Summer habitat use by Sonora chub in Sycamore Creek, Santa Cruz County, Arizona
Item Type text; Thesis-Reproduction (electronic)
Authors Carpenter, Jeanette, 1960-
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Download date 01/10/2021 06:41:41
Link to Item http://hdl.handle.net/10150/278119 INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600
Order Number 1348486
Summer habitat use by Sonora chub in Sycamore Creek, Santa Cruz County, Arizona
Carpenter, Jeanette, M.S.
The University of Arizona, 1992
UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106
SUMMER HABITAT USE BY SONORA CHUB
IN SYCAMORE CREEK, SANTA CRUZ COUNTY, ARIZONA
by
Jeanette Carpenter
A Thesis Submitted to the Faculty of the
SCHOOL OF RENEWABLE NATURAL RESOURCES
In Partial Fulfillment of the Requirements For the Degree of
MASTER OF SCIENCE WITH A MAJOR IN WILDLIFE AND FISHERIES SCIENCE
In the Graduate College
THE UNIVERSITY OF ARIZONA
1992 2
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 accurate acknowledgement of 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 judgement 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:
APPROVAL BY THESIS COMMITTEE
This thesis has been approved on the date shown below:
<9. 3 3 faen/ul 93* U O. E. Maughan 'O Date' -i Professor, Wildlife and Fisheries Science
d Q. WkiLs \j W. E. Matter Date Associate Professor, Wildlife and Fisheries Science
G. H. Rodda Date Research Associate 3
Acknowledgements
I sincerely thank Dr. O. Eugene Maughan, my major advisor, for suggesting that I work on this project, and for his continuous encouragement and professional and personal guidance. I thank the other members of my committee, Dr. Gordon Rodda and Dr. William Matter, for their many helpful suggestions and comments. I am grateful to Jim Terrell for letting me continue to work on my thesis, and supplying me with the excellent resources at the National Ecology Research Center. The statistical analysis would not have been possible without the advice and comments by Brian Cade. Special thanks to Pete Smith and Katrina Estrada for their dedication and hard work in assisting with the field work and data entry. Pat Collins and Joel Lusk also helped with data entry. I appreciate the friendship and discussions with my fellow graduate students, especially Lorena Wada, Paul Barrett, Dan Welch, and Genice Froehlich. Thanks also to Jeriy Stei'ferud and Bob Anderson of the U.S. Forest Service. This work was financially supported by the Arizona Game and Fish Department and the U.S. Forest Service. 4
TABLE OF CONTENTS
LIST OF FIGURES 7
LIST OF TABLES 8
ABSTRACT 9
INTRODUCTION 10
OBJECTIVES 12
STUDY AREA 12
METHODS 14
CLIMATE AND WATER TEMPERATURE 14
MACROHABITAT ANALYSIS 15
Macrohabitat Data Collection 15
Macrohabitat Data Analysis 17
MICROHABITAT ANALYSIS 18
Microhabitat Data Collection 18
Microhabitat Data Analysis 22
LIFE HISTORY NOTES 23
RESULTS 24
CLIMATE AND WATER TEMPERATURE 24
HABITAT LOSS 26
FISH LOSS 26
MACROHABITAT ANALYSIS 26
Changes in Abundance 26
Comparisons of Macrohabitat Variables 28
Depth. Width. Surface Area 29 5
Dominant Substrate 29
Floating Cover 29
MICROHABITAT ANALYSIS 32
Velocity 33
Mean Depth of Water Column 33
Position in Water Column 36
Depth Heterogeneity 36
Shade 36
Instream Cover 40
Distance to Cover 40
Maximum Substrate Size 43
DISCUSSION 45
JUSTIFICATION FOR METHODS 45
CLIMATE AND WATER TEMPERATURE 46
MACROHABITAT ANALYSIS 47
MICROHABITAT ANALYSIS 49
Significance of Microhabitat Variables 50
Velocity 50
Mean Column Depth. Position in Water Column, and Depth Heterogeneity 51
Sunlit versus Shaded Areas 52
Cover and Distance to Cover 52
Substrate 53
LIMITING FACTORS FOR SONORA CHUB 54
REPRODUCTIVE RESPONSE TO FLOODING 56 6
MANAGEMENT IMPLICATIONS 57
APPENDIX A: HABITAT TYPES 60
APPENDIX B: LIFE HISTORY INFORMATION 61
APPENDIX C: AIR AND WATER TEMPERATURES 66
APPENDIX D: DESCRIPTION OF MICROHABITAT SITES 69
APPENDIX E: RESULTS OF STATISTICAL COMPARISONS 72
REFERENCES CITED 79 7 LIST OF FIGURES
Figure 1. Sycamore Creek, Santa Cruz County, Arizona 13
Figure 2. Simplified map of a typical study site, depicting method for sampling available microhabitat. 19
Figure 3. Diagram of microhabitat sampling point 20
Figure 4. Monthly averages of air temperature and precipitation from vicinity of Sycamore Creek 25
Figure 5. Relative distributions of velocity at used and available (used + unused) sampling points 34
Figure 6. Relative distributions of mean depth of water column at used and avail able (used + unused) sampling points 35
Figure 7. Relative distributions of position in water column by different life stages of Sonora chub 37
Figure 8. Relative distributions of depth range at used and available (used + unused) sampling points 38
Figure 9. Relative distributions of shade at used and available (used + unused) sampling points 39
Figure 10. Relative distributions of instream cover at used and available (used + unused) sampling points 41
Figure 11. Relative distributions of distance to cover at used and available (used + unused) sampling points 42
Figure 12. Relative distributions of maximum substrate size at used and available (used + unused) sampling points 44 8
LIST OF TABLES
Table 1. Substrate classification (adapted from Bovee 1986) 16
Table 2. Loss of surface water area in Sycamore Creek during the summer, 1991 27
Table 3. Estimated loss of fish in Sycamore Creek during the summer, 1991 27
Table 4. Mean numbers of fish per macrohabitat in each macrohabitat category in May and July, 1991 28
Table 5. Mean macrohabitat characteristics of Sycamore Creek, in May 1991 30
Table 6. Means of selected macrohabitat characteristics of Sycamore Creek, in July 1991. Decrease is relative to depths in May 30
Table 7. Relative frequency of dominant substrates in July 1991 31
Table 8. Relative frequency of floating cover in July 1990 31 9
ABSTRACT
The Sonora chub (Gila ditaenia) is a small minnow that is federally-listed as threatened.
My research objectives were to quantify characteristics and persistence of macrohabitats used by this species through critical summer periods, and to quantify microhabitat selection in Sycamore
Creek, Arizona. By the end of the summer drought, macrohabitats containing adults were deeper, larger and decreased less rapidly than areas with only immature fish or unoccupied areas.
Loss of surface area was highest in inundated unoccupied areas and areas with only immature fish. Loss of Sonora chub from diying pools was highest in pools with immature fish. Ephemeral and unoccupied areas had higher percentages of floating algae and coarser substrates than persistent, occupied areas. General microhabitat characteristics selected by Sonora chub were areas with bedrock or fine substrates near cover and zero velocity. Microhabitat use usually differed among life stages, and availability and selection varied among pools. 10
INTRODUCTION
The Sonora chub (Gila ditaenia) is a small minnow endemic to the Rio de la Concepcion
basin. In Mexico, Sonora chub are widely distributed in the Rio Magdalena and Rio Altar
(Hendrickson and Ju&rez-Romero 1990). In the U.S., Sonora chub are found only in Sycamore
Canyon, Santa Cruz County, Arizona. Sycamore Creek is the headwaters of the Rio Altar; however, surface water is rarely continuous to the border, and the U.S. population is geographi cally isolated. The Sonora chub is federally listed as threatened (USFWS 1986) and state listed as endangered (AGFD 1988). The final rule for federal listing (USFWS 1986) designated critical habitat in Sycamore Creek watershed.
Little is known of the habitat requirements of Sonora chub, and published papers contain only qualitative accounts (Miller 1945, Branson et al. 1960, Minckley 1973, Hendrickson and
Judrez-Romero 1990). Habitat information on Sonora chub is needed to evaluate the status of the habitat in Sycamore Creek over time; to predict success of potential translocations; and to determine how to improve habitat
Sycamore Creek is an interrupted stream during low-flow years, consisting primarily of isolated pools. Surface water is continuous only after heavy rainfall, and even then the lower reaches may remain diy. Habitat conditions appear most limiting in May and June, when temperatures are highest, and precipitation and surface flow are lowest Therefore, information on characteristics of summer habitat is essential to understanding habitat requirements of this species.
Density of a target species is commonly used as a direct measure of habitat quality, although its use is currently under debate (Van Home 1983, Fagen 1988, Hobbs and Hanley
1990). However, strict regulations on handling Sonora chub precluded estimating density with depletion methods (e.g., electrofishing and seining), and snorkeling observations were impossible
because Sonora chub were strongly attracted to submerged observers. Using density as a
measure of habitat quality may not be an appropriate measure in a stream such as Sycamore
Creek; as the stream shrinks, larger numbers of fish are forced into the remaining inundated
areas. These areas are usually isolated from each other (with little or no opportunity for
movement of fish), and habitat quality undoubtedly continues to decrease as evaporation rates
increase. Therefore I defined high-quality habitats as areas that supported a variety of life stages.
This is similar to Van Home's (1986) definition as "those in which a species survives and
reproduces relatively well".
Wiens et al. (1987) recommend analyzing habitat use with a nested hierarchy of scales. I developed a methodology of using shoreline observations to investigate patterns of habitat use at
two different spatial scales: among stream reaches (macrohabitat) and within stream reaches
(microhabitat). These scales are analogous to the "local" and "individual" spatial scales described by Wiens et al. (1987). For the local scale analysis, distinct stream reaches were evaluated by relating presence of different life-stages of Sonora chub to habitat characteristics. Thus, presence of adults (in breeding colors), as opposed to density, was the index of macrohabitat quality.
Choosing adults with breeding colors as an indicator of habitat quality assumes that these fish will select macrohabitats best suited to satisfy their life requirements, as well as the requirements for eggs and fry. In addition, preliminary surveys suggested that macrohabitats with adults usually also contained juvenile and subadult fish. Microhabitat was evaluated by taking physical habitat measurements at observed points of use. 12 OBJECTIVES
This study had two objectives: 1) quantitatively describe habitat characteristics and persistence of macrohabitats used by different life stages of Sonora chub through critical summer periods; and 2) quantify microhabitat use and availability in Sycamore Creek.
STUDY AREA
Sycamore Creek is in the Pajarito Mountains in Santa Cruz County, Arizona, 40 km west of Nogales. Sycamore Creek Canyon is only about 11 km along the thalweg, from its beginning just above Ruby Road to the international border. Data were collected from just below Hank &
Yank Spring at 1220 m to the Mexican border at 1035 m (Figure 1). The vegetation changes from oak woodland-oak grassland at the headwaters to mesquite grassIand-Sonoran desert scrub at the international boundary (Toolin et al. 1980).
Sycamore Canyon forms a natural corridor from Mexico, and contains a diverse assem blage of Mexican flora and fauna (Miller 1945, Miller 1949, Goodding 1961, Mills 1977, Toolin et al. 1980, Bowers and McLaughlin 1982). Consequently, the U.S. Forest Service (USFS) established Sycamore Canyon as a Research Natural Area by 1988. 13
Yanks Spring Site 5 Locations of Site 13 Temperature Recorder, Goodding 1990 Research Natural .Site 24 Site 22
Location of Temperature Recorder, 1991 Site 46
Site 48 J
Site 66
_rc
ARIZONA
MEXICO
Figure 1. Sycamore Creek, Santa Cruz County, Arizona. Circles indicate microhabitat sampling sites. Black dots indicate locations of temperature recorder. 14 METHODS
Habitat use by Sonora chub was measured at the microhabitat and macrohabitat level.
"Microhabitat" is defined as the site where a fish is located at any point in time (Baltz 1990).
"Macrohabitat" is a distinct type of stream reach, such as a pool, riffle, or run. Most macro- habitats were described using Bisson et al. (1982); habitat types occurring in Sycamore Creek but not described in Bisson et al. (1982) are defined in Appendix A.
CLIMATE AND WATER TEMPERATURE
Long-term air temperature and precipitation data were taken from Sellers and Hill
(1974). Daily maximum and minimum air temperatures were obtained from the Nogales weather station (OSC 1990 and 1991). Daily precipitation was collected from June 1990 to September
1991 by Bob Anderson, the campground host at nearby White Rock Campground.
On 30 November 1989, a Ryan Tempmentor was installed in a pool occupied by Sonora chub; it collected hourly water temperature until 20 August 1990. Major flooding in July and
August 1990 filled this pool with sand, so Jerry Stefferud (USFS Fishery Biologist, Tonto National
Forest) moved the recorder to another occupied pool farther downstream on 20 August The recorder stopped on 5 May 1991, was replaced on 12 July, and removed on 30 September. Daily air temperatures from Nogales were correlated to average daily water temperatures with a linear regression.
Temperatures tolerated by Sonora chub were measured by systematically locating the warmest occupied areas during the hottest part of the day (1100 to 1500 hours). MACROHABITAT ANALYSIS
General hypotheses for the macrohabitat research were: 1) habitat variables differ between discrete macrohabitats where Sonora chub persist throughout the summer and those where Sonora chub do not persist throughout the summer, and 2) habitat variables differ among macrohabitats that'
A) support adult Sonora chub,
B) support only immature Sonora chub (subadults alone or in combination with
juveniles),
C) support only juveniles, and
D) areas that do not contain fish.
