the potential spatiotemporal lag between the rate of climate change and the reproductive and dispersal rates of species, described as a “velocity gap” between current and projected suitable habitat (Loarie et al., 2009; Sandel et al., 2011; Schloss et al.,
2012, IPCC, 2014). Such velocity gaps are further affected by other factors that influence species movement rates (Robinet & Roques, 2010; Schloss et al., 2012), such as barriers related to development and urban areas, agricultural lands, and elevational barriers, in addition to biotic barriers such as those related to competition or pollination (Cassidy &
Grue, 2000; Fischer & Lindenmayer, 2005; Harrison & Rasplus, 2006; Walther et al.,
2009). The uncertainty surrounding the ability of particular plant species and associated vegetative communities to mobilize in response to their current climate creates a significant challenge for natural resource managers, particularly in assessing risks for rare species (Caldow et al., 2007). Rare species often have population sizes too small to remain resilient to environmental stressors and as such, are suffering disproportionate loss with increased effects of climate change (Bro-Jørgensen, 2014; Cahill et al., 2013;
Convertino & Valverde, 2013).
Among the more vulnerable are species for which dispersal rates are expected to be slow relative to the velocity gap (Fordham et al., 2012; Hiddink et al., 2012, IPCC,
2014). Many endemic, specialized, and rare species (including threatened and endangered
36 | P a g e species) have been little studied and lack key information on dispersal rates. However, limited information associated with a given rare species, along with related literature such as that for the genus or functional type, can provide preliminary indications that a species may be slow dispersing. Information on population dynamics and dispersal requirements are often lacking in the literature due to the rare nature inherent to endangered species, with the exception of the charismatic and economically important endangered species that tend to be more fully researched (Bried & Mazzacano, 2010; Clucas et al., 2008;
Cove et al., 2013; Di Minin et al., 2013). In addition to challenges associated with locating rare individuals, experiments and data collection possibilities are often limited by the need for non-destructive methods (Ellender et al., 2012; Metcalf & Swearer,
2005). Additionally, access to existing data may be regulated in effort to protect the species from interference.
Nonetheless managers need to obtain timely assessments of potential climate impacts and management options despite data limitations. Expected ecosystem changes potentially influencing rare species include expansion of invasive species (Pyke et al.,
2008; Rahel & Olden, 2008; Van Klinken et al., 2009), species density changes (Bertrand et al., 2011; Rubidge et al., 2011; Thaxter et al., 2010), and species distribution shifts
(Chen et al., 2011; Parmesan & Yohe, 2003). In effort to understand the magnitude of these changes, managers often turn to species distribution modeling, also referred to as
“environmental niche modeling” (Elith & Leathwick, 2009; Segurado & Araujo, 2004) using an approach also referred to as “Machine Learning” (Bhattacharya, 2013).
Although there is no universally best approach to species distribution modelling (Austin,
37 | P a g e
2007; Bhattacharya, 2013), there are a variety of tools are used for predictive modeling, each with specific advantages and disadvantages.
Notable tools include the generalized linear model (GLM), the generalized additive model (GAM), the Genetic Algorithm for Rule Set Production (GARP), classification and regression trees (CART), artificial neural networks (ANN), and a variety of specialized interpolation and niche analysis tools typically designed to accommodate multi-faceted climate space (Anderson et al., 2003; Austin, 2007;
Bhattacharya, 2013; Elith & Leathwick, 2009; Segurado & Araujo, 2004; Stockwell &
Peterson, 2002). With the availability of so many different techniques, it is important to examine (1) the available data and (2) the desired outcome to select the best suited model for the species distribution analysis (Austin, 2002; Austin, 2007). One tool, the classification tree, has specific strengths which are particularly useful in predicting potential range for an under-informed species. Primarily, classification trees make no a priori assumptions between the variables used and the species’ preference, or frequency of species positive relationship to any particular variable (Bhattacharya, 2013; Pal &
Mather, 2003). They also allow for a predictive model to be created with only a few variables while maintaining predictive accuracy (Aitken et al., 2007). This gives classification trees additional flexibility when assessing variables that may have a non- linear relationship with the species (Pal & Mather, 2003). Classification tree results are such that step-wise rules are given to enable ease of decision making (Hannöver &
Kordy, 2005; Pal & Mather, 2003; Segurado & Araujo, 2004; Yeh et al., 2009).
Weaknesses of the classification tree include over-fitness, which in species distribution models may translate to an over-abundance of predicted suitable habitat for the given
38 | P a g e species and variables (Van Ewijk et al., 2014); appropriate spatial scales for analysis and sufficient number of verified presence locations relative to randomly determined locations increase classification tree effectiveness (Aitken et al., 2007). Additionally, since there is no assumption regarding the relationship between variables, the resulting classification tree can give no indication of the potential probability of arriving at any given branch of the tree (Haizhou & Chong, 2009). Despite these limitations, given the need for maximal insights based on limited available information for little-studied and likely slow-dispersing species, classification trees can be especially useful because they offer potential for high predictive ability without assuming a priori functional relationships between environmental variables used and species suitability.
Here we illustrate the functionality and usefulness of environmental modelling with classification trees via an example endangered desert species: the Pima Pineapple
Cactus (Coryphantha scheeri var. robustispina), hereafter referred to as C. scheeri, located in the southwestern U.S. and northern Mexico. We focus on the Arizona portion of the distribution using available data sets. Based on limited literature and functional type reviewed below, C. scheeri is likely a relatively slow-dispersing species, making identification of projected changes in potential suitable habitat particularly useful.
Management of this desert species requires consideration of extreme heat in the summer months in addition to great variability in rainfall patterns between localized heavy summer monsoons and broad-scale winter rainfall. Because there is no clear understanding of the climate envelope C. scheeri exists within, land managers are left guessing at the best way to manage C. scheeri habitat. As projected climate change scenarios take effect, a better understanding of the climate envelope for C. scheeri could
39 | P a g e aid in the management of protected lands and surrounding resources to promote the success of this and other threatened and endangered species.
Highlighting the more general problem where little information is available on the parameters that contribute to the distribution of a rare species, we use classification trees to explore associations between climate and environmental (elevation, soil texture) variables with C. scheeri distribution in both current and projected future climate conditions. More specifically, our objectives are to: (1) identify the primary factors associated with C. scheeri distribution, and (2) project how suitable C. scheeri habitat may shift in response to climate change. More generally, we highlight the utility of classification trees in determining the spatiotemporal gap between climate change and species distribution changes, particularly for little-studied and likely slow-dispersing species.
2. Methods
2.1 Study Species.
C. scheeri is a small, hemispherical species characterized by their tubercles, which differs from the typical ribs of most cactus species in the southwestern U.S., as well as by their starburst spine formation that includes a single spine protruding directly out from the top of the node (Roller, 1996; Schmalzel et al., 2004; U.S. Fish and Wildlife
Service, 2007). C. scheeri are located in the Santa Cruz and Altar Valleys in southern
Arizona (Baker, 2004; McDonald & McPherson, 2005; Roller, 1996; Schmalzel et al.,
2004) and are federally protected within the United States; their distribution does extend into northern Mexico (Baker, 2004). C. scheeri is small and difficult to detect, especially in grassy areas, and as a result, has not been widely surveyed across its range (U.S. Fish
40 | P a g e and Wildlife Service, 2007). Previous research provides some description of the physical landscape characteristics that C. scheeri is found in, such as on slopes <10% (U.S. Fish and Wildlife Service, 2007), but there is very little information published and available to the general public on the effects of climatic factors such as temperature and precipitation
(Table 1). The species depends on a specific pollinator bee species (Diadasia), and a trace study of pollination patterns suggested pollination may not be limiting, but average pollination distance was limited and less than the pollination area visited by individual
Diadasia (1 km; McDonald & McPherson, 2005). Little other data regarding reproduction and dispersal rates exist, but in general cacti as a functional group are expected to disperse relatively slowly (Butler et al., 2012; Gibbs et al., 2008; Godínez-
Alvarez & Valiente-Banuet, 2004); cacti species may have even slower dispersal rates than trees, which were the slowest functional group relative to velocity gaps in a recent synthesis (IPCC, 2014).
2.2 Spatial Data Sets
We conducted a spatial analysis with geographic information systems (GIS) location data of the C. scheeri, along with shapefiles of environmental variables, to quantify the association of these factors with the species’ distribution. The Arizona
Heritage Data Management System (AHDMS) group maintains spatial data for both invasive and endangered species across Arizona as part of the state’s land management and planning. The spatial data was first submitted to Arizona Game and Fish Department through various land surveys completed by environmental consultants, development agencies, land management agencies, and others as guided by the endangered species management program in the state, and was then managed by AHDMS. Spatial location
41 | P a g e data for all known C. scheeri in Arizona (6905 individuals, alive and dead) was provided by AHDMS for this study. This spatial data on the distribution of C. scheeri within
Arizona was combined with spatial data for each environmental factor to identify parameters associated with C. scheeri.
Spatial data for climate and environmental data used for this analysis was collected from a variety of sources and loaded into ArcMap 10.2. We identified six environmental factors for use in our analysis: (1) elevation (Fig. 1a), (2) soil texture type
(Fig. 1b), (3) precipitation (Fig. 1c), (4) maximum annual temperature (Fig. 1d), (5) minimum annual temperature (Fig. 1e), and (6) annual temperature range (Fig. 1f).
Elevation data (Elev) for the state of Arizona was obtained from the U.S. Geological
Survey (USGS) National Map Viewer (http://viewer.nationalmap.gov/viewer/) at a resolution of one-third arcsecond. The map was highlighted to select Arizona as the desired range within the National Map Viewer, which was then sent from USGS as 62 shapefiles. These files were imported into ArcMap and combined into a single raster; the extraneous data outside of Arizona was clipped and removed from the raster.
Information on soil texture was obtained from the U.S.D.A Natural Resources
Conservation Service SSURGO (http://sdmdataaccess.nrcs.usda.gov/), which is a digital soil survey database at a 1:24,000 resolution. The data were downloaded as a package containing several different soil characteristics for the state of Arizona. The SSURGO polygon shapefile was imported into ArcMap, and the polygon file was joined using the map unit key field (MUKEY) to join the polygons with the soil texture data, identifying the soil texture type for each polygon.
42 | P a g e
The PRISM climate group (http://www.prism.oregonstate.edu/) maintains a dataset of 30-year normal averages for precipitation (ppt) and maximum and minimum temperature (Tmax and Tmin, respectively), each at a resolution of 800m. The shapefiles were downloaded for each of these parameters and imported into ArcMap. Each file contained data for the entirety of the United States and was clipped down to the extent of
Arizona. Using the raster calculator, Tmin was subtracted from Tmax to identify Trange as a
6th factor to evaluate as a contributor to C. scheeri distribution. In addition, the effect of seasonal ppt was examined by using the raster calculator in ArcMap to sum monthly ppt data into six-month totals; winter ppt consisted of October – March data (Fig. 1g) and summer ppt consisted of April – August data (Fig. 1h).
Climate projection spatial data was obtained from WorldClim.org, a collection of different data layers available for ecological and modeling research (IPCC, 2013). For this analysis, projection data for the year 2050 via the CCSM4 model at 30 seconds resolution was used due to the robustness of the model and its inclusion of factors that directly impact the southwestern U.S., such as the El Nino Southern Oscillation (Gent et al., 2011). Projections for Tmax, Tmin, and ppt were downloaded for the four available climate change scenarios: R26, R45, R60, and R85. Each of these scenarios refer to a representative concentration pathway (RCP), which each respectively indicate an increasing level of climate change severity. R26 projections assume the maximum abatement efforts of greenhouse gases (GHG) leading to peak levels of GHG in between
2010-2020 and decreasing after; R45 and R60 each indicate longer durations of GHG concentration increases (approximately 2040 and 2060, respectively); and R85 assumes
GHG levels increase through the end of the century (IPCC, 2007). The monthly average
43 | P a g e data from each RCP scenario were averaged to create an annual average, spatially projected in GCS North American 1983 coordinate system, and clipped to the same extent as the current data projections.
