Roost selection of bridges by bats in an urban area

              ROOST SELECTION OF BRIDGES BY BATS IN AN URBAN AREA
Sandy A. Wolf
William W. Shaw
Arizona Game and Fish Department
Heritage Grant [U98007]
Final Report
October, 2002
DISCLAIMER 2
The findings, opinions, and recommendations in this report are those of the investigators who have received partial or full finding from the Arizona Game and Fish Department Heritage Fund. The findings, opinions, and recommendations do not necessarily reflect those of the Arizona Game and Fish Commission or the Department, or necessarily represent official Department policy or management practice. For further information, please contact the Arizona Game and Fish Department.
ACKNOWLEDGEMENTS 3
We are grateful to the many people who helped make this project possible. Arizona Game and Fish Department Region 5 personnel provided assistance throughout the study; we especially thank Scott Richardson. The City of Tucson Department of Transportation, Pima County Public Works Department, Arizona Department of Transportation, and Pima County Parks and Recreation also provided support and assistance. Tanque Verde Guest Ranch graciously allowed us to monitor the bridge on their property.
Dave Dalton provided invaluable assistance throughout this project, including technical assistance in equipment design, equipment manufacture, and data collection techniques. He volunteered countless hours helping with field work and manuscript review. He also shared his extensive knowledge of bat biology and natural history, and provided ideas and stimulating discussions. Brian Keeley of Bat Conservation International provided guidance many times; thank you for your willingness to share your knowledge and experience of bats and bridges. Ginny Dalton shared her expertise in bat handling techniques and species identification and helped with the initial stages of this study. This project would not have been possible without the hard work of our excellent and always cheerful field assistants: Alicia Allen, Melanie Bucci, Sherry Daugherty, Sarah Lantz, and Sarah Schmidt. Volunteers from the University of Arizona also helped with field work, as did friends and bat-friendly local residents such as Homer Bull. Bob Steidl provided statistical advice. The comments from anonymous reviewers greatly improved this manuscript. This project was funded by Arizona Game and Fish Department Heritage Grant U98007.
TABLE OF CONTENTS 4
LIST OF FIGURES.............................................................................................................8
LIST OF TABLES.............................................................................................................10
ABSTRACT.......................................................................................................................12
INTRODUCTION.............................................................................................................13
STUDY AREA..................................................................................................................17
METHODS........................................................................................................................19
Initial survey to locate potential roost sites....................................................................19
Determining bat presence, abundance, and seasonal patterns........................................23
Determining colony characteristics................................................................................26
Bat captures..................................................................................................................26
Banding........................................................................................................................27
Crevice measurements....................................................................................................28
Width...........................................................................................................................28
Depth...........................................................................................................................31
Height..........................................................................................................................31
Dimensions of crevices used by bats...........................................................................31
Bridge characteristics......................................................................................................32
Structure type...............................................................................................................32
Feature under bridge....................................................................................................33
Human activity.............................................................................................................33
Adjacent land use.........................................................................................................33
Vegetation restricting access.......................................................................................34
Bridge age....................................................................................................................34
Bridge compass orientation.........................................................................................34
Temperature....................................................................................................................35
Analyses.........................................................................................................................39
Monthly surveys..........................................................................................................39
Crevice dimensions......................................................................................................41
Bridge characteristics...................................................................................................42
RESULTS.........................................................................................................................46
General results................................................................................................................46
Overview......................................................................................................................46
Crevice dimensions......................................................................................................52
Bridges......................................................................................................................52
TABLE OF CONTENTS- continued 5
Use by bats................................................................................................................53
Dimensions associated with bat presence.................................................................53
Bridge characteristics...................................................................................................55