Macrohabitat Data Collection
The study of macrohabitats was conducted from 11 May to 30 June 1990 and from 16
May to 7 July 1991. Macrohabitat characteristics quantified were total length, maximum depth, maximum width, azimuth, and maximum arc between canyon walls; and visual estimates of percent of instream cover (in quartiles), percent of light penetrating overhead cover (estimated only in 1990), percent of floating cover (in quartiles), and dominant substrate size (Table 1).
Minimum numbers of fish were estimated visually for each 2.5-cm size class.
Repeatedly measuring all inundated areas in Sycamore Creek indicated which habitats persisted as well as changes in numbers and life-stages of Sonora chub present. It also allowed me to define the physical characteristics of persistent habitats. The summer rainy season began just as the last survey was finished in 1990 and just before the last survey was completed in 1991.
Therefore, measurements quantify habitat of Sonora chub at the lowest water levels of both years. 16
Table 1. Substrate classification (adapted from Bovee 1986).
Micro- Macro- habitat habitat Description Size (mm) Size (in) Code Code
Silt 0-2 0-0.08 0 Silt/Fines Sand 2-4 0.08 - 0.16 1 Sand Very fine gravel 4-8 0.16 - 0.3 2 Fine gravel 8-16 0.3 - 0.6 3 Gravel Medium gravel 16-32 0.6 - 1.3 4 Coarse gravel 32-64 1.3 - 2.5 5 Veiy coarse gravel 64-128 2.5-5 6 Rubble Small cobbles 128 - 256 5 - 10 7 Large cobbles 256 - 512 10-20 8 Boulders Small boulders 512 -1024 20-40 9 Medium boulders 1024 - 2048 40-80 10 Large boulders > 2048 > 80 11 Bedrock 12 Bedrock Plant Material 13
Fish were observed from shore. Surface glare was minimized by selecting elevated observation points (e.g., boulders) and using polarized lenses. Total lengths (TL in 2.5-cm size classes) of fish were visually estimated from shore. A sample of 5 to 10 fish were captured prior to surface observations to verify visual size estimates. There was a recognizable cohort of fish under 2.5 cm TL. These fish are referred to as juveniles. Subadults ranged from 2.5-5.0 cm TL; they had all the markings of adults except breeding colors. This definition of subadults follows that by Trautman (1981) as "similar to an adult and approaching adulthood in age and size, but still incapable of breeding". Subadults alone or in combination with juveniles are referred to as
"immature". The smallest fish observed with breeding colors was 5.8 cm TL. Therefore, I considered all Sonora chub over 5.0 cm to be potentially breeding adults. Macrohabitat Data Analysis
Each macrohabitat was classified according to the observed life stage it contained: any
macrohabitat that contained adults with breeding colors was in Category A. Category B
contained immature fish (either subadults alone or with juveniles), and Category C contained only
juveniles. Category D were areas with no fish.
A number of pools that were large in May reduced to a series of small isolated pools by
July. This pattern of drying greatly complicated the analysis. Eventually I omitted these areas
from the macrohabitat study. However, they made up only 5% and 8% of the total surface area
in May and July, and contained 6% and 11% of the total number of fish in these months.
Habitat loss was analyzed based on area with surface water measured in May and July
1991 to determine if there was differential drying relative to habitat category. Fish loss was
estimated only by considering stream reaches that were already isolated at the time of the initial
survey in May 1991 and that dried up by the final survey in July 1991.
G-tests were used for all contingency tables of categorical data (substrate, instream cover,
floating cover), and Kruskal-Wallis tests were used for continuous numerical data (Zar 1984).
Percent values (decrease in depth, light penetrating overhead cover) were arcsine transformed
before analysis. A Tukey studentized range was used to test for differences in macrohabitat
characteristics between habitat categories. Azimuth was tested with multiresponse permutation
procedures (MRPP). MRPP are based on permutations of the data, and make no assumptions about the distribution of the data (Biondini et al. 1988). MRPP are fairly recent distribution-free approaches successfully applied in biology (Zimmerman et al. 1985, Cade and Hoffman 1990,
Burger et al. 1991). MRPP test statistics based on euclidean distances were chosen because of their greater power with non-normal, or skewed data (Biondini et al. 1988). I used MRPP to compare the average within-group (Euclidean) distances between the groups of interest (e.g., azimuths of areas that persisted through the summer and those that did not) to determine if the 18 dispersion and/or location (median) are significantly different Statistical analysis was conducted using BLOSSOM software (Version 1290) available from the U.S. Fish and Wildlife Service
(USFWS) National Ecology Research Center in Fort Collins, Colorado.
MICROHABITAT ANALYSIS
General hypotheses for study of microhabitat were: 1) Sonora chub select specific patterns of individual microhabitat variables; 2) use of microhabitat differs among life stages of
Sonora chub.
Microhabitat Data Collection
Data from macrohabitat measurements were used to stratify the available habitats and identify discrete units for microhabitat sampling using a randomized block design. Sycamore
Creek was divided into four sections based on geomorphology and extent of continuous water.
Two sites from each of the four stream sections were randomly chosen as sites for the micro habitat study (Figure 1).
Microhabitat data were collected twice at each site (subsamples were separated by at least 2 days) from 13 June to 27 July 1991. Prior to data collection, each site was sketched on grid paper. Fifty points were located on each map by using a random numbers table to select x and y coordinates (Figure 2). These points were located in the creek and marked by referencing nearby instream features or by carefully placing pebbles for reference. The minimum distance between two points was 15 cm.
A fish was considered to occupy a sampling point if it occurred in an imaginaiy cylinder with a diameter of IS cm and extending throughout the water column, with a sampling point at the center (Figure 3). Each sampling point was scanned for IS seconds, and the numbers and 19
351
30
25
20 (J)
's 15
10
1 r—J 1 I 1 | 1 1—I 1 1 r—i 1 1 1 I 1 I 1 0 5 10 15 20 25 30 x axis
Figure 2. Simplified map of a typical study site, depicting method for sampling available microhabitat. Black dots indicate sampling points determined by random x and y coordinates (grid not shown). 20 N
Metal Washer
7.5 cm
7.5 cm
s
Figure 3. Diagram of microhabital sampling point. Five measurements of depth and substrate were taken: at the center, and 7.5 cm from the center in each ordinal direction. Black dots indicate measurement locations. Figure not drawn to scale. 21 life-stages of fish observed were recorded. These measures are "instantaneous use"; the use by fish of each of the 50 points was recorded for a given instant (IS seconds) in time. Position in the water column (surface, middle, or bottom) was estimated for each life stage. If fish in the same life stage occurred at more than one position, depth occupation was called "dispersed".
Before proceeding to the next sampling point, I also noted whether the sampling point was in shade or in sunlight, and if filamentous algae were present
After ten sampling points were observed, the central locations of each sampling point were marked with a metal washer numbered with permanent ink. Sampling points were observed and marked ten at a time to ensure that the markers were placed as accurately as possible to the actual observed sampling point A 2-meter pvc pipe with an L-shaped hook on one end was used to place washers with minimal disturbance to the fish and the substrate. This process continued until all sampling points were observed and marked.
Physical measurements were made immediately after completing all fish observations.
Depth and substrate were measured at the point marked by the metal washer, and at four ordinal points 7.5 cm from the center (Figure 3). The measurements provided an estimate of variability of depth and substrate around each sampling point Velocity, instream cover and distance to instream cover were measured from the center of the metal washer. Velocity was measured with a Marsh-McBirney model 201D current meter, accurate to 0.01 m/s. Because velocity was so low in many areas, a velocity category of "trace" was added, which was defined as visible but not measurable current
Substrate data were collected using a modified Brusven index as adapted by Bovee
(1986) (Table 1). The dominant and subdominant substrate categories were recorded, using visual estimates in percent quartiles.
Instream cover was defined as objects in the water. Instream cover types included rock crevices, rock walls, algae, and plant material. Instream cover was defined as extending 90 cm 22
from the nearest instream object The minimum rock size that could provide instream cover was
defined to be cobbles measuring at least 10 cm in diameter.
Microhabitat Data Analysis
Seven variables were used to describe microhabitat at each point: water velocity
(measurable velocity occurred in only two of the eight sites); mean depth of water column; range
of depth (an index of depth heterogeneity); instream cover and distance to cover; shade; and
maximum substrate size. All sampling points represent available microhabitat All sampling
points where fish were observed represent microhabitat used by fish. As recommended by Bovee
(1986), each sampling point used by fish was assigned a frequency of one, regardless of the
number of fish observed at that point
Null Statistical Hypotheses:
1) Distribution of available microhabitat is similar to microhabitat used by all life stages combined for a given habitat variable
2) Distribution of available microhabitat is similar to use by each life stage for a given habitat variable
3) Distribution of used microhabitat is similar among life stages for a given habitat variable (e.g., juvenile use = adult use)
Distribution-free, multiresponse permutation procedures (MRPP) were used to test if areas used by Sonora chub were different from available areas, and if areas used by one life stage were different from areas used by another life stage. The test statistic is sensitive to differences in median and dispersion, so data were plotted in frequency histograms to determine what characterized the significant differences in distribution. Water levels in Sycamore Creek changed rapidly, and changes were noticeable between sampling periods. Thus, it was necessary to determine whether use of habitat characteristics differed between subsamples (the two samples taken at one site). If the distribution of used microhabitat differed between subsamples by MRPP P-values greater than 0.40 for any variable, the data were pooled for analysis.
To compare distribution of used microhabitat between life stages, only non-inclusive data were used. In other words, in comparisons between adults and subadults, data from sampling points that contained both life stages were omitted from the analysis.
LIFE HISTORY NOTES
A literature review combined with my field observations on food habits and reproduction are included in Appendix B. 24
RESULTS
CLIMATE AND WATER TEMPERATURE
Sycamore Creek has mild winters and hot summers (Sellers and Hill 1974). Most of the yearly precipitation occurs in July and August (Figure 4). Air temperatures were higher than normal from March to June 1990 (Figure 4), but were closer to long-term averages in 1991.
Rainfall patterns were quite different over the 2 years. In July 1990, precipitation was nearly twice the long-term average for July. In 1991, precipitation was high in spring, but the summer rainy season was late and produced only about half as much rain as the long-term average
(Figure 4).
Average daily water temperatures were calculated from the hourly water temperature data recorded from Sycamore Creek. Daily air temperatures (maximum and minimum) were obtained from Nogales. Graphs of daily air and water temperatures are included in Appendix C.
Water temperatures follow Nogales air temperatures closely (r2 = 0.817).
The maximum water temperature recorded by the Ryan Tempmentor was 29.7° C on 28
June 1990; the minimum temperature was -2.4° C on 23 December 1990. In 1990, the maximum temperature measured with a field thermometer at a Sonora chub location was 37° C. Water temperatures increased until the summer rains began, then dropped and stabilized, ranging from
20° C to 25° C from July to mid-September. Overall, water temperatures in 1991 were lower than in 1990. The pools that contained the temperature recorder were shaded and persisted during both years. Therefore, temperatures reported were probably lower than in more exposed pools. Air Temperature
JAN FEB MAR APR MAY JUN
Precipitation
Nogales Ruby Bear Valley Arivaca 1966-72 1935-44 1943-56 1966-72
Figure 4. Monthly averages of air temperature and precipitation from vicinity of Sycamore Creek. Long-term climate data from Sellers and Hill (1974). Air temperature data for 1990 (—) and 1991 (--) from Nogales (OSC 1990 and 1991). Precipitation data for 1990 (••-•••) and 1991 ( ) collected at White Rock Campground. HABITAT LOSS
In 1991, at the end of the diy summer period, almost all macrohabitats that persisted were pools. In 1991, 65% of the areas occupied by fish in May had dried up by July (Table 2).
Habitats that contained adult fish (Categoiy A) had the largest area in May (46%); only 19% of the surface area of these sites dried up by July. Habitats supporting immature fish (Category B) made up 24% of total surface water in May but by July 70% of this area had dried up. Habitats supporting only juveniles (Categoiy C) represented 9% of available surface water in May.
However, 74% of the surface water at these sites dried up by July. Ninety-three percent of the surface water that did not support fish in May (Category D) did not persist by July.
FISH LOSS
At least 20% of the total numbers of Sonora chub estimated in May were in isolated pools that dried up by July (Table 3). However, only 12% of the fish in habitats with adults
(Category A) were lost when areas dried up. In contrast, 31% of immature fish (Categoiy B) and 54% of juvenile fish (Categoiy C) were lost when areas dried up.
MACROHABITAT ANALYSIS
Changes in Abundance
Habitats containing adult fish had higher fish abundance than habitats containing only immature or juvenile populations (Table 4). This trend was consistent in both May and July
1991. Also, mean numbers of fish increased from May to July in all habitat categories. 27
Table 2. Loss of surface water area in Sycamore Creek during the summer, 1991.
%of Total Total Loss of Loss as surface surface total % of May area area surface total by in May in May area categoiy (m2) % (m2) %
Habitats supporting:
• Adult fish (A) 89000 46.4 17054 19.2
• Immature fish (B) 49351 23.7 34377 69.7
• Juvenile fish only (C) 16515 8.6 12251 74.2
Areas with no fish (D) 37092 19.3 34790 93.8
Totals 191958 100
Table 3. Estimated loss of fish in Sycamore Creek during the summer, 1991.
Estimated Estimated %of number %of number total lost of fish total of fish from May in May in May lost to July
Habitats supporting:
• Adult fish (A) 21604 71 2614 12
• Immature fish (B) 4895 16 1493 31
• Juvenile fish only (C) 4074 13 2183 54
Totals 30573 100 6290 28
Table 4. Mean numbers of fish per macrohabitat in each macrohabitat category in May and July, 1991. The number of macrohabitats in each category (N) are in parentheses.