2.3 Classification Tree Validation/Verification
Classification trees created in MATLAB R2014a were then developed to estimate predicted presence or predicted absence of C. scheeri under specified environmental conditions. Spatial data for all six environmental factors (Tmax, Tmin, Trange, ppt, Elev, and soil texture) were imported as shapefiles into ArcMap along with C. scheeri location data. All layers originated in GCS North American 1983 except soil texture, which was transformed for consistency. In addition to verified presence data points specific to C. scheeri location, we also selected 7000 randomly-generated location points within the boundaries of Arizona for our analysis. Using a spatial analyst tool in ArcMap, the values for the six predictors for each C. scheeri verified presence point and random location point were extracted and exported into data tables.
Using a split sample design, we trained a classification tree in MATLAB with a data set that combined 75% of points for both C. scheeri verified presence and for random locations. Using the remaining 25% of the points of each type, we tested the predictive ability of the tree (Iverson & Prasad, 1998). Points that the tree predicted presence of C. scheeri (value = TRUE/1) were compared to points with verified presence of C. scheeri, while points that the tree predicted absence of C. scheeri (value =
FALSE/0), were compared to random locations. To validate our method we randomly shuffled the data five times, and each time recreated a classification tree.
44 | P a g e
2.4 Classification Tree Model Prediction
To look at the potential distribution of C. scheeri within Arizona, we rasterized the state into six grids of 695 x 742 pixels each. Using a single classification tree (Fig. 2), we established the predicted presence or predicted absence of C. scheeri for each pixel.
To do this, environmental factors for the entire state of Arizona (see Section 2.4) were exported as rasterized grids from ArcMap and loaded into MATLAB. The grids were reshaped into column vectors, and the values throughout Arizona were evaluated with the classification tree for C. scheeri habitat suitability. The resulting evaluation was then reshaped back to the original grid and, with cell sizes identified, imported into ArcMap displaying the state of Arizona.
Building on projected current suitable C. scheeri habitat, we next evaluated how that spatial distribution might shift under future climate change scenarios. The climate projection data (Tmin, Tmax, Trange, ppt) were also exported from ArcMap as rasterized grids and loaded into MATLAB. Elevation and soil texture type are not expected to change within this time scale, and as such, current spatial data was used to project future
C. scheeri habitat for 2050. Each future projection scenario was evaluated through the classification tree, and a projection of suitable C. scheeri habitat was created for each scenario.
Our initial classification tree included elevation as an important variable, but because elevation is considered within PRISM climate data calculations, it is removed to avoid redundancy. Additionally, in regards to plant distribution, elevation restrictions are due to climate variables, and as such, future climate change might remove elevation constraints associated with the first model (Beckage et al., 2008; Chen et al., 2011).
45 | P a g e
Therefore, to consider future projected suitable habitat not constrained by elevation, we developed another classification tree for current distribution that did not include elevation; we subsequently used this tree to estimate another set of projected changes under future climate. Finally, because seasonal precipitation distribution within the species distribution range has been shown to impact growth and distribution of local species (Cable & Huxman, 2004), we developed a classification tree that included totals for winter and for summer precipitation (in addition to the annual total previously considered; this tree also excluded elevation).
3. Results
The classification tree created as the predictor for C. scheeri distribution analysis was evaluated for accuracy between iterations to review for sampling bias within the creation data subset. All iterations of the classification tree (Table 2) demonstrate high accuracy in predicting both verified presence locations and random locations. In creating the second classification tree without elevation included as a predictor, the same review of sampling bias was conducted, and again, all iterations of the classification tree demonstrate high accuracy (Table 3). The third classification tree including winter and summer precipitation was subject to the same review (Table 4).
The classification tree created in the analysis with elevation (Fig. 2) indicates that
Tmax is the primary predictor in determining suitable C. scheeri habitat, followed by ppt and Elev as secondary predictors. These three predictors remain the dominant factors through the top tiers of the classification tree. Tmin and Trange appear as tertiary predictors with soil texture type as the least common predictor. In the classification tree created for the second analysis (Fig. 3) without Elev as a predictor, Tmax and ppt remain the primary
46 | P a g e and secondary predictors, respectively. Trange and Tmin remain the tertiary predictors, though in this analysis they are contributing factors notably higher on the classification tree. The least common predictor remains soil texture type in this analysis. The classification tree created in the third analysis (Fig. 4) that included winter and summer precipitation indicates that Tmax and summer precipitation are the primary predictors, with winter precipitation and Tmin as secondary predictors; Trange and ppt occur further down on this classification tree.
The projections of suitable C. scheeri habitat under current conditions (Fig. 5) are very similar between the analyses with and without elevation, and both are consistent with the current known distribution of C. scheeri. The projections indicate that current suitable C. scheeri habitat extends north and west of the currently observed C. scheeri habitat at the southern extent of the projections. The projection of suitable habitat in the analysis including seasonal precipitation (Fig. 5c) is consistent with current known distribution, though this tree projects less suitable habitat to the north and west than the first two analyses.
Projections of future C. scheeri habitat differ visibly between analyses with elevation, without elevation, and including seasonal precipitation. In analysis of future climate scenarios R26, R45, R60, and R85 with Elev as a predictor, there appears to be little C. scheeri habitat within the boundaries of Arizona (Fig. 6). In the analysis without
Elev as a variable, each future climate scenario indicates that while little C. scheeri habitat remains within the current range, there are extensive new areas with potential suitable habitat in northern and eastern Arizona (Fig. 7). Future projections including seasonal precipitation display some habitat loss within the current range, with sparse
47 | P a g e suitable habitat extending north and west of the current distribution (Fig. 8). For all analyses, projections were consistent across all climate scenarios, indicating little change between R26, R45, R60, and R85.
4. Discussion
Methodologically, the high accuracy of the classification tree as a model to predict C. scheeri suitable habitat is notable. This type of model typically has good results in habitat projection among species that exhibit a tight relationship with their environment, as highlighted here by a prediction rate over 95% for all iterations of the model for C. scheeri. The high reliability of the classification tree for current distribution increases our confidence in its utility for projections for 2050.
It is important to note that these results are evaluating the conditions of C. scheeri versus the current distribution. Location data for C. scheeri provided by AHDMS is often collected as a result of development and land use activities such that known locations may not indicate the full distribution of the species. Given the available resolutions of the data used here (see section 2.2), projections are sufficient for estimating general suitable
C. scheeri habitat, but where detailed site-specific information is needed, thorough surveys (e.g. using the approach detailed by Roller, 1996) should be completed to confirm presence or absence.
An additional caveat is that the spatial data used for this evaluation are mean values, and are not meant to indicate extreme climate conditions or events. Projections of suitable habitat displayed here are estimates of where C. scheeri may exist; they are not verified locations within the whole of the projected range. Results displayed are indicative of suitable habitat, but these associations with average climate conditions may
48 | P a g e not correspond to specific conditions determining plant mortality, which may be more related to extreme events (McDowell et al., 2008).
From a conservation perspective, the classification trees (with and without elevation, and without or including winter and summer precipitation) all project an enormous loss of suitable habitat within the species current range. This projection reinforces the need for protection, active management, and DNA archival of the species as climate change progresses. Given the severe loss projected, those actions may be insufficient to conserve the species. The classification tree without elevation as a constraint also highlights that large portions of Arizona outside the current C. scheeri range could become suitable habitat. However, given the likely slow dispersal rate of this species and its dependence on a specific pollinator bee species, the likelihood of this species tracking the velocity gap of climate change seems low. Also, these additional areas are not identified as suitable habitat when seasonal precipitation is considered.
Consequently, managers may need to actively intervene through microclimate manipulations that reduce the maximum temperature and add soil moisture to help preserve the species. Additionally or alternatively, managers could evaluate implementing assisted migration to new suitable habitat, but would need to account for pollinator issues, as well as uncertainties associated with several other aspects of the species’ biology.
A further challenge to management includes trans-border issues since the distribution of C. scheeri spans into northern Mexico. We expect that C. scheeri in
Mexico will be subject to the same climatic pressures as the population in southern
Arizona, without the protections assured by U.S. Federal regulations. As is the case for
49 | P a g e all species distributed across international borders, there is a matrix of land management and land use practices that contribute to overall species health and distribution. In regards to conservation efforts, international collaboration is often required from both governmental and independent agencies to maintain robust ecosystems across political boundaries.
More generally, our study illustrates how classification tree results for little- studied but likely slow-dispersing rare species can be useful in a variety of contexts: predicting existing suitable habitat within the current range of a rare species where other individuals are most likely to be located; identifying future potential habitat and associated vulnerabilities, including where climate velocity might be tracked, and mapping candidate locations for consideration in the context of assisted migration.
Approaches such as this are needed to aide conservation managers in making the best decisions possible for rare species in a timely manner despite substantial data limitations on dispersal rates and other aspects of species biology.
Acknowledgements
Support for this project was provided by the Raytheon Company Missile Systems
Environmental, Health, Safety, and Sustainability Department. Additional support was provided by NSF Critical Zone Observatory (NSF-EAR-1331408). We would like to acknowledge and thank the members of the Papuga and Breshears labs at the University of Arizona, School of Natural Resources and the Environment. Additionally, D. P.
Guertin and M. King have provided valuable advice in the approach of this study. A special thanks to J. Crawford with U.S. Fish and Wildlife Service and S. Tonn with the
50 | P a g e
Arizona Game and Fish Department for assistance in obtaining C. scheeri spatial location data.
51 | P a g e
References:
Aitken, M., Roberts D. W., & Shultz, L. M. 2007. Modeling distribution of rare plants in the great basin, western North America. Western North American Naturalist, 67(1), 26-38.
Anderson, R. P., Lew, D., & Peterson, A. T. 2003. Evaluating predictive models of species’ distributions: criteria for selecting optimal models. Ecological Modelling, 162(3), 211–232. doi:10.1016/S0304-3800(02)00349-6
Austin, M. 2007. Species distribution models and ecological theory: A critical assessment and some possible new approaches. Ecological Modelling, 200(1-2), 1–19. doi:10.1016/j.ecolmodel.2006.07.005
Austin, M. 2002. Spatial prediction of species distribution: an interface between ecological theory and statistical modelling. Ecological Modelling, 157(2-3), 101– 118. doi:10.1016/S0304-3800(02)00205-3
Baker, M. 2004. Phenetic analysis of Coryphantha, section Robustispina (cactaceae), part 1: stem characters.U.S. Fish and Wildlife Service report (pp. 1–21). Tempe, Arizona
Beckage, B., Osborne, B., Gavin, D. G., Pucko, C., Siccama, T., & Perkins, T. 2008. A rapid upward shift of a forest ecotone during 40 years of warming in the Green Mountains of Vermont. Proceedings of the National Academy of Sciences of the United States of America, 105(11), 4197–202. doi:10.1073/pnas.0708921105
Bertrand, R., Lenoir, J., Piedallu, C., Riofrío-Dillon, G., de Ruffray, P., Vidal, C., Pierrat, J.-C., & Gégout, J.-C. 2011. Changes in plant community composition lag behind climate warming in lowland forests. Nature, 479(7374), 517–20. doi:10.1038/nature10548
Bhattacharya, M. 2013. Bioclimatic modelling: a machine learning perspective. International Journal of Advanced Computer Science and Applications, 4(2), 1–8.
Bried, J. T., & Mazzacano, C. A. 2010. National review of state wildlife action plans for Odonata species of greatest conservation need. Insect Conservation and Diversity, 3(2), 61–71. doi:10.1111/j.1752-4598.2010.00081.x
Bro-Jørgensen, J. 2014. Will their armaments be their downfall? Large horn size increases extinction risk in bovids. Animal Conservation, 17(1), 80–87. doi:10.1111/acv.12062
52 | P a g e
Butler, C. J., Wheeler, E. A., Stabler, L. B., The, S., & Society, B. 2012. Distribution of the threatened lace hedgehog cactus (Echinocereus reichenbachii) under various climate change scenarios. Journal of the Torrey botanical Society, 139(1), 46–55.