Structure type............................................................................................................55
Bridges...................................................................................................................55
Use by bats.............................................................................................................58
Feature under bridge.................................................................................................60
Bridges...................................................................................................................60
Use by bats.............................................................................................................60
Human activity under bridge....................................................................................60
Bridges...................................................................................................................60
Use by bats.............................................................................................................64
Adjacent land use......................................................................................................66
Bridges...................................................................................................................66
Use by bats.............................................................................................................66
Vegetation obstructing access...................................................................................66
Bridges...................................................................................................................66
Use by bats.............................................................................................................69
Compass orientation of bridge..................................................................................69
Bridge age.................................................................................................................69
Temperature.................................................................................................................69
Intra-bridge comparisons..........................................................................................72
Inter-bridge comparisons..........................................................................................74
Seasonal comparisons...............................................................................................74
Summary...................................................................................................................77
Species results.................................................................................................................77
Mexican free-tailed bat................................................................................................80
Western pipistrelle.......................................................................................................83
Cave myotis.................................................................................................................83
Big brown bat...............................................................................................................88
Pallid bat......................................................................................................................88
Yuma myotis................................................................................................................93
California myotis.........................................................................................................93
Silver-haired bat...........................................................................................................93
DISCUSSION....................................................................................................................94
Census techniques and methods.....................................................................................94
Estimating abundance..................................................................................................94
Survey frequency.........................................................................................................95
Annual variation in bridge use.....................................................................................96
Proportion of bridges used by bats in Tucson.................................................................97
TABLE OF CONTENTS- continued 6
Unoccupied bridges.....................................................................................................99
Bat species found in bridges.........................................................................................100
Abundance of bats......................................................................................................101
Colony type................................................................................................................102
Seasonal patterns........................................................................................................104
Variation among individual bridges in use by bats.......................................................105
Bridge characteristics....................................................................................................108
Crevice dimensions....................................................................................................108
Depth.......................................................................................................................109
Height......................................................................................................................112
Why do Mexican free-tails roost in higher bridges?............................................113
Width.......................................................................................................................114
Measurement and analysis...................................................................................114
Widths used by bats.............................................................................................118
Conclusions..........................................................................................................120
Structure type.............................................................................................................120
Feature under bridge..................................................................................................124
Human activity under bridges....................................................................................124
Adjacent land use.......................................................................................................130
Vegetation obstructing access....................................................................................131
Temperature...............................................................................................................134