Habitats Supporting: May July
• Adult fish (A) 144 252 (150) (89)
• Immature fish (B) 49 86 (101) (43)
• Juvenile fish only (C) 61 71 (67) (7)
Tukey Test (P <; 0.05): A > B,C A > B,C
Comparisons of Macrohabitat Variables
There were no significant differences among macrohabitat categories (macrohabitats with adult fish, immature fish, juvenile fish only, or no fish) for the following variables: percent of instream cover (1990 data only, not measured in 1991), percent of light penetrating overhead canopy (1990 data only, not measured in 1991), azimuth, and maximum arc between canyon walls. Floating cover (1990), substrate (1991), maximum depth, surface area, and depth decrease were significantly different (P < 0.05) among macrohabitat categories. 29
Depth. Width. Surface Area
In May, habitats that contained adult fish were deeper and wider than all other habitats.
They also had greater surface areas than habitats that contained no fish or only juveniles (Table
5). Habitats occupied by immature fish had greater surface area than those that contained only juveniles and deeper than areas that contained no fish. There were no significant differences in size or depth between habitats that contained only juveniles versus areas that contained no fish.
Habitats that supported adults at the end of the summer were the deepest and largest habitats (Table 6). Habitats containing only juveniles were statistically similar to unoccupied areas. Habitats that supported immature fish in July were also similar in depth and surface area to unoccupied areas. Furthermore, some of the habitats that supported adults in May supported only immature fish in July (adults either succumbed or moved out of these areas). The depth de crease in these particular areas was similar to that in areas that contained no fish in July, and was significantly higher than sites in Category A in both May and July.
Dominant Substrate
Data from 1990 indicated no differences in substrate between categories. In 1991, dominant substrate was significantly different (P < 0.05) between areas that supported Sonora chub throughout the summer and those that did not (Table 7). Habitats with bedrock-sand substrates supported Sonora chub throughout the summer more frequently than expected, and habitats with rubble/gravel substrates supported Sonora chub less frequently than expected.
Floating Cover
In 1990, habitats where Sonora chub survived had lesser amounts of floating cover than areas that dried up or were uninhabited by fish (Table 8). Seventy-five percent of the 30
Table 5. Mean macrohabitat characteristics of Sycamore Creek, in May 1991. The number of macrohabitats in each categoiy (N) are in parentheses.
Maximum Maximum Maximum surface depth width (m) area (m2)
Habitats supporting: • Adult fish (A) 0.49 3.57 55.12 (N=150) • Immature fish (B) 0.30 2.60 44.95 (N=102) • Juvenile fish only (C) 0.25 1.93 22.56 (N=68) Areas with no fish (D) 0.22 2.04 43.07 (N=79)
Tukey Test: A>B,C,D A>B>C,D A>C,D (P a 0.05) B>D B>C
Table 6. Means of selected macrohabitat characteristics of Sycamore Creek, in July 1991. Decrease is relative to depths in May. The number of macrohabitats in each category (N) are in parentheses.
Maximum Decrease in Maximum Maximum Surface Maximum Depth Width (m) Area (m2) Depth (%)
Habitats Supporting: • Adult fish (A) (N=94) 0.37 2.67 29.89 19.6 • Immature fish (B) (N=49) 0.19 1.92 14.84 37.1 • Juvenile fish only (C) (N=7) 0.14 1.32 8.39 25.1 Areas with No Fish (D) (N=17) 0.15 1.79 7.34 45.9
Tukey Test (P s 0.05): A>B,C,D A>B,C,D A>B,C,D D,B>A 31
Table 7. Relative frequency of dominant substrates in July 1991. The number of macrohabitats (N) are in parentheses.
Bedrock/ Rubble/ Gravel/ sand Boulders gravel sand
Persistent habitats with Sonora chub 0.26 0.18 0.10 0.46 (34) (24) (13) (60)
Ephemeral or unoccupied 0.10 0.26 0.24 0.40 habitats (26) (65) (61) (102)
G = 25.5, df = 3, P < 0.0001 H0: Substrate is independent of habitat persistence
Table 8. Relative frequency of floating cover in July 1990. The number of macrohabitats (N) are in parentheses.
Percent of Floating Cover
0-25% 25-50% 50-75% 75-100%
Persistent habitats with Sonora chub 0.50 0.25 0.20 0.05 (28) (14) (11) (3) Ephemeral or unoccupied areas 0.39 0.17 0.22 0.21 (43) (19) (24) (23)
G = 9.06, df=3, 0.05 > P > 0.025 H0: Floating cover is independent of habitat persistence 32 habitats that supported Sonora chub contained less than 50% floating cover. In addition, only
5% of the habitats that supported Sonora chub throughout the summer had cover over 75% of the water surface. In 1991, there was little floating cover in Sycamore Creek and this variable was not analyzed for that year.
MICROHABITAT ANALYSIS
Definitions of microhabitat use, availability, and selection were based on Johnson (1980).
"Availability" is the distribution of a given habitat variable as represented by all sampling points in a site. "Use" is the distribution of a given habitat variable at all sampling points that were occupied by Sonora chub during the observation period; "use" may be for all life stages combined, or for individual life stages of Sonora chub. "Selection" indicates that a variable was used disproportionate (P<0.05) to availability.
Sonora chub selected areas with specific microhabitat characteristics (see Appendix D for detailed habitat characteristics of sites): differences in distributions occurred between areas used and areas available for most microhabitat variables in at least one site. The MRPP analyses tested differences in distribution of actual frequencies; however, for ease of interpretation, figures show distribution of relative frequencies. Graphs provide the sample size for each comparison: pooled data have an available sample size of 100, while subsamples have 50 locations. Sample sizes for data on areas used by fish do not represent number of fish, but number of sampling locations where fish were observed. Any available location could be used by more than one life stage; thus, summations of sample sizes of different life stages will not be equal to the sample size of locations used by all fish. Although differences between use by fish of different life stages are discussed, there are no graphs depicting these differences; statistical results of all microhabitat comparisons are in Appendix E. 33
The results includes all comparisons significant at the 0.05 level; however, there may be dependence of habitat use among size classes within the same pool. All three age classes were in six of the eight pools; only Site 13 and Site 48 had no juveniles. Therefore, most pools had four potentially dependent comparisons: use by all ages combined, use by juveniles, use by subadults, and use by adults. Thus, a P-value of 0.013 is equal to a Bonferroni simultaneous probability of
0.05 (Miller 1966), and P-values greater than 0.013 may need to be treated with caution.
Velocity
Only two of the eight sites (Site 5 and 13) had flowing water when microhabitat was measured, and Sonora chub used velocity in proportion to availability in Site 13. Sonora chub used areas with velocities ranging from 0 to 3.4 cm/s. In Site 5, adults and juveniles selected areas with no velocity (Figure 5; P=0.003), but subadults used velocities roughly in proportion to availability. When all life stages were combined, selection was towards areas with no velocity
(Site 5; P=0.027).
Mean Depth of Water Column
There was no significant selection for mean depth of water column by all life stages combined. However, in several sites specific life stages of fish used specific ranges of depth
(Figure 6). Adults primarily selected areas that were more than 10 cm deep (Site 24, P=0.043).
Subadults selected depths of 11 to 30 cm, and were found less frequently in shallower areas
(Site 48, P=0.043; Site 73,P=0.024). Juveniles selected shallow areas under 20 cm deep (Site 46,
P=0.007; Site 66, P<0.001). For example, in Site 66, juveniles were not observed in areas more than 20 cm deep, although over 38% of all available locations were deeper than 20 cm.
In pools where depth selection was determined, there was a direct relationship between fish size and depth occupied. Juveniles used areas that were shallower than those used by 34
o Site 5, Pooled: P = 0.003 100% B Available (N=100) £3 Use by Adults (N=20) O 80% - 1 H Use by Juveniles (N=20) Q_ • Use by Subadults (N=29) *NS
2 60% - CL 1 40% - CD > I — 20% - +-» CO — mm®?®*. J 0% 1 I. 0 trace - 3 >3 cc Velocity (cm/s)
Figure 5. Relative distributions of velocity at used and available (used + unused) sampling points. Site 24, Subsample 2: P = 0.043 60% 5 50%- •flso"by'X(Sjlts §J»23)
40% I 1 # 30%- ll pi 20%- *55 8 C 10%- 1 P O 0% 1 1 p i i i r i ' 10 20 30 40 SO 60 70 80
Site 48, Pooled: P = 0.043 Site 73, Subsample 2: P = 0.024 EJAvaBabte (N*=50) 50% - 0U»« by 8ubadutt$ (N=11)
40% r
30%-
20%-
10%-
20 30 40 50 60 70 80
CO Site 46, Pooled: P = 0.007 Site 66, Subsample 1: P < 0.001 0 60% 3Available (N-100) HAvaSable (N»50) rr 50% §iHUsa by Juveniles (N=40) •Use byJuvenfle# (N=11) R\ 40%- \ 40% - R 30%- 30% - I 20%- IN 20%- s v 10%- is 0% So K JS}_ isirfs 1 g; 10 20 30 40 50 60 70 80 Mean Depth of Water Column (cm)
Figure 6. Relative distributions of mean depth of water column at used and available (used + unused) sampling points. Values on X-axis indicate maximum end of sampling interval. 36 subadults (Site 5, P=0.009) and adults (Site 5, PcO.OOOl; Site 46, PcO.OOOl). Subadults used areas that were shallower than those used by adults (Site 5, P=0.003; Site 46, P=0.0001).
Position in Water Column
Adults were generally dispersed within the water column (Figure 7); they were seldom observed near the water surface. In Site 22, juveniles (P=0.030) and subadults (P=0.032) were more frequently found on the surface or on the bottom than adults. In Site 24, juveniles used the surface and mid-column (PcO.OOOl) and subadults used the surface (P=0.002) more than adults; adults more frequently used the bottom or were dispersed. In Site 22, adults used the mid- column more than subadults (P=0.032); whereas in Site 73, adults used the bottom more and the surface less than subadults (P=0.040). In Site 66, juveniles were mostly near the surface, while subadults were at mid-depths (P=0.010). In this site, adults used the same depths as subadults; these depths were significantly different from the relative depths used by juveniles (P=0.005).
Depth Heterogeneity
In general, Sonora chub did not show selection for depth heterogeneity. In pools where
Sonora chub selected depth heterogeneity, the preferred range in depth was between 2 and 4 cm
(Figure 8). Similar distributions were observed for all fish, as well as all life stages. However, subadults used a larger range of depth than juveniles in Site 46 (P=0.037).
Shade
There was no clear pattern in the selection of shade; however, sunlit areas were selected more frequently than shaded areas (Figure 9). Sonora chub selected areas with no shade in the morning in Site 22 (P=0.040) and in the afternoon in Site 73 (P=0.016). Adults (Site 22, 37 Site 22, Pooled: P = 0.030 Site 22, Pooled: P = 0.032 100% -T 100% •Juveniles (N=32) 0Subadults (N=46) 80% - ESAdults (N=69) B0% - EJAdults (N=69)
60% - 60%
40% - 40%- c o 20% - 20% - i 0% 1 0% ^3® mm Surface Middle Bottom Dispersed ISurface Middle 1Bottom Dispersed 1— o Q. Site 24, Pooled: P< 0.0001 Site 24, Pooled: P = 0.002 100% 100% O SJuveniles (N=38) •Subadults (N=44) 80% - •Adults (N=59) 80% - iAdults (N=59)
CL 60% - 60% -
40% - 40% - CD > 20% - 20% - P5 0% _Ei 0% £ •4—' • ISurface 1Middle Bottom Dispersed ISurface Middle Bottom Dispersed co
CD Site66, Pooled: Ps 0.010 Site 73, Pooled: P = 0.040 100% 100% DC •Juven. (N=16) •Subadults (N=31) 80% - [7L_ HSubad. (N=11) 80% HAdults (N=29) /p lAdults (N=9) 60% - '/•/ 60%- / 40% - / 40% / / 0 20% - / 20% - / A 1 0% o% 2 m Surface Middle Bottom Dispersed Surface Middle Bottom Dispersed Position in Water Column
Figure 7. Relative distributions of position in water column by different life stages of Sonora chub. 38 Site 24, Subsample 1: P = 0.007
H Available (N=50) EJ Use by Adults (N«36)
c o
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
o Site 22, Subsample 1: P < 0.028 Q_ ED Available (N=50) • Use by All Fish (N=39) O • Use by Subadults (N=22)
CD >
+-» CO CD Site 22, Subsample 2: P = 0.035 a: 70% U Available (N=50) 60% - H Use by Juveniles (N=21)
50% • *a5gI 40% - im
30%
20% -
10% -
0% § \\ Depth Range (cm)
Figure 8. Relative distributions of depth range at used and available (used + unused) sampling points. Note that Site 24 (top graph) has different x and y axes. 39 Site 22, Subsample 1: P = 0.040 Site 22, Subsample 1: P < 0.026 100% 100% « ^Available (N-50) ^Available (N-50) Use by All Fish (N-39) ElUse by Adults (N-33) 80% - Hi •Use by Juven. (N»11) SHF 60% - 60% •
40% -
20% - C : o 0% El No Shade Shade No Shads Shade
Site 48, Subsample 1: P = 0.029 Site 46, Pooled: P = 0.012 100% ^Available (N-50) ^Available (N=100) Use by Air Fish (N-31) EiUse by Adults (N-23) 80% -
• 60% - •11
40% - 40% - 1
20% - 1 J jWBjp , 0% EMS09S No Shade Shade No Shade Shade CO Site 73, Subsample 1: P = 0.016 Site 73, Subsample 1: P = 0.044 CD 100% [^Available (N»50) •Available (N=50) Use by All Fish (N=31) •Use by Juveniles (N-13) cc 80% - B0% -
60% -
40% -
No Shade Shade No Shade Shade Shade
Figure 9. Relative distributions of shade at used and available (used + unused) sampling points. 40
P=0.013; Site 46, P=0.012) and juveniles (Site 22, P=0.026; Site 73, P=0.044) selected areas with
no shade, while subadults used areas in proportion to availability at all sites. Site 48 was the only
pool where shaded areas were selected: by all fish as well as by individual age classes (PsO.035).