Cable, J. M., & Huxman, T. E. 2004. Precipitation pulse size effects on Sonoran Desert soil microbial crusts. Oecologia, 141(2), 317–24. doi:10.1007/s00442-003-1461-7
Cahill, A. E., Aiello-Lammens, M. E., Fisher-Reid, M. C., Hua, X., Karanewsky, C. J., Ryu, H. Y., Sbeglia, G. C., Spagnolo, F., Waldron, J. B., Warsi, O., & Wiens, J. J. 2013. How does climate change cause extinction? Proceedings. Biological Sciences / The Royal Society, 280(1750), 20121890. doi:10.1098/rspb.2012.1890
Caldow, R. W. G., Stillman, R. A, le V. dit Durell, S. E. A., West, A. D., McGrorty, S., Goss-Custard, J. D., Wood, P. J., Humphreys, J. 2007. Benefits to shorebirds from invasion of a non-native shellfish. Proceedings of the Royal Society Biological Sciences, 274, 1449–1455. doi:10.1098/rspb.2007.0072
Cassidy, K. M., & Grue, C. E. 2000. The role of private and public lands in in conservation of vertebrates State Washington. Wildlife Society Bulletin, 28(4), 1060–1076.
Chen, I.-C., Hill, J. K., Ohlemüller, R., Roy, D. B., & Thomas, C. D. 2011. Rapid range shifts of species associated with high levels of climate warming. Science, 333(6045), 1024–6. doi:10.1126/science.1206432
Clucas, B., McHugh, K., & Caro, T. 2008. Flagship species on covers of US conservation and nature magazines. Biodiversity and Conservation, 17(6), 1517–1528. doi:10.1007/s10531-008-9361-0
Convertino, M., & Valverde, L. J. 2013. Portfolio decision analysis framework for value- focused ecosystem management. PloS One, 8(6), e65056. doi:10.1371/journal.pone.0065056
Cove, M. V., Pardo Vargas, L. E., de la Cruz, J. C., Spínola, R. M., Jackson, V. L., Saénz, J. C., & Chassot, O. 2013. Factors influencing the occurrence of the Endangered Baird’s tapir Tapirus bairdii: potential flagship species for a Costa Rican biological corridor. Oryx, 48(03), 402–409. doi:10.1017/S0030605313000070
Di Minin, E., Fraser, I., Slotow, R., & MacMillan, D. C. 2013. Understanding heterogeneous preference of tourists for big game species: implications for conservation and management. Animal Conservation, 16(3), 249–258. doi:10.1111/j.1469-1795.2012.00595.x
Elith, J., & Leathwick, J. R. 2009. Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology, Evolution, and Systematics, 40(1), 677–697. doi:10.1146/annurev.ecolsys.110308.120159
53 | P a g e
Ellender, B. R., Becker, A., Weyl, O. L. F., & Swartz, E. R. 2012. Underwater video analysis as a non-destructive alternative to electrofishing for sampling imperilled headwater stream fishes. Aquatic Conservation: Marine and Freshwater Ecosystems, 22(1), 58–65. doi:10.1002/aqc.1236
Fischer, J., & Lindenmayer, D. B. 2005. The sensitivity of lizards to elevation: A case study from south-eastern Australia. Diversity and Distributions, 11(3), 225–233. doi:10.1111/j.1366-9516.2005.00139.x
Fordham, D. a, Watts, M. J., Delean, S., Brook, B. W., Heard, L. M. B., & Bull, C. M. 2012. Managed relocation as an adaptation strategy for mitigating climate change threats to the persistence of an endangered lizard. Global Change Biology, 18(9), 2743–55. doi:10.1111/j.1365-2486.2012.02742.x
Gent, P. R., Danabasoglu, G., Donner, L. J., Holland, M. M., Hunke, E. C., Jayne, S. R., Lawrence, D. M., Neale, R. B., Rasch, P. J., Vertenstein, M., Worley, P. H., Yang, Z.-L., & Zhang, M. 2011. The Community Climate System Model Version 4. Journal of Climate, 24(19), 4973–4991. doi:10.1175/2011JCLI4083.1
Gibbs, J. P., Marquez, C., & Sterling, E. J. 2008. The role of endangered species reintroduction in ecosystem restoration: tortoise–cactus interactions on Española Osland, Galápagos. Restoration Ecology, 16(1), 88–93. doi:10.1111/j.1526- 100X.2007.00265.x
Godínez-Alvarez, H., & Valiente-Banuet, A. 2004. Demography of the columnar cactus Neobuxbaumia macrocephala: a comparative approach using poipulation projection matricies. Plant Ecology, 174, 109–118.
Haizhou, D., & Chong, M. 2009. Study on constructing generalized decision tree by using DNA coding genetic algorithm. 2009 International Conference on Web Information Systems and Mining, 163–167. doi:10.1109/WISM.2009.41
Hannöver, W., & Kordy, H. 2005. Predicting outcomes of in-patient psychotherapy using quality management data: comparing classification and regression trees with logistic regression and linear discriminant analysis. Psychotherapy Research : Journal of the Society for Psychotherapy Research, 15(3), 236–47. doi:10.1080/10503300512331334995
Harrison, R. D., & Rasplus, J.-Y. 2006. Dispersal of fig pollinators in Asian tropical rain forests. Journal of Tropical Ecology, 22(06), 631. doi:10.1017/S0266467406003488
Hiddink, J. G., Ben Rais Lasram, F., Cantrill, J., & Davies, a. J. 2012. Keeping pace with climate change: what can we learn from the spread of Lessepsian migrants? Global Change Biology, 18(7), 2161–2172. doi:10.1111/j.1365-2486.2012.02698.x
54 | P a g e
IPCC, 2007: Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Pachauri, R.K and Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 104 pp.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp, doi:10.1017/CBO9781107415324
IPCC, 2014: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1132 pp.
Loarie, S. R., Duffy, P. B., Hamilton, H., Asner, G. P., Field, C. B., & Ackerly, D. D. 2009. The velocity of climate change. Nature, 462(7276), 1052–1055. doi:10.1038/nature08649
McDonald, C., & McPherson, G. 2005. Pollination of pima pineapple cactus (Coryphantha scheeri var. robustispina): does pollen flow limit abundance of this endangered species. USDA Forest Service Proceedings, 529–532.
McDowell, N., Pockman W. T., Allen C. D., Breshears D. D., Cobb N., Kolb T., Plaut J., Sperry J., West A., Williams D. G., & Yepez, E.. 2008. Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought? New Phytologist 178(4), 719-739
Metcalf, S. J., & Swearer, S. E. 2005. Non-destructive ageing in Notolabrus tetricus using dorsal spines with an emphasis on the benefits for protected, endangered and fished species. Journal of Fish Biology, 66, 1740–1747. doi:10.1111/j.1095- 8649.2005.00704.x
Pal, M., & Mather, P. M. 2003. An assessment of the effectiveness of decision tree methods for land cover classification. Remote Sensing of Environment, 86(4), 554– 565. doi:10.1016/S0034-4257(03)00132-9
Parmesan, C., & Yohe, G. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature, 421(6918), 37–42. doi:10.1038/nature01286
55 | P a g e
Pyke, C. R., Thomas, R., Porter, R. D., Hellmann, J. J., Dukes, J. S., Lodge, D. M., & Chavarria, G. 2008. Current practices and future opportunities for policy on climate change and invasive species. Conservation Biology : The Journal of the Society for Conservation Biology, 22(3), 585–92. doi:10.1111/j.1523-1739.2008.00956.x
Rahel, F. J., & Olden, J. D. 2008. Assessing the effects of climate change on aquatic invasive species. Conservation Biology : The Journal of the Society for Conservation Biology, 22(3), 521–33. doi:10.1111/j.1523-1739.2008.00950.x
Robinet, C., & Roques, A. 2010. Direct impacts of recent climate warming on insect populations. Integrative Zoology, 5(2), 132–42. doi:10.1111/j.1749- 4877.2010.00196.x
Roller, P. 1996. Distribution, growth, and reproduction of Pima pineapple cactus (Coryphantha scheeri Kuntz var robustispina Schott). University of Arizona.
Rubidge, E. M., Monahan, W. B., Parra, J. L., Cameron, S. E., & Brashares, J. S. 2011. The role of climate, habitat, and species co-occurrence as drivers of change in small mammal distributions over the past century. Global Change Biology, 17(2), 696– 708. doi:10.1111/j.1365-2486.2010.02297.x
Sandel, B., Arge, L., Dalsgaard, B., Davies, R. G., Gaston, K. J., Sutherland, W. J., & Svenning, J.-C. 2011. The influence of late quaternary climate-change velocity on species endemism. Science, 334(6056), 660–4. doi:10.1126/science.1210173
Schloss, C. A., Nuñez, T. A., & Lawler, J. J. 2012. Dispersal will limit ability of mammals to track climate change in the Western Hemisphere. Proceedings of the National Academy of Sciences of the United States of America, 109(22), 8606– 8611. doi:10.1073/pnas.1116791109/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1116791109
Schmalzel, R. J., Nixon, R. T., Best, A. L., & Jr, J. A. T. 2004. Morphometric Variation in Coryphantha robustispina ( Cactaceae ). Systematic Botany, 29(3), 553–568.
Segurado, P., & Araujo, M. B. 2004. An evaluation of methods for modelling species distributions. Journal of Biogeography, 31, 1555–1568.
Sodhi, N. S., Koh, L. P., Brook, B. W., & Ng, P. K. L. 2004. Southeast Asian biodiversity: an impending disaster. Trends in Ecology & Evolution, 19(12), 654– 60. doi:10.1016/j.tree.2004.09.006
Stockwell, D. R., & Peterson, A. T. 2002. Effects of sample size on accuracy of species distribution models. Ecological Modelling, 148(1), 1–13. doi:10.1016/S0304- 3800(01)00388-X
56 | P a g e
U.S. Fish and Wildlife Service. 2007. U . S . Fish & Wildlife Service 5-Year Review Determination Pima Pineapple Cactus. 70 FR 5460. Region 2 Regional Office (pp. 1–17).