Bridges....................................................................................................................134
Bats.........................................................................................................................135
Factors not measured.................................................................................................138
Does competition among species exist in bridge roosts?.............................................139
Species diversity and degree of urbanization...............................................................142
MANAGEMENT IMPLICATIONS...............................................................................145
The importance of bridges to bats in Tucson................................................................145
Relative importance of individual bridges.................................................................145
The importance of bridges compared to buildings and natural roost sites in Tucson146
Locating bridges with bats—suggested protocol..........................................................147
Identify bridges with the greatest potential for roost sites.........................................147
Initial site inspection..................................................................................................147
Identify structure type of bridge and potential roost sites......................................147
Survey for bats and evidence of bats......................................................................148
Follow-up surveys......................................................................................................148
Minimizing disturbance to bats..................................................................................149
Additional survey needs................................................................................................149
Maintenance, repair, and replacement of bridges.........................................................150
Advanced planning....................................................................................................150
TABLE OF CONTENTS- continued 7
Scheduling work on bridges......................................................................................151
Bridge design and mitigation structures.......................................................................152
Other issues in management of urban bats...................................................................155
Bridging the education gap........................................................................................155
Vandalism to roost sites.............................................................................................157
Providing foraging habitat.........................................................................................158
Bridges and bat conservation........................................................................................159
SUMMARY AND CONCLUSIONS..............................................................................161
LITERATURE CITED....................................................................................................167
APPENDIX A: 1999 Monthly survey dates and results for each bridge........................A-1
APPENDIX B: Structural data and measurements for each bridge................................B-1
APPENDIX C: Detailed temperature results..................................................................C-1
APPENDIX D: Mexican free-tailed bat (Tadarida brasiliensis)..................................D-1
APPENDIX E: Western pipistrelle (Pipistrellus hesperus)...........................................E-1
APPENDIX F: Cave myotis (Myotis velifer).................................................................F-1
APPENDIX G: Big brown bat (Eptesicus fuscus).........................................................G-1
APPENDIX H: Pallid bat (Antrozous pallidus).............................................................H-1
APPENDIX I: Yuma myotis (Myotis yumanensis).........................................................I-1
APPENDIX J: California myotis (Myotis californicus)..................................................J-1
APPENDIX K: Silver-haired bat (Lasionycteris noctivagans)....................................K-1
APPENDIX L: CONFIDENTIAL (under separate cover) Locations of bridges
occupied by bats and maps of bridge locations, Tucson, Arizona, 1999....................L-1
LIST OF FIGURES 8
Figure 1. Map of study area, Tucson, Arizona.................................................................18
Figure 2. Illustration of terms used in crevices and bridges.............................................21
Figure 3. Examples of potential day-roost sites in bridges in Tucson, Arizona
not included in this study..............................................................................................22
Figure 4. Example of a roost diagram for a hypothetical bridge......................................25
Figure 5. Concrete beam types found in bridges selected for study, Tucson, Arizona....56
Figure 6. Comparison of temperature profiles in different areas of Park/Ajo
bridge (9042) May 1-4, 2000........................................................................................75
Figure 7. Comparison of temperatures in 2 bridges, Broadway (9033) and Campbell
(9618), of the same structure type, crevice depth, and degree of surrounding
urbanization, monitored November 26-29, 2000, Tucson, Arizona.............................76
Figure 8. Moderation of ambient temperatures in bridge crevices in summer,
Tucson, Arizona............................................................................................................78
Figure 9. Moderation of ambient temperatures at the top of a crevice in winter,
Campbell bridge (9618), November 26-29, 2000, Tucson, Arizona............................79
Figure 10. Location of bridges occupied by Mexican free-tailed bats, Tucson,
Arizona, 1999................................................................................................................81
Figure 11. Monthly survey results for Mexican free-tailed bats for all bridges (n = 32)
combined in Tucson, Arizona, 1999.............................................................................82
Figure 12. Location of bridges occupied by western pipistrelles, Tucson,
Arizona, 1999................................................................................................................84
Figure 13. Monthly survey results for western pipistrelles for all bridges (n = 15)
combined in Tucson, Arizona, 1999.............................................................................85
Figure 14. Location of bridges occupied by cave myotis, Tucson, Arizona, 1999..........86
Figure 15. Monthly survey results for cave myotis for all bridges (n = 10) combined,
Tucson, Arizona, 1999..................................................................................................87
LIST OF FIGURES- continued
9
Figure 16. Location of bridges occupied by big brown bats, Tucson, Arizona, 1999......89