Instream Cover
Fish selected for instream cover at several sites (Figure 10). Rocky cover was selected by
all life stages in Site 73 (P=0.004) and Site 46 (P=0.040), by adults and subadults in Site 66
(PsO.024), and by juveniles in Site 46 (P=0.034). There was little plant cover substantial enough
to provide protection from increased flows at any of the sites. The plant material noted in Figure
10 was primarily fallen leaves, small branches, and submerged roots and aquatic plants. Juveniles
selected plant material in Site 24 (P=0.006) and algae in Site 5 (P=0.040). At Site 46, Sonora
chub selected areas with no cover as well as rock crevices. Adults used rocky crevices more and
plant cover less than subadults (Site 5, P=0.046) and juveniles (Site 24, P=0.043).
Distance to Cover
There was no clear pattern for selection of distance to cover (Figure 11). All life stages
combined selected distances less than 10 cm from cover in Site 66 (P=0.044) and 21 to 40 cm
from cover in Site 73 (P=0.001). Areas less than 20 cm from cover were avoided at Site 46
(P=0.015). Adults selected areas 21 to 50 cm from cover in Site 73 (P=0.020). Subadults selected areas more than 20 cm from cover in Site 46 (P=0.008), and areas less than 30 cm from
cover in Site 73 (P=0.005) and Site 48 (P=0.036). Subadults remained closer to cover than
adults in Site 5 (P=0.0001). 41 Site 46, Pooled: P = 0.040 Site 5, Subsample 2: P = 0.040
•Available (N=50) ish (N-63) •Use by Juveniles (N*>7)
60% -
c o a No Rock Plant No Rock Rock Algae Plant Cover Crevice Material Cover Crevice Wall Material
Site 73, Pooled: P = 0.004 Site 24, Subsample 1: P = 0.006
AvalabM (N»100) EJAvaiabt* (N-50) UwbyAIF)th(N-55) Sum byJuvente* (N»20)
J , PH No Rock Alga* Plant No Rock Rock Atga« Plant Cover Crevice Material Cover Crevice Wall Material
Site 66, Pooled: P < 0.025 Site 46, Pooled: P = 0.034 0 100% / ^Available (N=100) DC / •Use by Adult3 (N=9) 80% / •Use by Subadults (N=11) p" / / 60% / / s / 40% M / 40% - / $ / 20% KTw / / / 0% No Rock Rock Algae Plant No Rock Rock Algae Cover Crevice Wall Material Cover Crevice Wall Type of Instream Cover
Figure 10. Relative distributions of instream cover at used and available (used + unused) sampling points. 42 Site 73, Pooled: P = 0.001 Site 73, Pooled: P = 0.020 50% -| 50% •Available (N-100) •Use by All Fish (N-55) 1 40% - 40% luseby Aciults (??=29)
30% - 30%
20% - 20% c o 10% - 10% 0% 0% 1 III +-» 10 20 30 40 SO 60 70 80 90 10 20 30 40 50 60 70 80 90 1— o Site 66, Subsample 2: P = 0.044 Site 73, Pooled: P = 0.005 Q_ 50% 50% / ^Available (N=100) •Available (N-50) / O 67% •Use by All Fish (N-15) •Use by Subadults (N=31) 40% - 40%- %•/ •§/ 1/ CL 30% - 30% - I'fa/
20% - 20% - v, a) J71 10% - 10% - •/ / > / i/ \/_ 0% H pfi •+—> 0% 14 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 03
Site 46, Pooled: P = 0.015 Site 46, Pooled: P = 0.008 CD 50% r 50% DC ^Available (N-100) •Available (N=100) •Use by All Fish (N-63) •Use by Subadults (N=23) 40% - 40%
30% - 30%-
20% - 20% V\ |7| [71 10% - 10% / / / 1/ I' / / / 1/ / / 1 / 0% 0% / / 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 Distance to Cover (cm)
Figure 11. Relative distributions of distance to cover at used and available (used + unused) sampling points. Values on X-axis indicate maximum end of sampling interval. Maximum Substrate Size
At three of the five sites where bedrock occurred, it was selected by at least one age class. However, bedrock was usually overlain by a thin covering of sand and gravel. In Site 13, all life stages combined selected bedrock (Figure 12; Ps0.044). Adults (Site 22, P=0.0004) and subadults (Site 22, P=0.038; Site 46, P=0.015) selected bedrock when available. In Site 66, which contained no bedrock, use was higher than availability for coarse gravel to small cobbles, and small boulders (P=0.012). In Site S, which contained no bedrock, adults selected fine to coarse gravel (P=0.001). Subadults used much coarser substrates than adults in Site 5 (P=0.004). Site 5, Pooled: P = 0.001 100%
80% -
60% -
40%-
20% -
1 2 3 4 5 6 7 8 9 10 11 12 o Site 13, Subsample 1: P = 0.044 Site 22, Pooled: P £ 0.038 Q. 100% 100% SAvallable (N=50) O •Use by All Fish (N=10) H Available (N=100) 80% - 80% EH Use by Adults (N=69) 0 Use by Subadults (N=46)
60%- 60% -
40%- 40% CD > 20%- 20%
0% i n-n Uti1 1 a •+—> 0% —i—^'Y ' "i ' "i'' 'i "i 1 1 r 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 CO
Site 66, Subsample 2: Ps 0.049 Site 46, Pooled: P = 0.015 0 100% 100% ^Available (N=5Q\ ^Available (N=100) DC •Use by All Fish (N=15) HUso by Subadults (N=23) 80% - •Use by Subadults (N=9) 80%-
60% - 60% -
40% - 40%
20% - 20%
0% i 1 r" 0% 1 1 2 3 4 5 6 7 9 10 11 12 8 9 10 11 Maximum Substrate Size
Figure 12. Relative distributions of maximum substrate size at used and available (used + unused) sampling points. Explanation of substrate codes is in Table 1. 45
DISCUSSION
JUSTIFICATION FOR METHODS
Surface observations worked well in this study because the water in Sycamore Creek was
very clear, there was little overhead cover, and Sonora chub were not frightened by quiet
observers on the bank. This method has been used with variable results in other fish habitat
research (Shirvell and Dungey 1983, Bachman 1984, Baltz et al. 1991, Bozek and Rahel 1991a).
Several authors report that surface observations underestimate fish abundance in deep or
turbulent water (John 1964, Shirvell and Dungey 1983, Heggenes et al. 1990). Deep or turbulent
water was rare in Sycamore Creek during this study. Bozek and Rahel (1991b) compared streamside visual counts and electrofishing estimates of a species of trout They found that visual counts of fry were highly correlated to electrofishing estimates, but that visual counts of adults were poor estimators because adults used overhead cover more than fry. These limitations may
not apply to my study because Sonora chub adults seldom selected overhead cover.
John (1964) surveyed minnows in intermittent streams of the Chiricahua Mountains in southern Arizona; he reported that visual estimates undoubtedly underestimated populations in pools with more than 100 fish. Therefore, my visual estimates probably did underestimate fish density; especially in pools with large numbers of schooling fish. I assumed that the estimates of abundance of Sonora chub have the same level of error in similar-sized pools and thus could be used for comparisons between pools.
I have no reason to suspect that I was unable to determine the presence or absence of a given life stage or the relative depth occupied. The water was consistently clear and Sonora chub did not avoid the observer. Shirvell and Dungey (1983) tested accuracy of estimates of a model's height above the substrate from surface observations; they found no significant differences between actual and estimated heights of the model. There may have been a small percent of 46
cases where I assigned fish to the wrong life stage; however, based on field measurements
verifying my visual estimates, the error was minimal.
Randomly locating sampling points to determine microhabitat use of a mobile animal is
an unusual data collection method, probably because the chances of observing a target animal at
a random point would be veiy low. However, this method provided large sample sizes in
Sycamore Creek because Sonora chub densities in pools were so high and fish generally moved
freely within the pools.
Multiple Response Permutation Procedures (MRPP) were ideal for microhabitat data
collected as described above, where random sampling of resources determines available micro
habitat, and presence or absence of fish determines what portion of microhabitat is used. MRPP
can compare used with available (used + unused) distributions (Cade and Hoffman 1990), by
denoting all of the unused sampling points as an excess group (B. Cade, USFWS, personal
communication).
The macrohabitat analysis included a comparison of mean numbers of fish in different
habitat categories in May and in July 1991 (Table 4). These data illustrated how animal numbers can be a misleading indicator of habitat quality. There was a large increase in mean number of
fish in each habitat category, even though volume—and undoubtedly habitat quality—decreased.
CLIMATE AND WATER TEMPERATURE
The climate for the 2 years of my study were within the range of normal conditions. The
first summer had record high air temperatures and high precipitation during the rainy season, and the second summer had mild temperatures and low rainfall. The linear regression demonstrates that air temperatures in Nogales provide a good estimation of water temperatures in Sycamore
Creek. 47
MACROHABITAT ANALYSIS
The deepest pools that decreased in depth the least over the summer provided the best habitat (most likely to persist and support a variety of life stages) for Sonora chub in Sycamore
Creek. Areas with the least floating cover were also the most suitable; stagnant areas quickly developed thick pads of algae on the surface. Floating cover was not a significant variable in
1991; probably because milder temperatures did not stimulate algal development
Substrate type may influence water quantity. Areas dominated by rubble and gravel substrates persisted less frequently than expected by chance, and areas dominated by bedrock- sand persisted more than expected by chance. Low persistence may be related to the higher porosity and decreased water-holding capacity of the fractured substrates.
Adults selected habitats that were significantly larger and deeper than other areas. The habitats selected by adults persisted at a much higher rate than habitats containing only immature or only juvenile fish. Persistence is clearly related to maximum depth, and it would be adaptive for adult Sonora chub to select pools that do not dry out
By July, depth, surface area, and depth decrease were no longer different between areas supporting only immature fish and unoccupied areas with surface water. These sites dried up at a higher rate than habitats chosen by adults. Therefore, habitat quality for immature fish declined significantly over the summer.
The connectedness of habitats in Sycamore Creek is an important aspect to consider for interpreting the data. It would be inappropriate to assume that any area occupied by Sonora chub can be considered "habitat" if fish can not move to more suitable sites or move out of unsuitable areas. However, large sections of the stream within Sycamore Canyon are connected at least a few times a year following rains. During this 2-year study, I frequently observed Sonora chub moving between connected pools at times when surface flow had increased and also when 48
these areas were about to become isolated. Therefore, I assumed that, at least for adult Sonora
chub, they had freedom to choose a habitable site within a reach at some point in their life.
Habitat connectedness also affects interpretation of loss of numbers of fish. Loss of
habitat did not necessarily mean that all Sonora chub in those habitats were lost; some fish may
have escaped into connected areas. At the same time, fish losses were undoubtedly underesti
mated, since other ephemeral, occupied habitats could have become isolated between the initial
and subsequent surveys. Ninety-four percent of all areas with no fish dried up. Of these
unoccupied areas, 70% were isolated—sometimes by over 20 m of dry streambed—from
occupied, persistent habitats. Thus, many of these unoccupied areas may have become isolated
before I conducted the first surveys; these areas may have filled up with water during spring
rains. However, 30% of the unoccupied areas were connected to sites with fish. Even though
the variables we measured showed no difference between areas with no fish and areas with
juveniles only, the juveniles may be detecting (at least sometimes) environmental cues that allow
them to avoid the most ephemeral areas.
It is also possible that the fish occupying ephemeral pools have little choice. These
marginal areas may be dispersal sinks for subdominant animals (Van Home 1983). In other
words, juvenile and subadult Sonora chub may be forced out of better areas by competition or
intraspecific predation. Power and Matthews (1983) provide experimental evidence that smaller
fishes are limited to shallow areas because predaceous centrarchids change abundance and spatial
distribution of their prey. Although there are no other fish species besides Sonora chub in
Sycamore Creek, intraspecific predation is possible. Angermeier (1987) hypothesizes that
cyprinids should exhibit less avoidance of pools as they grow, because their risk of predation decreases with increasing size. It is also possible that younger fish are not as good as adult fish at detecting that a given area will dry up. Water levels changed rapidly in Sycamore Creek; channels between pools evaporated in a matter of hours. Thus, it is easy to foresee how juveniles 49
in side pools and channel edges can become isolated from the main sections of Sycamore Creek.
John (1964) reported that all cyprinids in temporary sections of a stream in the Chiricahua
Mountains were one-year-olds, whereas fish in permanent sections represented several age
classes. Schlosser (1982) reported that shallow, temporally variable habitats contained mostly age
0-1 cyprinids in the headwaters of a warmwater stream. Schlosser (1987) provides a conceptual
framework explaining how young-of-year cyprinids colonize small shallow pools with low habitat
heterogeneity, and how large pools provide refuge from harsh low flow conditions in summer.
Qualitative information is available for habitat used by Sonora chub in Mexico. Descrip
tions are briefly outlined in Appendix B. In Mexico, Sonora chub were generally found in pools
under 2 feet deep, near areas with a fairly swift current, over sand and gravel substrates. Sonora
chub were found near cover such as fallen logs, aquatic vegetation, and undercut roots. Results
from Sycamore Creek primarily differ in terms of selection for current and propensity for large
objects for cover. However, these differences may be due to the lack of availability of these
habitat characteristics rather than actual differences in habitat use.