Van Ewijk, K. Y., Randin, C. F., Treitz, P. M., & Scott, N. a. 2014. Predicting fine-scale tree species abundance patterns using biotic variables derived from LiDAR and high spatial resolution imagery. Remote Sensing of Environment, 150, 120–131. doi:10.1016/j.rse.2014.04.026
Van Klinken, R. D., Lawson, B. E., & Zalucki, M. P. 2009. Predicting invasions in Australia by a neotropical shrub under climate change: the challenge of novel climates and parameter estimation. Global Ecology and Biogeography, 18(6), 688– 700. doi:10.1111/j.1466-8238.2009.00483.x
Walther, G.-R., Roques, A., Hulme, P. E., Sykes, M. T., Pysek, P., Kühn, I., Zobel, M., Bacher, S., Botta-Dukat, Z., Bugmann, H., Czucz, B., Dauber, J., Hickler, T., Jarosik, V., Kenis, M., Klotz, K., Minchin, D., Moora, M., Nentwig., Ott, J., Panov, V. E., Reineking, B., Robinet, C., Semenchenko, V., Solarz, W., Thuiller, W., Vila, M., Vohland, K., & Settele, J. 2009. Alien species in a warmer world: risks and opportunities. Trends in Ecology & Evolution, 24(12), 686–93. doi:10.1016/j.tree.2009.06.008
Yeh, D.-Y., Cheng, C.-H., & Hsiao, S.-C. 2009. Classification knowledge discovery in mold tooling test using decision tree algorithm. Journal of Intelligent Manufacturing, 22(4), 585–595. doi:10.1007/s10845-009-0321-7
57 | P a g e
List of Tables:
Table 1: Journal publications and theses on C. scheeri and other species in the genus; also included non-comprehensively are some relevant reports……………………………….60
Table 2: An evaluation of the classification tree created with elevation for consistency and accuracy and to determine any bias in the data selected to create the tree………….61
Table 3: An evaluation of the classification tree built without elevation as a predictor for consistency and accuracy and to determine any bias in the data selected to create the tree………………………………………………………………………………………62
Table 4: An evaluation of the classification tree built including seasonal precipitation as a predictor (without elevation included) for consistency and accuracy and to determine any bias in the data selected to create the tree………………………………………………...63
58 | P a g e
List of Figures:
Figure 1: Environmental factors evaluated in CART model of C. scheeri distribution: elevation (a), soil texture type (b), precipitation (c), maximum annual temperature (d), minimum annual temperature (e), annual temperature range (f), winter precipitation (g), and summer precipitation (h)…………………………………………………………….64
Figure 2: The classification tree created in MATLAB® used to evaluate environmental factors for C. scheeri predicted presence or predicted absence………………………….65
Figure 3: The classification tree created in MATLAB® without elevation to evaluate environmental factors for C. scheeri predicted presence or predicted absence………….66
Figure 4: The classification tree created in MATLAB® including seasonal precipitation to evaluate environmental factors for C. scheeri predicted presence or predicted absence…………………………………………………………………………………...67
Figure 5: Projected suitable C. scheeri habitat under current climate conditions throughout the state of Arizona, with elevation as variable (a), excluding elevation from the analysis (b), and including winter and summer precipitation as a variable (c)………68
Figure 6: Projected suitable C. scheeri habitat under 2050 climate projections, with elevation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d)……………….69
Figure 7: Projected suitable C. scheeri habitat under 2050 climate projections, without elevation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d)……………….70
Figure 8: Projected suitable C. scheeri habitat under 2050 climate projections, with seasonal precipitation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d)….71
59 | P a g e
Yearsley, 2004 2005 Thomas, 2011 Martorell, Portilla-Alonso and Nobel, 1981 Valverde, 2008 & Martinez-Berdja al,Guadalupe et 2005 al, et Mandujano 2002 EmmaMarris, Sayre, 2005 2001 Consultation, US DOI Fox & 20006 Nino-Murcia, 2007 McDonald, Schmalzel al, et 2004 USFWS review, 5-year 2007 2005 & McPherson, McDonald al, et Phillips 1981 Roller, 1996 1991 Mills, Baker, 2004 Author
Table1: Journal publications and theses on scheeri C. and other species in thegenus; also included non comprehensivelya resome relevantreports. structure assumption. structure rate and segregate remove model that need age removal components corrects for species for the into and that age-a of the component structure dynamics, long-term. as compared to expand of model population Results include scope growth transient to to short-term near-term sensitivity examining for study species and effectiveness dynamic robustness modelling of transient the as study robbinsorum presented "Coryphantha on density. decreaseobserved to observed spp. breeding some previously restore though of the re-establishment to appeared population, vivipara of resiliance included "Coryphantha in fire study to damage. have C. vivipara found to 3.7x was higher rate and fire mortality damage disturbance-tolerant. relate chronic to disturbance, positivley seems whereas to growth seems negatively related, indicating C. werdermannii that is relatively werdermannii for effects of chronic studied "Coryphantha locations. anthropogenic Establishment and fecundity disturbance different at study cellular to prior affectindividuals. uptake damage. CO2 to appear Cold night temps vivipara for freezing studied below -15C kill nocturnal effects. Found "Coryphantha temps below some, -20C while kill temps all observed different levels All to included (if species of a soil any) nurse plant). moisture, of nosolar to strong response radiation. responded werdermannii between of cactusrelationship included"Coryphantha growth, soil moisture, solar rediation in contributions noted study (as the and determinediversity where conservation actions will effective. be most radians, salinensis, sulcata, and wohlschlageri of various evaluate locations to species included of study inventory species in represented leuchtenbergia,"Coryphantha georgii, glanduligera, guerkeana, macro robust. var. not macromeris, statistically found, but Relationship vaupeliana, pallida nurse on association includedplants. ocatcantha, with pulleineana,"Coryphantha in study delicata, nickelsiae, is additional^There danger in field Southernillegal within due Arizona to studies imigrant crossings and drug trafficking. andand activitiesfederal other on state lands more challenging. grazing lands^Private and temporary lands with exclusions are easier after fire; restore to threatened and endangered make species restoration developer to report dynamics within and population known threats, habitat, ^Overview of PPC increases demand the for cconservation banking. threatened and endangered to ^Impacts drivers is one of conservation strongest species of the banking and strong enforcement of regulations *Undergone 9 name changes var @ time (robus robus of article), extensive of description physical * over Schmalzel al, et 2004 based on comments of reviewers. *Determination * time) the and information (at habitat, known on *Extensiveflowering as well of typical description, as description physical and reproduction, *Extensive of description physical includes and on causethreats(e.g. speculations fire,surveys, of low density weevles, loss) habitat of *Description variations, considers of taxonomic robustispina *Analysis between coryphantha relationships Focus(* C. scheeri C. scheeri C.scheeri species distinctions are questionable, does not think distinctions between subspecies are are between distinctions subspecies significant. species questionable, distinctions think does not more be likely limited pollinated ifpollinator to near not another, but C. scheeri C. scheeri is focus, ^ is should remain listed as endangered, thorough overview of existing Baker, delineation literature.2004 species Accepts survey, description of habitat, physical description, and notes on specific populations through various biological on specific and populations notes description, physical of habitat, description survey, C.scheeri
C. scheeri
is mentioned, is "Genus related only) , germination and growth rates, flowering process, C. scheeri var. robustispinaC. scheeri var. C. scheeri .
separate. -
60 | P a g e
Predicted Predicted Predicted Presence Predicted Absence Trial Presence Correct Incorrect Absence Correct Incorrect 1 86% 14% 99% 1% 2 99% 1% 100% 0% 3 99% 1% 100% 0% 4 99% 1% 98% 2% 5 99% 1% 98% 2% Mean 96% 4% 99% 1% SD 0.0513 0.0513 0.0082 0.0082
Table 2: An evaluation of the classification tree created with elevation for consistency and accuracy and to determine any bias in the data selected to create the tree.
61 | P a g e
Predicted Predicted Predicted Presence Predicted Absence Trial Presence Correct Incorrect Absence Correct Incorrect 8 90% 10% 100% 0% 9 99% 1% 99% 1% 10 99% 1% 99% 1% 11 99% 1% 97% 3% 12 98% 2% 97% 3% Mean 97% 3% 98% 2% SD 0.0346 0.0346 0.0100 0.0100
Table 3: An evaluation of the classification tree built without elevation as a predictor for
consistency and accuracy and to determine any bias in the data selected to create the tree.
62 | P a g e
Predicted Predicted Predicted Presence Predicted Absence Trial Presence Correct Incorrect Absence Correct Incorrect 14 92% 8% 98% 2% 15 99% 1% 99% 1% 16 99% 1% 98% 2% 17 99% 1% 98% 2% 18 99% 1% 99% 1% Mean 98% 2% 99% 1% SD 0.0275 0.0275 0.0019 0.0019
Table 4: An evaluation of the classification tree built including seasonal precipitation as a
predictor (without elevation included) for consistency and accuracy and to determine any
bias in the data selected to create the tree.
63 | P a g e
a b
c d
e f
g h
Figure 1: Environmental factors evaluated in CART model of C. scheeri distribution: elevation (a), soil texture type (b), precipitation (c), maximum annual temperature (d), minimum annual temperature (e), annual temperature range (f), winter precipitation (g), and summer precipitation (h).
64 | P a g e
environmental factors for Figure
2
. The classification tree classification The . MATLAB® in created used evaluate to
C. scheeri C.
predicted
presence or
predicted predicted
absence.
65 | P a g e
environmental factors for Figure
3
.
The classification tree classification evaluate The elevation MATLAB® in created without to
C. scheeri C.
predicted
presence or
predicted predicted
absence.
66 | P a g e
predicted precipitation Figure4:
The classification t classification The
absence.
to evaluate to environmental factors for
reeMATLAB® in seasonal created including
C. scheeri C.
predictedp resence or
67 | P a g e
a
b
Figure 5: Projected suitable C. scheeri habitat under current climate conditions throughout the state of Arizona, with elevation as variable (a), excluding elevation from the analysis (b), and including winter and summer precipitation as a variable (c).
68 | P a g e
a
b
c
d
Figure 6: Projected suitable C. scheeri habitat under 2050 climate projections, with elevation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d).
69 | P a g e
a
b
c
d Figure 7: Projected suitable C. scheeri habitat under 2050 climate projections, without elevation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d).
70 | P a g e
a
b
c
d
Figure 8: Projected suitable C. scheeri habitat under 2050 climate projections, with seasonal precipitation included, for scenarios R26 (a), R45 (b), R60 (c), and R85 (d).
71 | P a g e
APPENDIX B: ECOHYDROLOGICAL CONTROLS ON CACTUS FLOWERING
PHENOLOGY AND PLANT REPRODUCTIVE CHARACTERISTICS: A TWO YEAR
OBSERVATIONAL STUDY OF AN ENDANGERED SPECIES
Amí L. Kidder1, Shirley A. Papuga1, David D. Breshears1, Darin J. Law1
1School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721 USA
Targeted for publication in Ecohydrology
72 | P a g e
Abstract
Flowering is a phenological event that plays a critical role in reproductive success, and therefore influences the health and success of plant populations. Flowering phenology can be particularly sensitive to changes in climate, especially in water-limited environments. However, cactus phenology has been little studied and may be relatively independent of pulse-dependent soil moisture dynamics because cacti are able to store water in their tissues, or, alternatively, could use their stored moisture to regulate reproductive phenology. Therefore an improved understanding of cactus phenology is needed, especially for rare and endangered cactus species that drive natural resource management decisions and associated economic costs. Our objective was to identify ecohydrological controls on the flowering phenology and associated plant reproductive characteristics of Pima Pineapple Cactus (Coryphantha scheeri var. robustispina), an endangered species inhabiting a limited region of the Sonoran Desert. We focused on a relatively healthy and dense population located at the northern limit of its range. Over a two year study period (2013 and 2014), we used this dense population to quantify the temporal dynamics of flowering phenology and link these dynamics to climate variables.
We examine the relationships between other abiotic factors such as soil texture and patterns of solar radiation and C. scheeri plant characteristics such as size and number of nodes. Both bud formation and flowering tended to be synchronous across plants.
Flowering events tended to occur within two to six days of precipitation and associated peak soil moisture, usually after events larger than 10 mm. C. scheeri were located mostly on sandy loam or sandy clay loam soils. Fraction of potential annual solar radiation input near the ground (DSF) was relatively high (>0.95). Higher infiltration
73 | P a g e rates where soils had greater sand content tended to be associated with smaller individuals and fewer nodes. These observations provide critical information for natural resource managers on ecohydrological controls on the variability of flowering phenology and associated plant physical characteristics for this endangered species. More generally, our results suggest that cactus flowering phenology can be sensitive to precipitation pulses, despite water stored within the individual.
74 | P a g e
1.0 Introduction
Phenology has been defined as the study of the seasonal timing, duration, and abundance of recurring life-cycle events (e.g. Morisette et al. 2009, Rathcke & Lacey
1985) including plant reproductive events such as budding, flowering, fruiting, seed dispersal and germination. Flowering is a phenological event that plays a critical role in reproductive success (e.g. Giménez-Benavides 2007, Rathcke & Lacey 1985). Because reproduction is one of the key processes that influences the health and success of plant populations, understanding controls on flowering phenology and associated plant characteristics is essential for effective conservation and management of rare, endemic, and endangered species (e.g. Kaye 1999, Lara-Romero et al., 2014, Walck et al., 1999).
Flowering phenology can be particularly sensitive to changes in climate
(Borchert, 1983; Inouye, 2008; Pfeifer et al., 2006), especially in water-limited environments (Bowers, 2007; Crimmins et al., 2010; Liancourt et al., 2012). In water- limited environments, vegetation is particularly responsive to precipitation pulses that deliver moisture to the soil (Holmgren et al., 2006; Huxman et al., 2004; Schwinning et al., 2004; Schwinning & Sala, 2004). In most semiarid ecosystems, these strongly seasonal precipitation pulses are characterized by brief, sometimes intense, rainfall events, segregated by periods of drought (Cable & Huxman, 2004). The primary abiotic factor affecting flowering phenology in water-limited ecosystems has been shown to be the dynamic soil water availability associated with these precipitation pulses (Borchert
1994, Borchert et al. 2004, Bowers & Dimmitt 1994).
Despite providing essential resources for a numerous insect and animal species within water-limited desert ecosystems (e.g. Fleming & Holland 1998, Grant et al. 1979,
75 | P a g e
Valiente-Banuet et al. 1996), cacti and other succulents are underrepresented in studies of precipitation pulse effects on phenological activity. Cactus phenology may be relatively independent of pulse-dependent soil moisture dynamics because they are able to store water in their tissues (Pavón & Briones, 2001). Alternatively, this stored moisture could provide a means to regulate reproductive phenological activity such as flowering and fruit production. In some cactus species, increases in moisture has been associated with improved reproductive phenology (Bowers, 1996; De La Barrera & Nobel, 2004), whereas other cactus species appear to be insensitive to or negatively impacted by increases in moisture (Petit, 2001; Ruiz et al., 2000). Overall, research identifying climatic controls on cactus phenology is scarce (Bustamante & Búrquez, 2008), and this is especially true for rare and endangered cactus species.
The Pima Pineapple Cactus (Coryphantha scheeri var. robustispina, hereafter C. scheeri), is an endangered species inhabiting a limited region of the southwestern United
States restricted to elevations between 700 m and 2000 m with slopes less than 10%
(Roller, 1996; Schmalzel et al., 2004; U.S. Fish and Wildlife Service, 2007). These populations are all located in Santa Cruz and Altar Valleys of Arizona with a small region within Sonora, Mexico (Baker, 2004; McDonald & McPherson, 2005; Roller,
1996; Schmalzel et al., 2004). C. scheeri is a small, hemispherical cactus, with nodes
(also referred to as “tubercles”) rather than ribs that are characteristic of many other cactus species (Roller, 1996; Schmalzel et al., 2004; U.S. Fish and Wildlife Service,
2007). Because of the limited range and the difficulty in locating individuals, C. scheeri research is limited, resulting in vague management options. Understanding the controls on phenological activity for this species is critically important because current
76 | P a g e populations are at risk from mining and expanding anthropogenic activities that degrade the habitat currently recorded and restrict new growth (Sayre, 2005; U.S. Depertment of
Interior, 2001; U.S. Fish and Wildlife Service, 2007).
The objective of our study was to identify ecohydrological controls on flowering phenology and associated plant characteristics of the endangered C. scheeri. We focused on a relatively healthy and dense population located at the northern limit of the C. scheeri range on United States Air Force Plant 44. Over a two year study period (2013 and 2014), we used this dense population to quantify the temporal dynamics of flowering phenology and link these dynamics to climate variables. We examined the relationships between other abiotic factors such as soil texture, infiltration, and patterns of solar radiation associated with C. scheeri characteristics, including size, number of nodes, and pup count.
2.0 Methods
2.1 Site Description
The study site was located at the 567 hectare Air Force Plant 44 (AFP44), approximately 13 km south of Tucson, Arizona. According to a 2010 survey, the AFP44 property hosts a relatively dense population of 308 C. scheeri (alive and dead) (Harrison
Environmental Report, 2010). The C. scheeri population at AFP44 is at the northern boundary of the population range (Figure 1). AFP44 and the surrounding area is predominantly industrial with Tucson International Airport residing to the north and east, and undeveloped land owned by the Tohono O’odham Nation and private property owners to the south and west. The landscape is dominated by creosotebush with patches of mesquite, cholla, and barrel cacti (Harris Environmental Report, 2010). AFP44 is
77 | P a g e dominated by sandy loam soils with variation is evident across the site. Mean annual temperatures range from 13.2oC - 28.4°C and, based on 33 years of data, AFP44 receives
294 mm of annual precipitation (http://www.wrh.noaa.gov/twc/climate/tus.php) with over half falling between July and September (http://www.wrh.noaa.gov/twc/climate/tus.php).
Using location data from the Arizona Heritage Data Management System
(http://www.azgfd.gov/w_c/edits/species_concern.shtml; AHDMS), a map of C. scheeri within the study areas was generated in ArcMap v10.2 (Figure 1). Using this map, at
AFP44, three study sites (AFP44 1, 2 and 3) were randomly selected and also had to meet the following minimum criteria: (1) three or more C. scheeri could be identified together within a 5 m2 area, and (2) the site had to be greater than five meters from any paved roads or structures. All data were carefully gathered at AFP44 with the guidance of U.S.
Fish and Wildlife Service (USFWS) so as not to harm or disturb the endangered C. scheeri plants.
2.2 Climate Data
Daily precipitation (precip, mm) and daily air temperature (temp, oC ) data were obtained from the nearby Tucson International Airport (TIA) meteorological station
(http://www.ncdc.noaa.gov/cdo-web/search). The TIA property is adjacent to AFP44, and the meteorological station is located near enough that measurements are representative of conditions at the three AFP44 sites.
Using the TIA air temperature data, growing degree days (GDDs), a measure of heat accumulation traditionally used by horticulturalists, were calculated as:
푇 + 푇 퐺퐷퐷 = 푚푎푥 푚푖푛 − 푇 2 푏푎푠푒
78 | P a g e
o where Tmax and Tmin are the maximum and minimum daily temperatures in C respectively and Tbase is a species-specific base temperature, which for C. scheeri we assumed to be
10oC, as assumed for other plant species (Mix et al., 2012). Traditionally GDD is calculated as an accumulation of heat by summing of all of the preceding GDD for each day for that calendar year. In our study, we began the accumulation of heat at the start of the water year (October 1) as opposed to the start of the calendar year (January 1).
2.3 Soil Moisture Data
In both years, soil moisture measurements were made daily at AFP44 from June into September to capture the moisture conditions both before and throughout the flowering season using a handheld Fieldscout TDR soil moisture probe (Spectrum
Technologies, Aurora, IL). By inserting the 12 cm probe vertically into the ground, soil moisture measurements reflected an averaged value for the top 12 cm of soil. At each individual C. scheeri, four soil moisture measurements were made at a 45° offset from the cardinal directions (NW, SW, SE and NE; Figure 2). These measurements were made
1 m away from the C. scheeri to prevent the probe from damaging any C. scheeri roots.
Daily average soil moisture at AFP44 was calculated by first averaging all four measurements made at each individual C. scheeri, then averaging those averages for each of the three individuals at each site, and finally averaging together the values for each of the three sites.
2.4 Phenological Data
Daily visual observations of buds and flowering events were made from June into
September for both years. Additionally, Moultrie I60 game cameras (Moultrie, Calera,
AL) were installed at each of the three AFP44 sites. These cameras were programmed to
79 | P a g e take hourly photos to capture the phenological activity of the three C. scheeri at each site and the surrounding vegetation. Each photo is stamped with barometric pressure, temperature, and time. Memory cards were switched out every 30 – 60 days, and batteries were changed every 120 – 150 days.
2.5 Other Abiotic Data
Measurements of infiltration rates were made around each individual in each of the cardinal directions (N, S, E, and W; Figure 2). For each infiltration measurement, a 15 cm-diameter ring infiltrometer was filled with water to a height of 7 cm and allowed to infiltrate into the soil. Infiltration rates were calculated by using the total time necessary for the entire 638.5 mL to infiltrate into the soil. Infiltration measurements were made 1 m away from each C. scheeri plant under the guidance of USFWS so as to protect the C. scheeri root structure from any potential damage by the infiltrometer. These measurements were obtained once during the dry season within the two-year study period. An average infiltration was calculated for each C. scheeri plant by averaging all four measurements made at each individual C. scheeri.
A 5-cm diameter soil core was extracted from the upper 10 cm, located 1 m south of each C. scheeri plant (9 total; Figure 2). Each soil sample was sieved to remove any gravel > 2 cm in diameter. The remaining soil was analyzed for soil texture via Particle
Size Analysis by the Center for Environmental Physics and Mineralogy (CEPM) lab at the University of Arizona.
Direct Site Factor (DSF) is an estimate of potential incoming solar radiation for a cloudless year that ranges from 0.0 to 1.0, where a value of 1.0 indicates complete exposure to incoming solar radiation and 0.0 indicates complete shading (Rich, 1989;
80 | P a g e
Rich et al., 1999). The fractional reductions are determined by nearby vegetation obstructions, as well as topographical effects (the latter are usually relatively small unless in close proximity to tall topographic features). In our study, we derived DSF using images taken with a Nikon Coolpix camera instrumented with a fish-eye lens to capture a full 180° image directly south of each C. scheeri plant (Figure 2). We took images at each
C. scheeri plant (9 total) using the fish-eye lens either at dawn or dusk when there was still daylight but when the sun was not in the sky to avoid the sun appearing in the photo.
The camera was placed on a small bean bag to level the camera and oriented with a marker to indicate North in the photo. The image height was at approximately 25 cm above the ground surface. The photographer was positioned on the north side of each C. scheeri plant so as not to shade the C. scheeri plant in any way. Each image was processed in Delta-T Devices HemiView v. 2.1 software to adjust the contrast so that the sky is white and all other objects which would create shade are black. DSF is the percent of the image that is white (open). We took these images once at each C. scheeri plant in the two-year study period, which, based on the path of the sun in the sky at that location, yields an annual value of DSF.
2.6 C. scheeri Plant Physical Characteristics
The height and diameter of each individual C. scheeri was measured with a measuring tape; for circumference, a string was used to measure around the C. scheeri between the spines. Height and diameter were multiplied together to calculate a C. scheeri size; this size was normalized by dividing by the largest size in our dataset.
Additionally, the nodes on each individual C. scheeri were counted, as well as the number of pups. Number of nodes and number of pups were also normalized by dividing
81 | P a g e by the largest number of nodes and largest number of pups in our dataset. C. scheeri were classified as clusters if they had multiple heads and many more pups than a typical C. scheeri. The height, diameter, and circumference of the entire cluster were measured, though nodes were only counted on the C. scheeri heads, excluding the pups.
3.0 Results
3.1 Climate
Throughout the study period, temperature ranged from approximately -10°C in the winter to 40°C in the summer. In both years, temperature increased in mid-late spring
(~DOY 100) until mid-summer (~DOY 180), when temperatures leveled off after the summer wet season began (Figure 3a, b). Average temperature for during the warmer half of the year (here designated as April 1 – September 7) was 28oC for both 2013 and 2014
(www.ncdc.noaa.gov). In both 2013 and 2014, GDD increased relatively steadily each year, with a slight inflection occurring between May 20 and June 9 (Figure 3c, d).
Winter precipitation was greater in 2013 – 2014 (93 mm; Table 1) than in 2012 –
2013 (73 mm; Table 1). This difference between years was further amplified in total water year precipitation (248 mm in 2014 versus 170 mm in 2013) due in part to contributions from a late summer hurricane in 2014. Within the warmer half of the year, precipitation was sparse prior to the wet season in both years, which began around July 1
(Figure 3e, f). Summer 2013 received fewer large (> 10 mm) storms than summer 2014
(Figure 3e, f).
3.2 Soil Moisture
Prior to the wet season (measured in 2014 only, Figure 3f), soil moisture (upper
12 cm) was low. Soil moisture responded rapidly to precipitation events, reflected in
82 | P a g e increased water content in same-day measurements, when feasible. Subsequently, soils dried rapidly, progressing to background soil moisture levels unless altered by additional precipitation within only a couple of days (Figure 3e, f). Temporal variation in soil moisture was most influenced by a few large events, which generally yielded longer subsequent intervals of above-background soil moisture (Figure 3e, f).
3.3 Flowering Phenology
Bud formation (observed only in 2014) tended to be synchronous across plants.
The initial bud event (5 of 9 plants) occurred on single day (June 10; Table 1), with three of the remaining four plants budding within the next 22 days. This initial bud event was associated with an inflection point in GDD while temperature was increasing (Figure 3b, d) and was prior to the onset of the wet season (Figure 3f). For this initial bud event,
GDD reached 2858 and water year precipitation accumulated to 93 mm (Table 1).