MICROHABITAT ANALYSIS
Sonora chub generally selected slow-moving water, rocky cover, deeper water as they
grew larger, and bedrock if it was available; otherwise, sand and fine gravel were used. There
was no clear pattern for shade and distance to cover.
Sonora chub did not select the same habitat variables (no significant difference between
use and availability) in all eight sites. Lack of significance may be attributable to biological
reasons (e.g., the variable was not a significant factor in describing habitat use in that site); or, statistical reasons (e.g., the sample size was not large enough to detect a difference).
Comparisons between use of microhabitat between life stages must be treated with caution since the sampling protocol was not designed to address specifically these differences. 50
However, the fact that many comparisons between life stages were similar between pools
strengthens the likelihood that the observed patterns were real. Helfman (1978) describes how
most species show ontogenetic differences in habitat preference.
Significance of Microhabitat Variables
Velocity
Three of the seven comparisons of velocity were significant at P < 0.03 (in Site 5: selection by all sizes combined, juveniles, and adults); it is unlikely that this distribution occurred
by chance. Few sites occupied by Sonora chub had observable velocity. In addition, higher velocities occurred in areas where the chances of observing use were rare: cascades, slides, and
riffles between pools. These areas are not frequently used by Sonora chub not just because of increased velocity, but also because these areas are veiy shallow (primarily under 6 cm), have no cover, and frequently have a sharp gradient However, movement between pools via the cascades and slides was observed during the course of the study. Therefore, Sonora chub do tolerate velocities higher than those indicated by the microhabitat data.
The lack of surface flow in mid-summer eliminates invertebrate drift as a potential food resource. Thus, in these isolated pools, Sonora chub may have few food resources during this time. Additional data from the laboratory and field observations also suggest limited food resources in Sycamore Creek. When well-fed Sonora chub were marked with colored epoxy in the lab, the marks persisted relatively well and no fish were damaged. In the field, Sonora chub were aggressive feeders. The same marked fish in Sycamore Creek were attacked, usually until the marks were pulled off. Two fish were found dead the day after marking, and the marks were intact but many scales had been removed from the dorsum of the fish. Also, Sonora chub would constantly nibble on observers in the stream. 51
Mean Column Depth. Position in Water Column, and Depth Heterogeneity
In two of eight sites, subadults selected mid-range depths of 11 to 30 cm. In two of six
sites, juveniles selected depths of under 20 cm; these comparisons were significant at the
P s 0.007 level, so they are probably not due to chance. In two sites, adults selected deeper
areas than both juveniles and subadults. Adults were usually observed near the bottom or
dispersed throughout the water column. This pattern between age or size and column depth has
been observed with many other stream fish (Moyle and Baltz 1985, Heggenes 1988, Baltz et al.
1991) including cyprinids (Gorman 1987, Schlosser 1987, Bestgen and Propst 1989).
Sonora chub juveniles occurred primarily in shallow water. Juveniles also were found in
areas with zero velocity, plant or algae cover, and fine gravel; these characteristics are indicative
of stream margins in Sycamore Creek. Riparian vegetation and shallower depths along the
channel edge would provide some protection from flooding and from potentially cannibalistic
adults. Sonora chub have not been observed eating their young; however, minimal recruitment
occurs in the display tank at the Arizona Sonora Desert Museum (H. Lawler, ASDM, personal
communication), which contains no other species and has little cover for the hatched fry. The
margins and areas with dense algae in Sycamore Creek were frequently warmer than deeper
sections of the stream. John (1964) reported that young of the year are more tolerant of high
temperatures than adults. Several researchers have reported that cyprinid fry disperse to stream
margins (Minckley and Barber 1971, Bestgen and Propst 1989).
Adult fish of several species use deeper water than other age classes in the summer
(Schlosser 1982, Moyle and Baltz 1985, Heggenes 1988, Bestgen and Propst 1989, Baltz et al.
1991). Deeper areas are cooler, and provide better protection from predators such as coatis
(Nasua narica) and herons (Family Ardeidae).
Some sites changed dramatically in water level, and thus in extent of available habitat,
between subsamples. Site 24 dropped 12.5 cm in 2 days, and data pooled for both subsamples 52 could not be used at this site. Site 22 only dropped 2.5 cm, but there were major differences in use between subsamples for mean depth and depth heterogeneity (P<0.05). However, the frequency histograms comparing use with availability for mean depth showed little difference between subsamples for Site 22. The fact that use of depth did not change even though water level in Site 48 decreased over 12 cm between subsamples emphasizes that depths were selected by Sonora chub.
In general, Sonora chub did not show a preference for range in depth. Only two of eight sites indicated a significant difference between use and availability; fish selected a depth range of
2 to 4 cm.
Sunlit versus Shaded Areas
There was no clear pattern of selection for shaded or sunlit areas. Out of six sites where selection was possible, Sonora chub selected sunlit areas in two sites, and selected shaded areas in one site. No consistent differences were found among age classes with reference to sunlight
Cover and Distance to Cover
Rocky cover was the primary cover type selected by all life stages of Sonora chub in two sites, and by individual life stages in other sites. There was only one site where Sonora chub showed strong selection for rocky cover, however, and it is possible that these distributions occurred by chance. Nonetheless, there are ecological reasons to suspect that this pattern is real: rock crevices supply potential protection from instream and terrestrial predators, and from high velocity. Rock crevices may be an increasingly important cover type in late summer. Pools at this time are shrinking, which increases vulnerability to predation; and flash floods can occur.
Also, rock cover in the form of larger substrates, such as medium-sized boulders, eventually provides pockets of water protected from direct sunlight These pockets of water increase the 53 chances that the smaller macrohabitats will survive the summer drought. Thus, selection of rock crevices as microhabitat could lead to selection of a macrohabitat that allows survival through drought periods. Six sites contained juveniles; in two, juveniles selected algae or plant cover.
Juveniles may be found in this cover because it frequently occurs in the shallow margins of many sites (see previous discussion on depth).
There was no clear pattern of selection for distance to cover. There were significant differences between use and availability at three of the eight sites, yet the patterns of selection differed among these sites.
Substrate
In several sites, Sonora chub (combined and individual age classes) selected areas dominated by bedrock. Four of the 37 comparisons were significant below the 0.016 level; they are probably not due to chance. The microhabitat results present data only on the maximum substrate size selected; however, sand and gravel were frequently the most abundant substrate, and bedrock was usually overlain by sand and gravel. The finer substrates may provide the best food resources in these isolated sites. Most feeding fish were grazing on the substrate, as opposed to surface feeding, although they usually attempted to eat items that dropped onto the pool surface. Since there was limited drift in the eight microhabitat sites (six of which were isolated pools), most aquatic insects present would be in the sandy substrate (e.g., chironomids, ephemeropteran nymphs). The fish also grazed on attached algae, which was available on the surfaces of bedrock and large boulders, as well as lying atop the sand. Gorman (1987) found that a species that grazed on substrates was the only one of six minnows where substrate selection was significant 54
In Sites 5 and 66, subadults selected coarser substrates than adults. I suggest that coarser substrates provide visual cover, especially from above. It is also possible that subadults may be pushed out of more productive sites (those with sandy substrates) by adults.
Analysis of substrate as a factor of fish habitat is difficult (Bovee 1982, Bain et al. 1985).
Substrate is usually measured by size strata; however, bedrock, one of the most common substrates in Sycamore Creek, is undefinable in terms of substrate size. The solution I chose was to give bedrock the highest substrate rating. This decision posed problems. For example, bedrock and sand were at different ends of the substrate code but would be equally poor in providing cover from flooding or predation. In addition, measures of substrate heterogeneity were difficult because of the ambiguity of this definition.
LIMITING FACTORS FOR SONORA CHUB
Summer habitat conditions that may affect Sonora chub habitat selection are: drying pools; flash flooding; low oxygen and high temperatures; lack of drift; and predation in shrinking pools. I doubt that high temperatures are ever a significant problem. The highest water temperature that I recorded where adults were observed with no apparent mortality was 37° C; however, most pools in Sycamore Creek rarely got above 30° C.
Low oxygen levels are likely a limiting factor during summer drought In 1990,1 found about 14 dead Sonora chub on 29 and 30 June 1990; this was shortly after record high air temperatures combined with cloudy skies (from 25 to 28 June): this weather combination is frequently associated with oxygen depletion in pools (Herman and Meyer 1990). The dead fish appeared to be among the largest individuals (35-100 mm TL) occupying these pools. Tramer
(1977) observed mortality of stream fishes in shrinking pools; he found that fry were the most resistant to mortality and that gulping increased survival. Lowe et al. (1967) also reported that smaller fish survived the lowest oxygen levels. Water temperatures did not appear to be critical 55
in the pools that contained the dead fish on the days I found them, when maximum air tempera
tures had decreased by only 2°C from the record days. The highest temperature recorded in
those pools, collected at the hottest part of the day, was 30° C. I was not able to record oxygen
levels prior to observing the dead fish. However, in 1990 and 1991,1 measured dissolved oxygen
in several pools occupied by Sonora chub. In sites with apparently healthy adult fish, dissolved
oxygen levels were as low as 1.0 ppm.
I also observed fish gulping at the surface in two diying pools. One of these pools dried
up several hours after these fish were observed, and a band of coatis apparently ate the fish.
Sonora chub survived in the other pool; summer rains replenished this pool shortly after I
observed the fish gulping at the surface.
Microhabitat selection varied among sites. Variation in selection may reflect differences
in availability but it may also reflect variations in limiting factors between sites. Sonora chub are
generalists, and their inherent flexibility in responding to environmental variables may produce different habitat use patterns between different pools. Angermeier (1987) found that cyprinids were highly flexible in habitat associations and selectivity in small streams of Illinois, and suggested that growth and survival were not directly limited by habitat characteristics. His data also indicated that a population may exhibit different levels of habitat selectivity in a given reach or year. Harrell (1978) reported that two-thirds of the dominant fish species in Devil's River increased in abundance after severe flooding and also shifted their habitat use; he suggested that such flexibility reflects a high degree of ecological plasticity. Schlosser (1982) found that shallow unstable habitats were inhabited primarily by generalized insectivores. Fausch and Bramblett
(1991) also reported generalized habitat use by dominant cyprinids in a plains stream that exper ienced a highly variable flow regime. 56 REPRODUCTIVE RESPONSE TO FLOODING
Sonora chub showed responses to flooding; after the 1990 summer floods, spawning was
evident, adults were observed attempting to move upstream, and fish were absent from only a
few places in the stream. There is little information on reproductive biology of the Sonora chub.
However, Sonora chub have been observed with breeding colors in August (Branson et al. 1960),
October (Miller 1949), and March (Miller 1945). Also, young have been collected in early spring
(Minckley 1973) as well as in November (Hendrickson and JuArez-Romero 1990). I was in
Sycamore Creek in October, November, December, March and April through September: I
observed fiy in April, May and September in 1990 and 1991. These months correspond to spring
and summer post-flooding conditions.
Bell (1984) qualitatively surveyed Sycamore Canyon one month after a major flood in
October 1983, and reported the number of pools available for chub had decreased and numbers
of adult chub appeared drastically reduced. However, he observed small groups of fry (<15 mm)
throughout the canyon. Fiy undoubtedly hatched from spawning that occurred after the flooding.
Spawning during or immediately after midsummer floods have been reported in other
cyprinids by John (1963), Harvey (1987), and Cross and Moss (1987). Spawning continuously or
after flooding allows persistence despite harsh flow regimes (Fausch and Bramblett 1991). Meffe
(1984) found that even 1-day old Sonoran topminnows (Poeciliopsis occidentalis) show an innate
response and resistance to flooding. John (1964) reported that only young-of-year speckled dace
(Rhinichthys osculus) were noticeably affected by flash floods. Minckley and Meffe (1987) concluded that most native southwestern fishes resist and survive flooding.
Sonora chub may breed as a result of floods (Bell 1984), or may be capable of breeding
throughout the year (Hendrickson and Juarez-Romero 1990). Floods occurring soon after spawning would cause losses of juvenile fish (John 1964, Harvey 1987), especially if a flood that induced spawning was followed by another flash flood. 57
MANAGEMENT IMPLICATIONS
In 1990 and 1991, Sonora chub in Sycamore Canyon appeared healthy and successful spawning was evident However, several phenomena indicate that managers should be cautious about concluding that the Sycamore Creek population is stable.
There is strong evidence of sediment loading in Sycamore Canyon. After the October
1983 flood, a large number of pools filled with sediment (Bell 1984, Hale and Jarchow 1988).
The results of my study demonstrate that the most suitable Sonora chub habitats are the deep pools with impervious substrates; and that adult Sonora chub prefer deeper areas. After flooding in July and August 1990,1 found many of the pools that previously supported large numbers of adult Sonora chub had completely filled with sand. In 1991,12% of the population that was observed in areas containing adult Sonora chub were lost because these fish were in isolated areas that did not persist Some of these groups of fish may have been lost because the in creased instability of the streambed decreased the ability of the Sonora chub to choose appro priate habitat
One possible cause of sediment loading is livestock grazing. Miller (1961) cites the introduction of livestock to the Southwest as leading to increased floods, severe erosion, and sedimentation. He also stated that filling of deep pools by sediments has eliminated species that require these areas for survival. The microhabitat analysis demonstrates that adult Sonora chub frequently selected flne substrates; however, this does not mean that increased sedimentation is desirable. Persistence of pools is not enhanced by instability of the stream bed. In 1991, during a wet spring and mild yet diy summer, many Sonora chub selected pools that eventually dried up; this reduced the U.S. population by at least 20%. In the face of a sustained drought, the
Sycamore population may be at risk.