Flowering resulting from this bud event occurred on July 11 (8 of 8 budded plants; Table
1), two days following a large 9 mm precipitation event. A second synchronous bud event
(7 of 9 plants; July 15) occurred during a large (14 mm) precipitation event. The majority of the flowering resulting from this second bud event occurring only five days later (6 of
7 budded plants; July 20; Table 1); the final budded plant flowering the next day (July
21). Three days after the second flowering event, a third bud event occurred (4 of 9 plants; July 23) with no associated precipitation within eight days. Flowering resulting from this third bud event occurred on August 6 (4 of 4 budded plants), five days following a large 11 mm precipitation event.
Flowering events (observed in both years) occurred in mid-summer after onset of the wet season (Figure 3e, f; Table 1). In both years, the first flowering event occurred
83 | P a g e after the first large (>10 mm) event. Flowering events tended to occur within two to six days of precipitation and peak soil moisture (Figure 3e, f), usually after events larger than
10 mm (an exception was the August 11, 2013 flowering event following a 4 mm precipitation event). Consequently, flowering generally occurred while soil moisture was above baseline levels (Figure 3e, f). No clear relationship was observed between flowering and temperature or GDD. Each individual bud flowered for only one day.
Some plants had a single lagging flower that opened one or two days after the flowering date of other buds on that same individual. Duration between initial and final flowering events was 30 to 40 days for both years, within which up to four unique flowering events occurred per individual cactus.
3.4 C. scheeri Plant Physical Characteristics
The three C. scheeri plant physical characteristics—size, number of nodes, and number of pups—all exhibited skewed frequency distributions with many more individuals in the smallest size class of each and few individuals being in the largest class
(Figure 4). C. scheeri at our study site ranged from 12.9 – 28.7cm tall and 10.5 – 45.0 cm diameter. Node count ranged from 47 to 462 nodes. For non-cluster C. scheeri individuals, number of pups ranged from two individual C. scheeri having no pups, to five individual C. scheeri having between one and ten pups. The smaller of the two C. scheeri clusters had 18 pups and the largest had 33 pups.
3.5 Abiotic Factors Influencing C. scheeri Plant Physical Characteristics
C. scheeri individuals spanned a range of the three plant physical characteristics
(size, # of nodes, and # of pups) for three abiotic factors: two soil related metrics
84 | P a g e
(infiltration rate and % sand) and a radiation input metric (fraction of potential incoming solar radiation at a location; DSF).
Both size and node count appear to be related to infiltration, with higher infiltration rates associated with both smaller and larger individuals, creating a concave- up relationship (Figure 5a, d). Maximum pup count appears to decrease with infiltration rate (Figure 5g). At the scale of an individual (for which there were 4 measurements of infiltration surrounding it; Figure 2), there was little variation in infiltration (0.017 SD) although the point most downslope of any individual C. scheeri plant tended to exhibit the slowest rates (personal observation). C. scheeri individuals spanned soil texture types, but were mostly found on sandy loam (7 sites) or sandy clay loam (2 sites). Sand ranged from 54-76% across individual C. scheeri locations; associated silt and clay composition each ranged from approximately 10-30%. Sand composition did not appear to relate to any of the three size metrics (Figure 5b, e, h). Fraction of potential solar radiation input near the ground (DSF) was relatively high for all sites ranging from 0.95 to 0.99 (Figure
5c, f, i).
When each of the three the size metrics are aggregated into three size categories, additional patterns were suggested. There is not a distinct trend with DSF but higher infiltration rates (Figure 6a, b, c) and sand content (Figure 6d, e, f) appear to be associated with larger individuals, more nodes, and high pup counts.
4.0 Discussion
4.1 Ecohydrological Controls on C. scheeri Flowering Phenology
The appearance of the first bud of the season in 2014 was associated with an inflection in the accumulation of GDDs (Figure 3d). Though this observation was only
85 | P a g e made during a single year, the onset of budding and thus the initiation of the flowering season for C. scheeri appears to be associated with an accumulation of energy.
Conversely, based on the observations of our two-year study, C. scheeri flowering events do not appear to be associated with energy or temperature triggers (Figure 3a, b, d).
Instead, flowering events appear to have a moisture trigger associated with large rainfall events and their associated peak in soil moisture (Figure 3e, f). This is not a surprise since previous research has shown that C. scheeri flower in response to the monsoon precipitation pulses (Baker, 2004; McDonald & McPherson, 2005; Roller, 1996;
Schmalzel et al., 2004). However, our research specifically shows that while repeat bud events tend to occur soon after the C. scheeri plants recover from their last flowering event, C. scheeri appear to wait to flower until specific moisture conditions are met.
These conditions seem to occur two to six days following a large (> 10 mm) precipitation event and associated peak in soil moisture. While the flowering phenology of some cactus species has been shown to be associated with the onset of increasing spring temperatures (Bustamante & Búrquez, 2008; Crimmins et al., 2010; Fleming et al., 2001), our results are consistent with other cactus species that flower in the summer (Crimmins et al., 2011). Notably, our study indicates cactus flowering occurring within a fixed window of days following a precipitation event that exceeded a minimum threshold (10 mm). This relationship has important implications for climate change impacts. For reproductive success of C. scheeri precipitation events will need to include appropriately timed, larger events (generally >10 mm). Current projections indicate precipitation will be reduced in the future (Garfin et al., 2013), but information on the distribution and timing of event sizes remains uncertain.
86 | P a g e
4.2 Ecohydrological Controls on Plant Physical Characteristics
In this study, we examined the associations between three physical plant characteristics—plant size, number of nodes, and number of pups—and three abiotic factors—infiltration, %sand, and potential incoming solar radiation. All C. scheeri individuals were located on loam soils, generally sandy. Capillary forces tend to be the major contributor to matric potential in these coarser-grained soils, as opposed to finer- textured soils where adsorptive surface forces dominate. These types of soils tend to have low runoff potential and high infiltration rates even when thoroughly wetted (NRCS
Hydrologic Soil Groups). Interestingly, for our study populations higher infiltration rates and sandier soils tend to be associated with smaller cactus and with less nodes (Figure 6).
While counterintuitive, this may be a result of quick draining soils decreasing the amount of plant available moisture for C. scheeri, as reported in studies with other plant species
(Borchert et al. 2004).
For all C. scheeri locations, the fraction of potential solar radiation input near the ground (DSF) was relatively high (> 0.95; Figure 5c, f, i). At AFP44, creosotebush and cholla are the dominant plant species, with bare ground between cactus and shrubs
(Harris Environmental Report, 2010). We suspect that because of the lack of grass cover at AFP44, C. scheeri have more full sun exposure and reduced competition for shallow soil moisture resources. This is consistent with other studies in water-limited areas where cactus are competing with other species for limited water resources (Coffin & Lauenroth,
1990).
87 | P a g e
4.3 Implications for Management and Conservation Practices
Although previous research has suggested that precipitation pulses trigger responses in C. scheeri (McDonald & McPherson, 2005; Roller, 1996), there is still a paucity of knowledge related to the size and distribution of the pulse needed to trigger these responses. This leaves species managers without a sufficient understanding of the conditions C. scheeri require for reproductive success. Our study suggests that C. scheeri are larger on loamy soils with some sand, perhaps because infiltration can occur but not too quickly. Soil texture may also be important in the context of its effect on evaporation loss. Additionally, we show that individuals in this high-density population of C. scheeri were located in sparsely vegetated and sunny locations, especially those with minimal competition for surface water resources. With this knowledge, locations that host known
C. scheeri with these abiotic conditions could be preferentially protected to maximize successful conservation. For reproductive success, our study also suggests that C. scheeri is dependent on large (> 10 mm) precipitation events that occur in the warm summer season. With this knowledge, managers could focus on conservation strategies associated with harvesting seeds (Fox & Nino-Murcia, 2005) in years with good precipitation.
Additionally, this understanding could be used to focus on relocation management strategies (Bouzat et al., 2008; Massei et al., 2010) in which individuals could be relocated to areas with ideal abiotic conditions which may be converting to a precipitation regime conducive to C. scheeri establishment.
We note that the results of this study are based on only 1 year of bud data and two years of flowering data, both within one population of C. scheeri within the known size and distribution of this species. Data collected here present C. scheeri response to climate
88 | P a g e events within the study period and are not indicative of species response to extreme climate events. Furthermore, results describing flowering phenology and individual site characteristics presented here do not indicate the health or distribution of the overall population across its range.
5.0 Conclusions
Here we identify ecohydrological controls on the flowering phenology and associated plant physical characteristics of an endangered cactus species inhabiting a limited region of the Sonoran Desert. We used a dense population to quantify the temporal dynamics of flowering phenology and link these dynamics to climate variables and examine the relationships between abiotic conditions such as soil texture and patterns of solar radiation and the reproductive plant characteristics such as size and number of nodes. We show that while bud events appear to be unassociated with moisture, flowering events are preceded by large rainfall events and their associated peaks in soil moisture. We also show that C. scheeri prefer loamy soils, with some sand so that infiltration can occur but not too quickly and that C. scheeri prefer sparse sunny locations, especially those with minimal competition for surface water resources. With this knowledge, locations that host known C. scheeri with these ecohydrological controls should be preferentially protected to maximize successful conservation.
Acknowledgements
Support for this project was provided in part by the U.S. Air Force Life Cycle
Management Center and Raytheon Company Missile Systems Environmental, Health,
Safety, and Sustainability Department. Soil moisture data collection was made possible thanks to Dr. Michael Crimmins, and soil particle size analysis measurements were
89 | P a g e greatly aided by Dr. Prakash Dhakal and the CEPM lab (Center for Environmental
Physics and Mineralogy). Special thanks to Thomas Hafkemeyer, Dianna Holgate, Katie
Merems, and Heather Spitzer for additional assistance for the field campaign. We would like to acknowledge and thank the members of the Papuga and Breshears labs at the
University of Arizona, School of Natural Resources and the Environment for their insights and support. A special thanks to Julie Crawford with U.S. Fish and Wildlife
Service and Sabra Tonn with the Arizona Game and Fish Department for assistance in obtaining PPC spatial location data.
90 | P a g e
References
Baker, M. 2004. Phenetic analysis of Coryphantha, section Robustispina (cactaceae), part 1: stem characters.U.S. Fish and Wildlife Service report (pp. 1–21). Tempe, Arizona
Borchert, R. 1983. Phenology and control of flowering in tropical trees. Biotropica, 15, 81–89. doi:10.2307/2387949
Borchert, R. 1994. Induction of rehydration and bud break by irrigation or rain in deciduous trees of a tropical dry forest in costa-rica. Trees-Structure and Function, 8, 198–204. doi:10.1007/BF00196847
Borchert, R., Meyer, S. A., Felger, R. S., & Porter-Bolland, L. 2004. Environmental control of flowering periodicity in Costa Rican and Mexican tropical dry forests. Global Ecology and Biogeography, 13, 409–425. doi:10.1111/j.1466- 822X.2004.00111.x
Bouzat, J. L., Johnson, J. A., Toepfer, J. E., Simpson, S. A, Esker, T. L., & Westemeier, R. L. 2008. Beyond the beneficial effects of translocations as an effective tool for the genetic restoration of isolated populations. Conservation Genetics, 10(1), 191– 201. doi:10.1007/s10592-008-9547-8
Bowers, J. E. 1996. More flowers or new cladodes? environmental correlates and biological consequences of sexual reproduction in a Sonoran Desert prickly pear cactus, Opuntia engelmannii. Bulletin of the Torrey Botanical Club, 123, 34–40.
Bowers, J. E. 2007. Has climatic warming altered spring flowering date of sonoran desert shrubs? The Southwestern Naturalist, 52(3), 347–355.
Bowers, J. E., & Dimmitt, M. A. 1994. Flowering phenology of six woody plants in the northern sonoran desert. Bulletin of the Torrey Botanical Club, 121(3), 215–229.
Bustamante, E., & Búrquez, A. 2008. Effects of plant size and weather on the flowering phenology of the organ pipe cactus (Stenocereus thurberi). Annals of Botany, 102, 1019–1030. doi:10.1093/aob/mcn194
Cable, J. M., & Huxman, T. E. 2004. Precipitation pulse size effects on Sonoran Desert soil microbial crusts. Oecologia, 141(2), 317–24. doi:10.1007/s00442-003-1461-7
Coffin, D. P., & Lauenroth, W. K. 1990. A gap dynamics simulation model of succession in semiarid grassland. Ecological Modelling, 49, 229–266.
91 | P a g e
Crimmins, T. M., Crimmins, M. A, & Bertelsen, C. D. 2011. Onset of summer flowering in a “Sky Island” is driven by monsoon moisture. The New Phytologist, 191(2), 468–79. doi:10.1111/j.1469-8137.2011.03705.x
Crimmins, T. M., Crimmins, M. A., & David Bertelsen, C. 2010. Complex responses to climate drivers in onset of spring flowering across a semi-arid elevation gradient. Journal of Ecology, 98, 1042–1051. doi:10.1111/j.1365-2745.2010.01696.x
De La Barrera, E., & Nobel, P. S. 2004. Carbon and water relations for developing fruits of Opuntia ficus-indica (L.) Miller, including effects of drought and gibberellic acid. Journal of Experimental Botany, 55, 719–729. doi:10.1093/jxb/erh084
Fleming, T. H., & Holland, J. N. 1998. The evolution of obligate pollination mutualisms: Senita cactus and senita moth. Oecologia, 114, 368–375. doi:10.1007/s004420050459
Fox, J., & Nino-Murcia, A. 2005. Status of species conservation banking in the united states. Conservation biology, 19(4), 996–1007. doi:10.1111/j.1523- 1739.2005.00231.x
Garfin, G., Jardine, A., Merideth, R., Black, M., & LeRoy, S. 2013. Assessment of Climate Change in the Southwest United States. (G. Garfin, A. Jardine, R. Merideth, M. Black, & S. LeRoy, Eds.). Washington, DC: Island Press/Center for Resource Economics. doi:10.5822/978-1-61091-484-0
Giménez-Benavides, L., Escudero, A., & Iriondo, J. M. 2007. Reproductive limits of a late-flowering high-mountain Mediterranean plant along an elevational climate gradient. New Phytologist, 173, 367–382. doi:10.1111/j.1469-8137.2006.01932.x
Grant, V., Grant, K. A., & Hurd, P. D. 1979. Pollination of Opuntia lindheimeri and related species. Plant Systematics and Evolution, 132, 313–320. doi:10.1007/BF00982393
Holmgren, M., Stapp, P., Dickman, C. R., Gracia, C., Graham, S., Gutiérrez, J. R., Hice, C., Jaksic, F,. Kelt, D. A., Letnic, M., Lima, M., Lopez, B. C., Meserve, P. L., Milstead, W. B., Polis, G. A., Previtali, M. A., Richter, M., Sabate, S., & Squeo, F. A. 2006. Extreme climatic events shape arid and semiarid ecosystems. Frontiers in Ecology and the Environment, 4(2), 87–95.
Huxman, T. E., Cable, J. M., Ignace, D. D., Eilts, J. A., English, N. B., Weltzin, J., & Williams, D. G. 2004. Response of net ecosystem gas exchange to a simulated precipitation pulse in a semi-arid grassland: the role of native versus non-native grasses and soil texture. Oecologia, 141(2), 295–305. doi:10.1007/s00442-003- 1389-y
92 | P a g e
Inouye, D. W. 2008. Effects of climate change on phenology, frost damage, and floral abundance of montane wildflowers. Ecology, 89, 353–362. doi:10.1890/06-2128.1
Kaye, T. N. 1999. From flowering to dispersal: Reproductive ecology of an endemic plant, Astragalus australis var. olympicus (Fabaceae). American Journal of Botany, 86, 1248–1256. doi:10.2307/2656772
Lara-Romero, C., García-Camacho, R., Escudero, A., & Iriondo, J. M. 2014. Genetic variation in flowering phenology and reproductive performance in a Mediterranean high-mountain specialist, Armeria caespitosa (Plumbaginaceae). Botanical Journal of the Linnean Society, 176(3), 384–395. doi:10.1111/boj.12208
Liancourt, P., Spence, L. A., Boldgiv, B., Lkhagva, A., Helliker, B. R., Casper, B. B., & Petraitis, P. S. 2012. Vulnerability of the northern Mongolian steppe to climate change: insights from flower production and phenology. Ecology, 93, 815–824. doi:10.1890/11-1003.1
Massei, G., Quy, R. J., Gurney, J., & Cowan, D. P. 2010. Can translocations be used to mitigate humanwildlife conflicts? Wildlife Research, 37, 428–439. doi:10.1071/WR08179
McDonald, C., & McPherson, G. 2005. Pollination of pima pineapple cactus (Coryphantha scheeri var. robustispina): does pollen flow limit abundance of this endangered species. USDA Forest Service Proceedings, 529–532.
Mix, K., Lopes, V. L., & Rast, W. 2012. Growing season expansion and related changes in monthly temperature and growing degree days in the Inter-Montane Desert of the San Luis Valley, Colorado. Climate Change, 114, 723-744 doi 10.1007/s10584-012- 0448-y.
Morisette, J. T., Richardson, A. D., Knapp, A. K., Fisher, J. I., Graham, E. A., Abatzoglou, J., Wilson, B. E., Breshears, D. D., Henebry, g. M., Hanes, J. M., & Liang, L. 2009. Tracking the rhythm of the seasons in the face of global change: Phenological research in the 21 st century. Frontiers in Ecology and the Environment.7(5), 253-260 doi:10.1890/070217
Pavón, N. P., & Briones, O. 2001. Phenological patterns of nine perennial plants in an intertropical semi-arid Mexican scrub. Journal of Arid Environments, 49, 265–277. doi:10.1006/jare.2000.0786
Petit, S. 2001. The reproductive phenology of three sympatric species of columnar cacti on Curaçao. Journal of Arid Environments, 49, 521–531. doi:10.1006/jare.2001.0801
93 | P a g e
Pfeifer, M., Heinrich, W., & Jetschke, G. 2006. Climate, size and flowering history determine flowering pattern of an orchid. Botanical Journal of the Linnean Society, 151, 511–526. doi:10.1111/j.1095-8339.2006.00539.x
Rathcke, B., & Lacey, E. P. 1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics, 16(1), 179–214. doi:10.1146/annurev.es.16.110185.001143
Rich P. M. (1989) A Manual for Analysis of Hemispherical Canopy Photography. Los Alamos National Laboratory Report LA-11733-M. Los Alamos National Laboratory, Los Alamos.
Rich P. M.,Wood J., Vieglais D. A., Burek K. & Webb N. (1999) Guide to Hemiview: Software for Analysis of Hemispherical Photography. Delta-T Devices, Cambridge.
Roller, P. 1996. Distribution, growth, and reproduction of Pima pineapple cactus (Coryphantha scheeri Kuntz var robustispina Schott). University of Arizona.
Ruiz, A., Santos, M., Cavelier, J., & Soriano, P. J. 2000. Phenological study of Cactaceae in the dry enclave of Tatacoa, Colombia. Biotropica, 32, 397–407.
Sayre, N. F. 2005. Interacting effects of landownership, land use, and endangered species on conservation of Southwestern U. S. Rangelands. Conservation Biology, 19(3), 783–792.
Schmalzel, R. J., Nixon, R. T., Best, A. L., & Jr, J. A. T. 2004. Morphometric variation in Coryphantha robustispina (Cactaceae). Systematic Botany, 29(3), 553–568.
Schwinning, S., & Sala, O. E. 2004. Hierarchy of responses to resource pulses in arid and semi-arid ecosystems. Oecologia, 141(2), 211–20. doi:10.1007/s00442-004-1520-8
Schwinning, S., Sala, O. E., Loik, M. E., & Ehleringer, J. R. 2004. Thresholds, memory, and seasonality: understanding pulse dynamics in arid/semi-arid ecosystems. Oecologia, 141(2), 191–3. doi:10.1007/s00442-004-1683-3
U.S. Depertment of Interior. 2001. United States Department of the Interior, U. S. Fish and Wildlife Service AESO/SE 2-21-01-F-271 (pp. 1–16). Phoenix, Arizona.
U.S. Fish and Wildlife Service. 2007. U . S . Fish & Wildlife Service 5-Year Review Determination Pima Pineapple Cactus. 70 FR 5460. Region 2 Regional Office (pp. 1–17).
Valiente-Banuet, A., Arizmendi, M. D. C., Rojas-Martínez, A., & Domínguez-Canseco, L. 1996. Ecological relationships between columnar cacti and nectar-feeding bats in Mexico. Journal of Tropical Ecology, 12, 103. doi:10.1017/S0266467400009330
94 | P a g e
Walck, J. L., Baskin, J. M., & Baskin, C. C. 1999. Ecology of the endangered species Solidago shortii. VII. Survivorship and flowering, and comparison with common, geographically-widespread Solidago species. Journal of the Torrey Botanical Society, 126, 124–132. doi:10.2307/2997288
95 | P a g e
List of Figures:
Figure 1: C. scheeri habitat range with AFP44 highlighted. Distribution of C. scheeri is presented in dark grey and study site is indicated by black area………………………...99
Figure 2: Diagram indicating the measurement design for field data collected.
Measurements depicted here include volumetric water content, infiltration, soil texture type, and hemispherical photos. Per guidance from Arizona Fish and Wildlife, invasive measurements were made 1m away from the C. scheeri so as not to risk damaging the root structure……………………………………………………………………………100
Figure 3: Daily measurements of temperature (a, b), growing degree days (GDD) (c, d), precipitation and volumetric water content (e, f) are presented here in conjunction with phenological events (e.g. bud formation, square, and bloom events, diamonds) for both study years……………………………………………………………………………...101
Figure 4: A histogram of C. scheeri size indicators: size (a), node count (b), and pup count (c) frequencies……………………………………………………………………102
Figure 5: Physical C. scheeri habitat characteristics as they related to C. scheeri size indicators. Site characteristics include infiltration rate (a, d, g), % sand as an indicator of soil texture type (b, e, h), and Direct Site Factor (DSF) for fraction of potential annual radiation input (c, f, i). Size indicators are normalized for context, and include size (a, b, c), node count (d, e, f), and pup count (g, h, i)………………………………………….103
Figure 6: C. scheeri histograms of grouped data for infiltration (a, b c), % sand (d, e, f), and Direct Site Factor (DSF) for fraction of potential annual radiation input (g, h, i); each as related to size (a, d, g), node count (b, e, h), and pup count (c, f, i). Different letters above bars indicate significant differences within a given panel; significance was
96 | P a g e determined by the Friedman’s Rank Test, and is set at P<0.10. Size is determined to be effected by infiltration (a) and %sand (d) at p=0.096. Node count is determined to be effected by infiltration (b) and %sand (e) at p=0.079. Other trends are determined to be non-significant………………………………………………………………………….105
97 | P a g e
List of Tables:
Table 1: Data describing chronology of climatic events and phenological response which enhances the time series data presented.
98 | P a g e
Figure 1: C. scheeri habitat range with AFP44 highlighted. Distribution of C. scheeri is presented in dark grey and study site is indicated by black area.
99 | P a g e
N
Soil Moisture
W E
1m S
45° Infiltration
1m Soil Core Hemi photo
Soil core
Figure 2: Diagram indicating the measurement design for field data collected.
Measurements depicted here include volumetric water content, infiltration, soil texture type, and hemispherical photos. Per guidance from Arizona Fish and Wildlife, invasive measurements were made 1m away from the C. scheeri so as not to risk damaging the root structure.
100 | P a g e
Figure 3: Daily measurements of temperature (a, b), growing degree days (GDD) (c, d), precipitation and volumetric water content (e, f) are presented here in conjunction with phenological events (e.g. bud formation, square, and bloom events, diamonds) for both study years.
101 | P a g e
10 a. b. c.
5
AFP44 Frequency
0-355 355-710 710-1065 0-160 160-320 320-480 0-15 15-30 30-45 Size (cm2) Nodes (#) Pups (#)
Figure 4: A histogram of C. scheeri size indicators: size (a), node count (b), and pup count (c) frequencies.