The entire population of Sonora chub in Sycamore Creek is within the Goodding
Research Natural Area (RNA). The RNA is supposed to be closed to livestock grazing; however, since December 1989,1 periodically saw cattle or recent signs of cattle from Yank's
Spring to the Mexican border. Certain sections of Sycamore Creek are obviously impacted by trampling and grazing (G. Froehlich, USPS biologist, personal communication). Two to five head of cattle, possibly from Mexico, appeared to be permanent residents in the lower reach in 1990 and 1991. In the allotment directly upstream of the Research Natural Area (Bear Valley
Pasture) there appeared to be virtually no riparian vegetation beyond the northern boundary of the RNA. In 1990 and 1991, cattle were frequently seen in this pasture during times when the
USPS said it was closed to grazing. Biologists from Arizona Game and Fish Department
(AGFD) contend that grazing practices within the watershed outside of the RNA can impact the volume of water flowing in Sycamore Creek, and these changes in flow volume may eventually affect the survival of this population during a drought (AGFD letter by T. Reynolds, W. Hayes, and K. Stirn, to Desert Fishes Recovery Team).
Although there is no current mining within the watershed, mining remains a potential threat to Sonora chub. Major mining operations could result in increased siltation, runoff, and toxic inflow. There are uranium deposits in the watershed above the RNA and mining claims do exist within the drainage.
The reason for extirpation of the Tarahumara frog (Rana tarahumarae) in Sycamore
Canyon has never been determined (Hale and Jarchow 1988). Although few dead frogs were found in 1991, numerous dead frogs—primarily leopard frogs (Rana chiricaguensis or R. yavapaiensis)—were observed in 1990. It is possible that some unknown environmental factor still threatens the aquatic residents of Sycamore Creek.
Managing the Sonora chub in the U.S. may include proposals for translocations. The results of my microhabitat research indicate that translocations into cattle tanks or tinajas with steep slopes would probably support adult Sonora chub; however, recruitment may only be possible if there are shallow areas and cover for the juveniles. I suggest that the Sonora chub population should be routinely monitored either during the summer dry season (late June), when habitat conditions are most critical, or shortly after the summer rainy season (September to October). The late June survey would provide information on which habitats and populations are permanent, and the degree of habitat loss due to drought
It would also allow determination of the age-class structure of the population that survived the dry part of the summer. The post-flooding survey would provide information on spawning after the summer rains, as well as survival after flash-floods. 60
APPENDIX A: HABITAT TYPES
The following definitions were used to describe macrohabitats; these are based primarily on Bisson et al. (1982):
Areas of minor groundwater inflow or outflow of pool that were less than 5 cm deep. Considered "minor" because water volume is so low or vegetation level so high that fish movement would be very limited.
Pool: Velocity only at head and tail of habitat
Glide: Less than 25% surface turbulence; velocity observable throughout the habitat
Riffle: More than 25% turbulence; velocity observable throughout the habitat
Cascade: Obvious gradient from head to tail; velocity observable throughout the habitat; may consist of a series of alternating small waterfalls and small shallow pools.
Connected: Adjacent habitats were considered connected if the minimum depth was greater than 2 cm (just deep enough for all age classes to move through).
The habitat types were not used in the macrohabitat analysis but aided in delimiting and identifying areas of surface water when repeating stream surveys. By the time of the final surveys in late June or early July, nearly all habitat types other than pools had dried up. 61 APPENDIX B: LIFE HISTORY INFORMATION, INCLUDING LITERATURE REVIEW, ON SONORA CHUB
TAXONOMY The Sonora chub was first collected in 1893 by Edgar A. Mearns in Sycamore Canyon, Arizona and identified as Richardsonius gibbosus (= Gila intermedia) by J.O. Snyder. It was not until 1945, however, that this form was recognized as a distinct species (Miller 1945). The holotype for Sonora chub was collected in 1940 by R.G. Miller in the Rio Magdalena, Sonora, Mexico. R.R. Miller (1945) identified the chub as Gila ditaenia, and described this species from the holotype and paratypes collected primarily from Mexico. The Sonora chub is in the Cyprinidae family, and Miller (1945) classified it with the Yaqui chub (G. purpurea) and the arroyo chub (G. orcuttii) in the subgenus Temeculina. DeMarais (1991) recently described a new species, G. eremica, from Mexico, which he included within this subgenus.
MORPHOLOGICAL DESCRIPTION The Sonora chub is a small, dark-colored minnow with two prominent, black lateral lines, one above and one below the lateral line (hence the species name, ditaenia). In Arizona it is rarely larger than 125 mm (Minckley 1973), although in Mexico it may reach 190 mm (Hendrick- son and Juarez-Romero 1990). The Sonora chub has small scales (63 to 75 in the lateral line) with prominent radii on all fields; an inferior mouth; a dental formula of 2,5-4,2; and dorsal, anal, and pelvic fins with eight rays (the dorsal fin rarely has nine). Breeding colors are Chinese red at the axils of the pectoral and pelvic fins; at the base of the anal fin, and at the corners of the shoulder and mouth (Miller 1945).
DISTRIBUTION Sonora chub are endemic to the Rio de la Concepcion Basin. In Mexico, they occur in the Rio Magdalena, the Rio Altar, and in some of their tributaries. The U.S. population is restricted to a single watershed, Sycamore Canyon. This canyon is the headwaters of the Rio Altar. Surface water is not continuous throughout the canyon. The Sonora chub is found in Sycamore Creek from Yank's Spring to about 0.4 km below the juncture of Penasco Canyon, and in three side canyons of Sycamore Creek: Pefiasco Canyon and the canyons of Tinaja Spring and Little Tinaja Spring. During years of heavy rainfall, water may flow continuously from Yank's Spring to the International Border, undoubtedly increasing the chub's range in wet years (USFWS 62
1986). However, the Sycamore population may now be permanently isolated from the Mexican population due to the extent of diy streambed south of the border. Water diversions in the upper reaches of the Rio Altar may have increased the separation between populations (Hendrickson and Ju&rez-Romero 1990). Miller (1949) observed people removing chub from Yank's Spring into live containers, and suggested that it may have been introduced into other drainages. However, a USFS survey of nearby canyons reported no other populations of Sonora chub (Bell 1984).
REPRODUCTION Research has not been conducted on reproductive biology of the Sonora chub. However, the chub has been observed with breeding colors in August (Branson et al. 1960), October (Miller 1949), and March (Miller 1945). Also, young have been collected in early spring (Minckley 1973) as well as in November (Hendrickson and Ju4rez-Romero 1990). I observed fry and juveniles (ranging from 15-18 mm) in April, May, and September in 1990 and 1991. This suggests that spawning occurred after the spring and monsoon rains. Bell (1984) also observed fry after heavy flooding, and suggested that post-flood spawning is a survival mechanism evolved by this species (Bell 1984). Hendrickson and Ju&rez-Romero (1990) suggest that Sonora chub is capable of breeding any time of the year if conditions are adequate.
SONORA CHUB FOOD HABITS No quantitative information is available on Sonora chub feeding habits: Minckley (1973) simply states that several collected stomachs contained aquatic and terrestrial insects and algae, in decreasing order of volume. I noted Sonora chub actively feeding in my footprints in fine gravel substrates, and found the pool bottom contained numerous chironomid larvae. Later, Sonora chub ate Chironomids from my hand. While determining substrate sizes, several other aquatic insects were collected and fed to Sonora chub, including larvae of Simuliidae and Ephemer- optera, and an arachnid.
HABITAT All previous information on Sonora chub habitat are from work in Mexico. The type locality for Sonora chub is a reach of the Rio Magdalena near La Casita; the habitat was described as 4 to 5 feet wide, clear water about 1 ft deep, and with a fairly swift current over a 63 sand and gravel substrate (Miller 1945). The principal vegetation was watercress, and most of the collection was from a pool about 3 to 4 ft deep underneath the roots of a fallen tree. Branson et al. (1960) described a collection site on the Rio Magdalena as very turbid and rapidly flowing water, varying in depth from 0.3 to 2 feet, over a substrate of mud and/or sand and gravel. Hendrickson and Juarez-Romero (1990) provided general habitat characteristics from electrofishing data, which suggested that Sonora chub concentrated in deeper areas with cover. They stated that Sonora chub preferred fallen logs, dense aquatic vegetation such as Rorippa acraticum, and undercut root masses in low to medium velocity waters, and were less abundant where current velocity was high. However, based on capture per unit effort, they concluded that Sonora chub preferred areas with the greatest flow, highest velocities, and a predominance of coarse substrate. Floods may improve or degrade Sonora chub habitat. Bell (1984) qualitatively surveyed Sycamore Canyon after a major flood in October 1983 and reported that the number of pools available for chub had decreased and the numbers of adult chub appeared to be drastically reduced. K. Stirn (AGFD, personal communication) was concerned that action needed to be taken to mitigate habitat damage from this flood. However, during a reconnaissance survey in December 1989, B. LeFever (USPS, personal communication) said the pools looked larger than before 1983, and W. Hayes (AGFD, personal communication) was surprised at the number of chub observed and remarked that dcwatered reaches were much smaller than what he expected. These observations suggest that Sonora chub habitat in Sycamore Canyon may have stabilized since the 1983 flood.
INTERACTION WITH OTHER SPECIES Potential threats to Sonora chub from other species include competition, predation, hybridization, and introduction of exotic parasites. Three other fish species have been reported in Sycamore Canyon: longfin dace (Agosia chrysogaster) and the introduced mosquitofish (Gambusia affinis) and green sunfish (Lepomis cyanellus). Records of occurrence in Sycamore Canyon are inconsistent for these species, especially the longfin dace. Silvey et al. (1984) and a few unpublished papers reported that longfin dace have been collected in Sycamore Canyon. These papers refer to Minckley (1973); however, a thorough search through Minckley (1973) has not produced this information. In fact, Minckley (1985) clearly states that longfin dace has not been recorded in Sycamore Canyon, and Sonora chub is the only indigenous fish. Other papers that report Sonora chub is the only native fish in Sycamore Canyon include Miller and Lowe 64
(1964) and USFS (1988). D. Hendrickson studied longfin dace extensively in Arizona, and believes that the report of this species in Sycamore Creek is an error (University of Texas, personal communication). In 1982, Brooks (1982) conducted electrofishing surveys and collected green sunfish in the lower portion of Sycamore Creek. Brooks did not mention longfin dace in this report In 1983, Bell (1984) conducted several visual surveys and although longfin dace was not mentioned, he reported finding both green sunfish and mosquitofish. Brooks (1982) noted that green sunfish appeared to be restricted to the deeper pools. He suggested that large sunfish may prey upon smaller chub; however, this impact did not appear significant and all size ranges of chub were present in pools with sunfish. Sunfish are sensitive to high velocities (Minckley and Meffe 1987), therefore, flash-floods may prevent sunfish from becoming established in Sycamore Canyon. Mosquitofish may be a serious problem for native desert fish species (Minckley and Deacon 1968); however, this species probably did not survive in Pefiasco Canyon (USFWS 1986), since they were located in an ephemeral pool (Bell 1984). Storage tanks in Pefiasco Canyon were believed to be the seed source for both exotic species (Brooks 1982). In the two years that I studied Sonora chub in Sycamore Creek, I searched intensively for these three other species while collecting macrohabitat data. The only fish species seen was Sonora chub. Potential predators that I saw in Sycamore Canyon include belastomid hemipterans (including the fish-consuming Lethocerus medius), leopard frogs (Rana spp.), coati (Nasua narica), and garter snakes (Thamnophis spp.). Other potential predators include the bull frog (Rana catesbeiana), raccoon (Procyon lotor), belted kingfisher (Ceryle alcyon), and herons (Family Ardeidae) (DFRT, in prep.). The only attacks on Sonora chub that I actually observed were by dytiscid and hydrophilid beetles in a tiny shrinking pool. I saw strong circumstantial evidence of coati predation and scavenging (see discussion of microhabitat analysis). In Mexico, Sonora chub are sympatric with several other fish species. Branson et al. (1960) collected Sonora chub on the Rio Magdalena near San Ignacio with longfin dace and Gila topminnow (Poeciliopsis occidentalis). Hendrickson and Juarez-Romero (1990) observed Sonora chub in Mexico with these two natives as well as with several exotic species: green sunfish, bluegill (Lepomis macrochirus), Hlapia sp., and black bullhead (Ameiurus melas). Sonora chub comprised the major portion of total biomass of all fish sampled, and exotic species had very low densities in stream habitats (1.95% by numbers and 16.85% by biomass). This study suggests that 65 the Sonora chub has limited and localized interactions with exotic fish species in Mexico. One avenue of interaction with exotic species is the introduction of exotic parasites. Hendrickson and Juarez-Romero (1990) noted areas in Mexico where Sonora chub were heavily infested by a nematode. Several papers discuss the potential threat to the Mexican population of Sonora chub from hybridization with Gila purpurea (USFWS 1986; Hendrickson and Juarez-Romero 1990). However, a recent paper clarifies that the possible hybrids are not with G. purpurea but with a newly described species, Gila eremica, which was apparently introduced into the Rio de la Concepcion (DeMarais 1991). APPENDIX C: AIR AND WATER TEMPERATURES, 1990 AND 1991 67 Deg. C Air
30 -
20 - /v" * '/ - x • \i / ' 10 - ' -:A • • \s-./
3° -{ y \
\. / 'V V V- "'i J. .»..-"-/\J ••. /• .a '•••". ..• /n""' r
:,H / v / "v * - t Summer rains began Recorder moved to Site 27
10 -
04-May-90 24-May-90 13-Jun-90 03-Ju!-90 23-Jul-90 12-Aug-90 01-Sep-90
30 •
20 H V "',$J \ / V i\A\ .•% / '•/ \ . .a.. •, .< N,/ v 10 - V y\A '"/* / \/V ! •-•. /\'l ^ K/ ' /*" V; .v \ - ' >" \ . / n /\ '••'! \ # w-^ x iI i/ft.!v.*\\ I vy
02-Sep-90 22-Sep-90 12-0ct-90 01-Nov-90 21-Nov-90 11-Dec-90 31-0ec-90
Figure C-l. Average daily temperatures for 1990. Air temperatures from Nogalcs weather station (OSC 1990). Water temperatures recorded with Ryan Tempmentor at Sites 12 and 27 in Sycamore Creek, Arizona. 68 Deg. C Air Water
v/ vA'' As'
01-Jan-91 21-Jan-91 10-Feb-91 02-Mar-91 22-Mar-91 11-Apr-91 01-May-91
30 -
A 20 - AJ\ r-*"
10 -
< I I 04-May-91 24-May-91 13-Jun-91 03-JUI-91 23-JUI-91 12-Aug-91 01-Sep-91
02-Sep-91 22-Sep-91 12-Oot-91 01-Nov-91 21-NOV-91 11-Dec-91 31-Dec-91
Figure C-2. Average daily temperatures for 1991. Air temperatures from Nogales weather station (OSC 1991). Water temperatures recorded with Ryan Tempmentor at Site 27 in Sycamore Creek, Arizona. 69
APPENDIX D: DESCRIPTION OF MICROHABITAT SITES
The following is a detailed description of the eight sites used for the microhabitat
analysis. Locations of all sites are indicated on Figure 1. As described in the data collection
methods, Sycamore Creek was divided into four similar sections, and two sites were randomly
chosen from each section. Physical characteristics of each site are in Table D-l.