102 | P a g e
1.0 a. b. c. 0.75
0.50 Size
Normalized 0.25
1.0 0.03 0.04 0.05 0.06 40 50 60 70 80 0.9 0.92 0.94 0.96 0.98 d. e. f. 0.75
0.50 Normalized Node Count Node 0.25
1.0 0.03 0.04 0.05 0.06 40 50 60 70 80 0.9 0.92 0.94 0.96 0.98 g. h. i. 0.75
0.50 Pup Pup Count Normalized 0.25
.03 .04 .05 .06 40 50 60 70 80 .90 .92 .94 .96 .98 Infiltration Rate (mm/s) %Sand DSF
Figure 5: Physical C. scheeri habitat characteristics as they related to C. scheeri size indicators. Site characteristics include infiltration rate (a, d, g), % sand as an indicator of soil texture type (b, e, h), and Direct Site Factor (DSF) for fraction of potential annual radiation input (c, f, i). Size indicators are normalized for context, and include size (a, b, c), node count (d, e, f), and pup count (g, h, i).
103 | P a g e
0.06 a. b. c. a a a
0.04 a
b b a
0.02 Infiltration (mm/s)
80.0 1 2 3 1 2 3 1 2 3 d. e. f. 70.0 a a a a 60.0
%Sand b b a 50.0
1.00
g. h. a i. 0.98 a a a a a a
0.96
0.94 DSF 0.92
0.90
Small Med Large Low Med High None Yes Cluster Size Node Count Pups
Figure 6: C. scheeri histograms of grouped data for infiltration (a, b c), % sand (d, e, f), and Direct Site Factor (DSF) for fraction of potential annual radiation input (g, h, i); each as related to size (a, d, g), node count (b, e, h), and pup count (c, f, i). Different letters above bars indicate significant differences within a given panel; significance was determined by the Friedman’s Rank Test, and is set at P<0.10. Size is determined to be effected by infiltration (a) and %sand (d) at p=0.096. Node count is determined to be effected by infiltration (b) and %sand (e) at p=0.079. Other trends are determined to be non-significant.
104 | P a g e
Y2013 Y2014 # of unique bud/flower events 4/4 3/3 Date of 1st bud June 10 Date of 1st flower July 10 July 11 Water year GDD at 1st bud 2187 Water year GDD at 1st flower 2830 2858 Water year precip 1st bud 75 mm 93 mm Water year precip 1st flower 100 mm 111 mm Temp at 1st bud 32.5°C Temp at 1st flower 30.3°C 31.4°C
Table 1: Data describing chronology of ecohydrological events and phenological response which enhances the time series data presented.
105 | P a g e
APPENDIX C: SUPPLEMENTAL DATA FROM SANTA RITA EXPERIMENTAL
RANGE ON PLANT PHYISCAL CHARACTERISTICS AND ASSOCIATED
ABIOTIC CHARACTERISTICS
106 | P a g e
1.0 Methods
1.1 Site Description
The Santa Rita Experimental Range (SRER) is a large, multi-use area located approximately 40 km south of Tucson, Arizona and covers about 21043 hectares (Figure
1). The area is used for camping, grazing, hunting, and has been a location for study and experiments for over 100 years, managed initially by the U.S. Forest Service in 1902, and transferring authority to the University of Arizona in 1987 (http://cals.arizona.edu/srer/).
Soil types and vegetation cover vary greatly within the SRER as a consequence of an elevational gradient and a variety of different anthropogenic activities (e.g. grazing)
(http://cals.arizona.edu/srer/). Mean annual temperature at the SRER is 18.44°C
(McClaran & Wei, 2014) and based on 80 years of data, mean annual precipitation ranges between 275 mm – 450 mm across the range (McClaran, 2010).
Using location data from the Arizona Heritage Data Management System
(http://www.azgfd.gov/w_c/edits/species_concern.shtml; AHDMS), maps of C. scheeri within the study areas were generated in ArcMap v10.2 (Figure 1). Maps generated using
AHDMS data were used to identify locations of C. scheeri within the SRER. However, because the small C. scheeri are very sparse within the SRER, individuals were difficult to locate. In the end, our study made use of the randomly selected C. scheeri individuals that could be successfully located within the SRER (11 total). All data were carefully gathered with the guidance of U.S. Fish and Wildlife Service (USFWS) so as not to harm or disturb the endangered C. scheeri plants.
107 | P a g e
1.2 Abiotic Data
A 5-cm diameter soil core was extracted from the upper 10 cm, located 1 m south of each C. scheeri plant (11 total). Each of these soil samples was sieved to remove any gravel > 2 cm in diameter. The remaining soil was analyzed for soil texture via Particle
Size Analysis by the Center for Environmental Physics and Mineralogy (CEPM) lab at the University of Arizona.
Direct Site Factor (DSF) is an estimate of potential incoming solar radiation for a cloudless year that ranges from 0.0 to 1.0, where a value of 1.0 indicates complete exposure to incoming solar radiation and 0.0 indicates complete shading (Rich, 1989;
Rich et al., 1999). The fractional reductions are determined by nearby vegetation obstructions, as well as topographical effects (the latter are usually relatively small unless in close proximity to tall topographic features). In our study, we derived DSF using images taken with a Nikon Coolpix camera instrumented with a fish-eye lens to capture a full 180° image directly south of each C. scheeri plant. We took images at each C. scheeri plant (11 total) using the fish-eye lens either at dawn or dusk when there was still daylight but when the sun was not in the sky to avoid the sun appearing in the photo. The camera was placed on a small bean bag to level the camera and oriented with a marker to indicate North in the photo. The image height was at approximately 25 cm above the ground surface. The photographer was positioned on the north side of each C. scheeri plant so as not to shade the C. scheeri plant in any way. Each image was processed in
Delta-T Devices HemiView v. 2.1 software to adjust the contrast so that the sky is white and all other objects which would create shade are black. DSF is the percent of the image that is white (open). We took these images once at each C. scheeri plant in the two-year
108 | P a g e study period, which, based on the path of the sun in the sky at that location, yields an annual value of DSF.
1.3 C. scheeri Plant Physical Characteristics
The height and diameter of each individual C. scheeri was measured with a measuring tape; for circumference, a string was used to measure around the C. scheeri between the spines. Height and diameter were multiplied together to calculate a C. scheeri size; this size was normalized by dividing by the largest size in our dataset.
Additionally, the nodes on each individual C. scheeri were counted, as well as the number of pups. Number of nodes and number of pups were also normalized by dividing by the largest number of nodes and largest number of pups in our dataset, which included data from Appendix B.
2.0 Results
2.1 C. scheeri Plant Physical Characteristics
The three C. scheeri plant physical characteristics—size, number of nodes, and number of pups—all varied substantially among individuals (Figure 2) and exhibited skewed frequency distributions with many more individuals in the smallest size class of each and few individuals being in the largest class (Figure 3). C. scheeri at our study sites ranged from 3.5 -12.0 cm tall and 6.3 -12.5 cm diameter. Node count ranged from 20 to
56 nodes. Number of pups ranged from five individual C. scheeri having no pups, to six individual C. scheeri having between one and ten pups.
2.2 Abiotic Factors Influencing C. scheeri Plant Physical Characteristics
C. scheeri individuals spanned a range of the three plant physical characteristics
(size, # of nodes, and # of pups) for two abiotic factors: a soil metric (% sand) and a
109 | P a g e radiation input metric (fraction of potential incoming solar radiation at a location; DSF).
C. scheeri individuals spanned soil texture types, but were mostly found on sandy loam
(6 sites) or sandy clay loam (4 sites), with one individual on clay loam. Sand ranged from
40-78% across individual C. scheeri locations; associated silt and clay composition each ranged from approximately 10-30%. Sand composition did not appear to relate to any of the three size metrics (Figure 4b, e, h). Fraction of potential solar radiation input near the ground (DSF) was relatively high for all sites ranging from 0.90 to 0.99, with most >
0.95.
When each of the three the size metrics are aggregated into three size categories relative to C. scheeri size values studied at another site (Appendix B), additional patterns were suggested. Higher DSF values (Figure 4f) appear to be associated lack of pups.
There were not pronounced differences in size with % sand (Figure 4c).
110 | P a g e
References
McClaran, M.P., D.M. Browning, and C. Huang. 2010 Temporal dynamics and spatial variability in desert grassland vegetation. In R.H. Webb, D.E. Boyer, and R.M. Turner (eds.). Repeat Photography: Methods and Applications in the Natural Sciences. Island Press. Pages 145-166.
McClaran, M. P., & Wei, H. Y. 2014. Recent drought phase in a 73-year record at two spatial scales: implications for livestock production on rangelands in the Southwestern United States. Agricultural and Forest Meteorology, 197, 40-51. doi: 10.1016/j.agformet.2014.06.004
Rich P. M. (1989) A Manual for Analysis of Hemispherical Canopy Photography. Los Alamos National Laboratory Report LA-11733-M. Los Alamos National Laboratory, Los Alamos.
Rich P. M.,Wood J., Vieglais D. A., Burek K. & Webb N. (1999) Guide to Hemiview: Software for Analysis of Hemispherical Photography. Delta-T Devices, Cambridge.
111 | P a g e
List of Figures:
Figure 1: C. scheeri habitat range with SRER highlighted. Distribution of C. scheeri is demonstrated in dark grey and study sites are indicated by black area………………...113
Figure 2: Physical C. scheeri habitat characteristics as they related to C. scheeri size indicators. Site characteristics include % sand as an indicator of soil texture type (a, c, e), and Direct Site Factor (DSF) for fraction of potential annual radiation input (b, d, f). Size indicators are normalized for context, and include size (a, b), node count (c, d), and pup count (e, f)………………………………………………………………………………114
Figure 3: A histogram of C. scheeri size indicators: size (a), node count (b), and pup count (c) frequencies……………………………………………………………………115
Figure 4: C. scheeri histograms of grouped data for % sand (a, b, c), and Direct Site
Factor (DSF) for fraction of potential annual radiation input (d, e, f); each as related to size (a, d), node count (b, e), and pup count (c, f). Different letters above bars indicate significant differences within a given panel; significance was determined by the
Friedman’s Rank Test, and is set at P<0.10. DSF is determined to have a relationship with pup count with a p-value of 0.054. Other trends are determined to be non- significant……………………………………………………………………………….116
112 | P a g e
Figure 1: C. scheeri habitat range with SRER highlighted. Distribution of C. scheeri is demonstrated in dark grey and study sites are indicated by black area.
113 | P a g e
0.50
a. b.
0.25
Size Normalized
0.50 40 50 60 70 80 0.9 0.92 0.94 0.96 0.98 c. d.
0.25
Normalized Node Count Node
0.5 40 50 60 70 80 0.9 0.92 0.94 0.96 0.98 0.4 e. f.
0.3
0.2
Pup Pup Count Normalized
0.1
0 40 50 60 70 80 .90 .92 .94 .96 .98 %Sand DSF
Figure 2: Physical C. scheeri habitat characteristics as they related to C. scheeri size indicators. Site characteristics include % sand as an indicator of soil texture type (a, c, e), and Direct Site Factor (DSF) for fraction of potential annual radiation input (b, d, f). Size indicators are normalized for context, and include size (a, b), node count (c, d), and pup count (e, f).
114 | P a g e
10 a. b. c.
5
SRER Frequency
0-355 355-710 710-1065 0-160 160-320 320-480 0-15 15-30 30-45 Size (cm2) Nodes (#) Pups (#)
Figure 3: A histogram of C. scheeri size indicators: size (a), node count (b), and pup count (c) frequencies.
115 | P a g e
80.0 a. b. c. 70.0 a a
60.0 %Sand 50.0
1.00 d. e. a f. 0.98
0.96 b
0.94 DSF 0.92
0.90
Small Med Large Low Med High None Yes Cluster Size Node Count Pups
Figure 4: C. scheeri histograms of grouped data for % sand (a, b, c), and Direct Site
Factor (DSF) for fraction of potential annual radiation input (d, e, f); each as related to size (a, d), node count (b, e), and pup count (c, f). Different letters above bars indicate significant differences within a given panel; significance was determined by the
Friedman’s Rank Test, and is set at P<0.10. DSF is determined to have a relationship with pup count with a p-value of 0.054. Other trends are determined to be non- significant.
116 | P a g e