Section 1
The two most upstream sites (Sites 5 and 13) contained more than one habitat type:
they consisted of pools connected by low-gradient riffles. The other six sites were pools only;
thus, only Sites 5 and 13 contained measurable velocity.
Site 5 consisted of one main pool over a substrate of gravel interspersed with small
boulders. It flowed into a short riffle of predominantly bedrock-cobble. This riffle flowed into a
small ephemeral cascade that bars fish movement to and from Site 6, depending on flow condi
tions. Low flows during the sample period appeared to prevent fish movement
Site 13 began as a small pool with a gravel and sand bottom with much plant cover from
tree roots. This pool flowed over a short cascade and run over a predominantly bedrock and
sand substrate. No juveniles were in Site 13 during sampling.
Section 2
Sites 22 and 24 were fairly similar, although Site 24 had much deeper sections. They were against the west side of the canyon as the creek flows south; the bank was fairly sharp from recent floods, and exposed roots from trees created shade and instream structure at the upper end of both sites. The substrate was predominantly bedrock overlain with gravel and sand,
interspersed with large boulders along the stream margin and smaller boulders instream. Site 24 was isolated at the time of the sample period. However, Site 22 was contiguous with the next
downstream pool
Section 3
Site 46 was a long, deep, isolated pool along the west side of the canyon. Substrate was
predominantly bedrock overlain with sand, with small-medium boulders interspersed throughout
The west bank was well vegetated with large trees, and the upper end of the east bank was thick
with willows. Like many pools in Sycamore Creek, the downstream end was fairly exposed.
Site 48 was a plunge pool cut into the bedrock at a point in the canyon that undoubtedly
experiences tremendous flooding pressure. The canyon is not treacherously narrow here as it is
in other areas (see Site 73); however, evidence of flooding is obvious from the entirely exposed
bedrock that is continuous from each canyon wall. There is also a sharp drop in elevation from
Site 48 to the pool below: hikers attempting to travel through this section must climb over a
fairly precarious wall. Site 48 had a bedrock bottom overlain with a thick layer of gravel-sand
interspersed with large boulders. The bedrock slide between Sites 48 and 49 is frequently
passable for downstream fish and probably for upstream fish as well; however, at the time of the sample period, the water level had dropped so that Site 48 was isolated. It is connected to the upstream part of the creek by a 1-m cascade that is usually impassible for upstream fish, even after recent flooding. There were very few juveniles in this pool at the time of sampling.
Section 4
Site 66 was a well-shaded pool below the confluence of Peiiasco Creek. Like many pools in this section, it had large boulders at the upstream end, which provide the deepest sections of the pool, and the pool opened up with finer substrate at the downstream end. Substrate in the middle of this pool was predominantly medium-sized cobble with patches of gravel-sand.
Vegetation along the banks was mature trees.
Site 73 was an isolated pool surrounded by large (>3 m) boulders that provide substan tial shade. The pool was just below the narrowest section of Sycamore Canyon; less than 20 m from the beginning of Site 73, the canyon walls are about 4.3 m apart It is not surprising that heavy flooding impact was evident in this area. This pool was isolated during sampling, although in late spring and after summer rains it is frequently connected to downstream portions of the creek. The pool contains a substrate of gravel-sand interspersed with medium-sized boulders, especially along the margins of the pool.
Table D-l. Habitat characteristics of sites that were used for microhabitat analysis.
Number of Sampling Maximum Total Estimated Presence Locations Depth Volume* Number of Age Site (cm) (m3) of Fish" Classes+ Used Unused
5 29.85 1.466 152 J,S,A 57 43 13 27.31 1.441 84 S,A 28 72 22 29.21 4.012 332 J,S,A 81 19 24 47.63 2.588 297 J,S,A 79 21 46 82.55 12.839 269 J,S,A 63 37 48 74.93 4.509 234 SA 71 29 66 41.91 4.170 86 J,SA 30 70 73 60.33 8.673 284 J,SA 55 45
Total volume determined by total surface area X mean depth
This is an index; it is based on total number observed during microhabitat study
J=juveniles; S=subadults; A=adults APPENDIX E: RESULTS OF STATISTICAL COMPARISONS 73
Table E-1. Use vs. availability by juveniles, subadults, adults, and all ages combined. Pooled data: N for availability is 100; Subsample: N for availability is 50. NS = Nonsignificant (P > 0.1 or N < 5).
Juveniles Subadults Adults All Fish
Stand. Stand. Stand. Stand. Type of Test P- Test P- Test P- Test P- Site Dataset N Stat. value N Stat. value N Stat. value N Stat. value
Velocity (Code: 0,1,2)
Site 5 Pooled 20 -3.11 0.003 29 -1.50 0.073 20 -3.11 0.003 57 -2.06 0.027
Site 13 Pooled ... NS NS 28 -1.52 0.073
Mean Depth of Uater Column (cm)
Site 5 Pooled NS NS NS 57 -1.63 0.060
Site 13 Pooled ... NS NS NS
Site 22 Subsample 1 NS NS NS NS Subsample 2 NS NS 23 -1.32 0.099 42 -1.79 0.051
Site 24 Subsample 1 NS NS NS NS Subsample 2 NS NS 23 -1.78 0.043 NS
Site 46 Pooled 40 -2.54 0.007 23 -1.48 0.068 NS NS
Site 48 Pooled ... 39 -1.86 0.043 NS NS
Site 66 Subsample 1 11 -3.27 0.001 NS NS NS Subsample 2 NS NS NS NS
Site 73 Subsample 1 NS NS NS NS Subsample 2 NS 11 -2.15 0.024 NS NS
Depth Range (cm)
Site 5 Subsample 1 NS NS NS NS Subsample 2 NS NS 13 -1.48 0.073 NS
Site 13 Subsample 1 NS NS NS Subsample 2 ... NS NS NS
Site 22 Subsample 1 NS 22 -2.16 0.019 33 -1.28 0.105 39 -2.08 0.028 Subsample 2 21 -1.85 0.035 NS NS 42 -1.43 0.089
Site 24 Subsample 1 NS NS 36 -2.91 0.007 NS Subsample 2 NS NS NS NS
Site 46 Pooled NS NS NS NS
Site 48 Subsample 1 NS NS NS Subsample 2 ... NS NS NS
Site 66 Subsample 1 NS NS NS NS Subsample 2 NS NS NS NS
Site 73 Pooled NS NS NS NS Table E-1. Continued.
Juveniles Subadults Adults All Fish
Stand. Stand. Stand. Stand. Type of Test P- Test P- Test P- Test P- Site Dataset N Stat, value N Stat, value N Stat, value N Stat, value
Shade (Code: 0,1)
Sites 5 and 13 were entirely in sunlight
Site 22 Subsample 1 11 -2.49 0.026 NS 22 -2.73 0.013 39 -1.97 0.040 Subsample 2 NS NS NS NS
Site 24 Subsample 1 NS NS NS NS Subsample 2 NS NS NS NS
Site 46 Pooled NS NS 23 -2.58 0.013 63 -1.51 0.074
Site 48 Subsample 1 20 -2.43 0.035 31 -2.62 0.029 31 -2.62 0.029 Subsample 2 NS NS NS
Site 66 Subsample 1 NS NS NS NS Subsample 2 NS NS NS NS
Site 73 Subsample 1 13 -2.06 0.044 NS NS 31 -2.67 0.016 Subsample 2 NS 11 -1.31 0.094 NS 24 -1.17 0.104
Cover (Code: 0-4)
Site 5 Subsample 1 NS NS NS NS Subsample 2 7-1.90 0.040 NS NS NS
Site 13 Pooled --- NS NS NS
Site 22 Subsample 1 NS NS NS NS Subsample 2 NS NS NS NS
Site 24 Subsample 1 20 -3.96 0.006 NS NS NS Subsample 2 NS NS NS NS
Site 46 Pooled 40 -2.02 0.034 NS NS 63 -1.87 0.040
Site 48 Pooled --- NS NS NS
Site 66 Pooled NS 11 -2.54 0.014 9 -2.25 0.024 30 -1.48 0.078
Site 73 Pooled 25 -1.65 0.055 31 -1.38 0.087 29 -1.29 0.102 55 -2.89 0.004 Table E- 1. Continued.
Juveniles Subadults Adults All Fish
Stand. Stand. Stand. Stand. Type of Test P- Test P- Test P- Test P- Site Dataset N Stat. value N Stat. value N Stat. value N Stat. value
Distance to Cover (cm)
Site 5 Pooled NS NS NS NS
Site 13 Subsample 1 ... NS NS NS Subsample 2 NS NS NS
Site 22 Pooled NS NS NS NS
Site 24 Pooled NS NS NS NS
Site 46 Pooled NS 23 -3.08 0.008 NS 63 -2.43 0.015
Site 48 Subsample 1 20 -1.87 0.036 31 -1.57 0.068 31 -1.57 0.068 Subsample 2 ... NS NS NS
Site 66 Subsample 1 NS NS NS NS Subsample 2 NS NS NS 15 -1.68 0.044
Site 73 Pooled NS 31 -2.81 0.005 29 -2.16 0.020 55 -3.43 0.001
Maximum Substrate Size
Site 5 Pooled NS NS 20 -3.96 0.0007 57 -1.54 0.067
Site 13 Pooled ... 7 -1.39 0.093 NS 10 -1.92 0.044
Site 22 Pooled NS 46 -1.96 0.038 69 -4.21 0.0004 81 -1.34 0.097
Site 24 Pooled NS -44 , -1.26 0.109 59 -1.49 0.076 NS
Site 46 Pooled NS 23 -2.49 0.015 NS NS
Site 48 Pooled ... NS NS NS
Site 66 Subsample 1 NS NS NS NS Subsample 2 NS 9 -1.75 0.049 NS 15 -2.48 0.012
Site 73 Pooled NS NS NS NS 76
Table E-2. Comparisons of use among age classes. Sample sizes for variable "Position in water column" are different than other variables due to different (non-exclusive) sampling protocol. NS = Nonsignificant (P > 0.1 or N < 5).
Juveniles Juveniles Subadults vs. Subadults vs. Adults vs. Adults
Stand. Stand. Stand. Test P- Test P- Test P- Site N Stat. value Stat. value N Stat. value
Velocity
Site 5 18,18 -2.22 0.042 NS 19,10 -3.00 0.020
Site 13 NS NS NS
Mean Depth of Uater Column (cm)
Site 5 18,27 -3.97 0..009 19,19 -11.93 >0.0001 19,10 -5.27 0.003
Site 13 NS NS NS
Site 22 NS NS NS
Site 24 NS NS 12,27 1.77 0.063
Site 46 NS 32,15 -15.70 >0.0001 19,19 -8.71 >0.0001
Site 48 NS NS 7,32 -2.13 0.044
Site 66 13,8 -3.53 0..010 14,7 -6.85 0.0004 NS
Site 73 NS 20,24 -1.39 0.089 14,12 -1.62 0.073
Position in Water Column
Site 5 NS 20,20 -1.82 0.060 NS
Site 13 NS NS NS
Site 22 NS 32,69 -2.59 0.030 46,69 -2.51 0.032
Site 24 NS 38,59 -9.28 >0.0001 44,59 -5.97 0.002
Site 46 NS NS NS
Site 48 NS NS NS
Site 66 16,11 -3.93 0.,010 16,9 -4.64 0.005 NS
Site 73 NS NS 31,29 -2.24 0.040 Table E-2. Continued.
Stand. Stand. Stand. Test P- Test P- Test P- Site N Stat. value N Stat. value N Stat. value
Depth Range (cm)
Site 5 18,18 -1.68 0.068 NS 19,10 -3.85 0.008
Site 13 NS NS NS
Site 22 NS NS NS
Site 24 NS NS NS
Site 46 NS 32,15 -2.55 0.029 19,19 -1.24 0.105
Site 48 NS NS NS
Site 66 NS NS NS
Site 73 NS 20,24 -1.73 0.065 14,12 -1.27 0.103
Shade (Code: 0,1)
Sites 5 and 13 were entirely in sunlight
Site 22 NS NS 8,29 -2.18 0.043
Site 24 NS NS NS
Site 46 NS NS 19,19 -4.48 0.005
Site 48 NS NS NS
Site 66 NS NS NS
Site 73 NS NS NS
Cover (Code: 0-4)
Site 5 NS NS 19,10 -2.09 0.046
Site 13 NS NS 6,6 -1.43 0.086
Site 22 NS NS NS
Site 24 NS 14,35 -2.19 0.043 NS
Site 46 NS NS NS
Site 48 NS NS NS
Site 66 13,8 -1.73 0.065 14,7 -1.39 0.092 NS
Site 73 NS NS NS Table E-2. Continued.
Juveniles Juveniles Subadults vs. Subadults vs. Adults vs. Adults
Stand. Stand. Stand. Test P- Test P- Test P- Site N Stat. value N Stat. value N Stat. value
Distance to Cover (cm)
Site 5 NS NS 19,10 -8.18 0.0001
Site 13 NS NS NS
Site 22 19,33 -2.24 0.041 NS NS
Site 24 NS 14,35 -1.49 0.083 12,27 -1.75 0.064
Site 46 28,11 -1.83 0.059 NS 19,19 -1.81 0.060
Site 48 NS NS NS
Site 66 NS NS NS
Site 73 15,21 -1.30 0.096 NS 14,12 -1.69 0.067
Maximum Substrate Size
Site 5 NS NS 19,10 -5.04 0.004
Site 13 NS NS NS
Site 22 NS 8,43 -2.78 0.025 NS
Site 24 NS NS NS
Site 46 NS NS 19,19 -1.39 0.088
Site 48 NS NS NS
Site 66 NS NS NS
Site 73 NS NS NS 79
REFERENCES CITED
Angermeier, P.L. 1987. Spatiotemporal variation in habitat selection by fishes in small illinois streams. Pages 52-60 in W.J. Matthews and D.C. Heins, editors. Community and evolu tionary ecology of North American stream fishes. University of Oklahoma Press, Norman.
Arizona Game and Fish Department (AGFD). 1988. Threatened native wildlife in Arizona. Arizona Game and Fish Department, Phoenix.
Bachman, R.A 1984. Foraging behavior of free-ranging wild and hatchery brown trout in a stream. Transactions of the American Fisheries Society 113:1-32.
Bain, M.B, J.T. Finn, and H.E. Booke. 1985. Quantifying stream substrate for habitat analysis studies. North American Journal of Fisheries Management 5:499-506.
Baltz, D.M. 1990. Autecology. Pages 583-605 in C.B. Shreck and P.B. Moyle, editors. Methods for fish biology. American Fisheries Society, Bethesda, Maryland.
Baltz, D.M., B. Vondracek, L.R. Brown, and P.B. Moyle. 1991. Seasonal changes in micro- habitat selection by rainbow trout in a small stream. Transactions of the American Fisheries Society 120:166-176.
Bell, G. 1984. Sonora chub, Sycamore Canyon, Nogales RD. U.S. Forest Service, Coronado N.F., inter-office memo.
Bestgen, K.R., and D.L. Propst. 1989. Distribution, status, and notes on the ecology of Gila robusta (Cyprinidae) in the Gila River Drainage, New Mexico. Southwestern Naturalist 34:402-412.
Biondini, M.E., P.W. Mielke, Jr., and E.F. Redente. 1988. Permutation techniques based on Euclidean analysis spaces: a new and powerful statistical method for ecological research. Coenoses 3:155-174.
Bisson, P.A, J.L. Nielsen, R.A Palmson, and L.E. Grove. 1982. A system of naming habitat types in small streams, with examples of habitat utilization by salmonids during low streamflow. Pages 62-73 in N.B. Armantrout, editor. Acquisition and utilization of aquatic habitat inventory information. American Fisheries Society, Western Division, Bethesda, Maryland.
Bovee, K.D. 1982. A guide to stream habitat analysis using the instream flow incremental methodology. U.S. Fish and Wildlife Service, FWS/OBS-82/26.
Bovee, K.D. 1986. Development and evaluation of habitat suitability criteria for use in the Instream Flow Incremental Methodology. Instream flow information paper 21. U.S. Fish and Wildlife Service Biological Report 86(7).
Bowers, J.E., and S.P. McLaughlin. 1982. Plant species diversity in Arizona. Madrono 29:227- 233. 80
Bozek, M.A., and F.J. Rahel. 1991a. Assessing habitat requirements of young Colorado River cutthroat trout by use of macrohabitat and microhabitat analyses. Transactions of the American Fisheries Society 120:571-581.
Bozek, M.A., and F.J. Rahel. 1991b. Comparison of streamside visual counts to electrofishing estimates of Colorado River cutthroat trout fry and adults. North American Journal of Fisheries Management 11:38-42.
Branson, B.A., C.J. McCoy Jr., and M.E. Sisk. 1960. Notes on the freshwater fishes of Sonora with an addition to the known fauna. Copeia 1960:217-221.
Brooks, J. 1982. Sycamore Creek survey - Gila ditaenia. Arizona Game and Fish Department, inter-office memo. August 20,1982. 2pp.
Burger, L.W., Jr., M.R. Ryan, D.P. Jones, and A.P. Wywialowski. 1991. Radio transmitters bias estimation of movements and survival. Journal of Wildlife Management 55:693-697.
Cade, B.S., and R.W. Hoffman. 1990. Winter use of douglas-fir forests by blue grouse in Colorado. Journal of Wildlife Management 54:471-479.
Cross, F.B., and R.E. Moss. 1987. Historic changes in fish communities and aquatic habitats in Plains streams of Kansas. Pages 155-165 in W.J. Matthews and D.C. Heines, editors. Community and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman.
DeMarais, B.D. 1991. Gila eremica, a new cyprinid fish from northwestern Sonora, Mexico. Copeia 1991: 178-189.
Desert Fish Recovery Team (DFRT). In prep. Sonora chub recovery plan. USFWS, Albuquer que, New Mexico.
Fagen, R. 1988. Population effects of habitat change: a quantitative assessment. Journal of Wildlife Management 52:41-46.
Fausch, K.D., and R.G. BrambletL 1991. Disturbance and fish communities in intermittent tributaries of a western Great Plains river. Copeia 1991:659-674.
Goodding, L.N. 1961. Why Sycamore Canyon in Santa Cruz County should be preserved as a natural sanctuary or natural area. Journal of the Arizona Academy of Science 1:113-115.
Gorman, O.T. 1987. Habitat segregation in an assemblage of minnows in an Ozark Stream. Pages 33-41 in W.J. Matthews and D.C. Heins, editors. Community and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman.
Hale, S.F., and J.L. Jarchow. 1988. The status of the Tarahumara frog (Rana tarahumarae) in the United States and Mexico: Part II. Prepared for Arizona Game and Fish Depart ment, Phoenix, Arizona, and U.S. Fish and Wildlife Service, Albuquerque, New Mexico. 81
Harrell, H.L. 1978. Response of the Devil's River (Texas) fish community to flooding. Copeia 1978:60-68.
Harvey, B.C. 1987. Susceptibility of young-of-the-year fishes to downstream displacement by flooding. Transactions of the American Fisheries Society 116:851-855.
Heggenes, J. 1988. Effects of short-term flow fluctuations on displacement of, and habitat use by, brown trout in a small stream. Transactions of the American Fisheries Society 117:336- 344.
Heggenes, J., A. Brabrand, and S.J. Saltveit 1990. Comparison of three methods for studies of stream habitat use by young brown trout and Atlantic salmon. Transactions of the American Fisheries Society 119:101-111.
Helfman, G.S. 1978. Patterns of community structure in fishes: summary and overview. Environmental Biology of Fishes 3:129-148.
Hendrickson, D.A., and L.R. Juarez-Romero. 1990. Fishes of the Rio de la Concepcion basin, Sonora, Mexico, with emphasis on determination of status of the Sonora chub, Gila ditaenia, a threatened species. Southwestern Naturalist 36(2).
Herman, R.L., and F.P. Meyer. 1990. Fish kills due to natural causes. Pages 41-44 in Meyer, F.P and L.A. Barclay, editors. Field manual for the investigation of fish kills. U.S. Fish and Wildlife Service Resource Publication 177.
Hobbs, N.T., and T.A. Hanley. 1990. Habitat evaluation: do use/availability data reflect carrying capacity?. Journal of Wildlife Management 54:515-522.
John, K.R. 1963. The effect of torrential rains on the reproductive cycle of Rhinichthys osculus in the Chiricahua Mountains, Arizona. Copeia 1963:286-291.
John, K.R. 1964. Survival of fish in intermittent streams of the Chiricahua Mountains, Arizona. Ecology 45:112-119.
Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71.
Lowe, C.H., D.S. Hinds, and E.A. Halpern. 1967. Experimental catastrophic selection and tolerances to low oxygen concentration in native Arizona freshwater fishes. Ecology 48:1013-1017.
Meffe, G.K. 1984. Effects of abiotic disturbance on coexistence of predator-prey fish species. Ecology 65:1525-1534.
Miller, L. 1949. Field notes on the minnow, Gila ditaenia in southern Arizona. Copeia 1949:148-150.
Miller, R.G., Jr. 1966. Simultaneous statistical inference. McGraw-Hill. New York, New York. 82
Miller, R.R. 1945. A new cyprinid fish from southern Arizona, and Sonora, Mexico, with the description of a new subgenus of Gila and a review of related species. Copeia 1945:104- 110.
Miller, R.R. 1961. Man and the changing fish fauna of the American southwest Pages 365-404 in S. Baker, editor. Papers of the Michigan Academy of Science, Arts, and Letters. Vol. XLVI.
Miller, R.R., and C.H. Lowe. 1964. An annotated check list of the fishes of Arizona. Pages 133- 151 in C. Lowe, editor. The vertebrates of Arizona. The University of Arizona Press, Tucson, Arizona.
Mills, G.S. 1977. New locations for the five-striped sparrow in the United States. Western Birds 8:121-130.
Minckley, W.L. 1973. Fishes of Arizona. Arizona Game and Fish Department Phoenix, Arizona.
Minckley, W.L. 1985. Native fishes and natural aquatic habitats in U.S. Fish and Wildlife Service Region II west of the continental divide. Report to the U.S. Fish and Wildlife Service, Albuquerque, New Mexico. Department of Zoology, Arizona State University, Tempe, Arizona.
Minckley, W.L., and W.E. Barber. 1971. Some aspects of biology of the longfin dace, a cyprinid fish characteristic of streams in the Sonoran desert Southwestern Naturalist 15:459-464.
Minckley, W.L., and J.E. Deacon. 1968. Southwestern fishes and the enigma of "endangered species". Science 159:1424-1432.
Minckley, W.L., and G.K. Meffe. 1987. Differential selection for native fishes by flooding in the American southwest Pages 93-104 in W.J. Matthews and D.C. Heines, editors. Commu nity and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman.
Moyle, P.B., and D.M. Baltz. 1985. Microhabitat use by an assemblage of California stream fishes: developing criteria for instream flow determinants. Transactions of the American Fisheries Society 114:695-704.
OSC (Office of the State Climatologist). 1990. Arizona climate summary. Office of Climatology, Arizona State University, Tempe, Arizona.
OSC (Office of the State Climatologist). 1991. Arizona climate summary. Office of Climatology, Arizona State University, Tempe, Arizona.
Power, M.E., and W.J. Matthews. 1983. Algae-grazing minnows (Campostoma anomalum), piscivorous bass (Micropterus), and the distribution of attached algae in a small prairie margin stream. Oecologia 60:328-332. 83
Schlosser, I J. 1987. The role of predation in age- and size-related habitat use by stream fishes. Ecology 68:651-659.
Sellers, W.D., and R.H. Hill, editors. 1974. Arizona Climate, 1931-1972. University of Arizona Press, Tucson, Arizona.
Shirvell, C.S., and R.G. Dungey. 1983. Microhabitats chosen by brown trout for feeding and spawning in rivers. Transactions of the American Fisheries Society 112:355-367.
Silvey, W., J.N. Rinne, and R. Sorenson. 1984. Index to the natural drainage systems of Arizona-a computer compatible digital identification of perennial lotic waters. USDA, Forest Service Wildlife Unit Technical Report, Fort Collins, Colorado.
Toolin, L., T.R. Van Devender, and J.M. Kaiser. 1980. The flora of Sycamore Canyon, Pajarito Mountains, Santa Cruz County, Arizona. Journal of the Arizona-Nevada Academy of Science 14:66-74.
Tramer, E.J. 1977. Catastrophic mortality of stream fishes trapped in shrinking pools. Ameri can Midland Naturalist 97:469-478
Trautman, M.B. 1981. Fishes of Ohio, revised edition. Ohio State University Press, Columbus.
U.S. Fish and Wildlife Service (USFWS). 1986. Endangered and threatened wildlife and plants: final rule to determine the Sonora chub to be a threatened species and to determine its critical habitat Federal Register 51(83): 16042-16047.
U.S. Forest Service (USFS). 1988. Establishment record for the Goodding Research Natural Area within the Coronado National Forest, Santa Cruz County, Arizona.
Van Home, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901.
Van Home, B. 1986. Summary: When habitats fail as predictors—the researcher's viewpoint. Pages 257-258 in J. Verner, M.L. Morrison, and C.J. Ralph, editors. Wildlife 2000, Modeling habitat relationships of terrestrial vertebrates. University of Wisconsin Press, Madison.
Wiens, J.A., J.T. Rotenberiy, and B. Van Home. 1987. Habitat occupancy patterns of North American shrubsteppe birds: the effects of spatial scale. Oikos 48:132-147.
Zar, J.H. 1984. Biostatistical analysis, 2nd edition. Prentice-Hall, Englewood Hills, New Jersey.
Zimmerman, G.M., H. Goetz, and P.W. Mielke, Jr. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology 66:606-611.