Figure 17. Monthly survey results for big brown bats for all bridges (n = 9) combined
in Tucson, Arizona, 1999..............................................................................................90
Figure 18. Location of bridges occupied by pallid bats, Yuma myotis, California
myotis, and silver-haired bat, Tucson, Arizona, 1999..................................................91
Figure 19. Monthly survey results for pallid bats in one bridge in Tucson, Arizona,
1999..............................................................................................................................92
Figure 20. Comparison of monthly trends of cave myotis in a mine and the bridge
nearest to the mine in Tucson, Arizona, 1999............................................................106
Figure 21. Comparison of distributions of widths in crevices in 2 bridges with the
same mean width (11 mm)..........................................................................................116
LIST OF TABLES
Table 1. Criteria for categorizing bridges as primary or incidental use by bats based 10
on surveys, Tucson, Arizona, 1999...............................................................................43
Table 2. Bat species found in bridges in the Tucson metropolitan area, 1999.................47
Table 3. Species found in each bridge, categorized as primary or incidental use,
Tucson, Arizona, 1999..................................................................................................48
Table 4. List of top 5 bridges ranked by abundance (maximum number counted
in a monthly survey) for each species, Tucson, Arizona, 1999....................................50
Table 5. List of top 5 bridges ranked by density of bats for the 4 most
commonly found species, Tucson, Arizona, 1999........................................................50
Table 6. Monthly abundance of bats found in 43 bridges surveyed January-December,
1999, Tucson, Arizona..................................................................................................51
Table 7. Crevice dimensions used by each species in bridges in Tucson, Arizona,
1999..............................................................................................................................54
Table 8. Bridges grouped by structure type, Tucson, Arizona, 1999...............................57
Table 9. Bat use by bridge structure type, Tucson, Arizona, 1999...................................59
Table 10. Bridges grouped by feature under bridge, Tucson, Arizona, 1999...................61
Table 11. Bat use of bridges by the feature under the bridge, Tucson, Arizona, 1999....62
Table 12. Bridges grouped by human activity under bridge, Tucson, Arizona, 1999......63
Table 13. Bat use of bridges by human activity under bridges in Tucson, Arizona,
1999..............................................................................................................................65
Table 14. Bridges grouped by adjacent land use, Tucson, Arizona, 1999........................67
Table 15. Bat use of bridges by adjacent land use, Tucson, Arizona, 1999.....................68
Table 16. Bridges grouped by percent vegetation obstructing access to bridge,
Tucson, Arizona, 1999..................................................................................................70
LIST OF TABLES- continued
Table 17. Bat use of bridges by percent vegetation obstructing access to bridge, 11
Tucson, Arizona, 1999..................................................................................................71
Table 18. Summary of temperature monitoring dates for selected bridges, Tucson,
Arizona, 1999................................................................................................................73
Table 19. Crevice widths used by bat species roosting in bridges: a comparison of
results from 2 studies.................................................................................................120
ABSTRACT
12
Selection of appropriate roosts is crucial to the survival and reproduction of bats. In Tucson, Arizona, bridges seem to be an important source of roost sites for the crevice dwelling species able to live in urban environments. We studied bats that use bridges in the Tucson metropolitan area to provide information to wildlife managers for bat management and conservation. We studied 43 bridges that were constructed with crevices between concrete beams creating potential for roost sites. We found bats in 81% of these bridges and identified 8 species: Tadarida brasiliensis, Eptesicus fuscus, Pipistrellus hesperus, Antrozous pallidus, Lasionycteris noctivagans, Myotis velifer, M. californicus, and M. yumanensis. We conducted monthly surveys January through December 1999 to follow changes in abundance and diversity and identify maternity colonies. We identified some characteristics of bridges that are correlated with the presence of bats, such as adjacent land use, crevice width, crevice depth, and bridge height. State, county, and city departments of transportation can use study results to make engineering decisions that will benefit bats or minimize negative effects on bats using bridges.
80
Mexican Free-tailed Bat
• found in larger numbers than any other bat species in bridges studied
• 25 bridges housed colonies of >100 bats
• up to 11,400 bats (conservative estimate) in 1 bridge at 1 time
• most widespread of bat species using bridges
• found in 74% of 43 bridges studied
• found in bridges all over study area (Figure 10)
• used every channel beam bridge
• present in all months in fluctuating numbers, peak in October (Figure 11)
• seasonal patterns differed among bridges
• some used only during migration
• some used mostly in winter
• some used year round
• probably different populations in winter vs. summer
• both sexes present but segregation of sexes frequently occurs
• parturition occurs late June-mid July
• pregnant females may leave bridges to give birth elsewhere
• only small numbers of young observed
• abundance at most bridges decreased drastically in July
• presence of bats strongly associated with deepest bridge crevices
81
82
050001000015000200002500030000JanFebMarAprMayJunJulAugSepOctNovDec# of bats
Figure 11. Monthly survey results for Mexican free-tailed bats for all bridges (n=32) combined in Tucson, Arizona, 1999.
83
• bat abundance at a bridge may increase as bridge height, amount of vegetation restricting access to bridge, and the number of meters of preferred crevice width increases
• may be sensitive to certain types of disturbance
• not found in bridges over streets
• roost above the least disturbed areas of a bridge
Western Pipistrelle
• found in 15 bridges, in numbers ranging from 1-164 bats in a bridge
• found primarily in north and east parts of the city along washes (Figure 12)
• present in all months with peak in August, minimum number in January and February (Figure 13)
• abundance changes frequently at a bridge, individuals come and go, staying for varying periods of time
• bats roost singly scattered throughout the bridge except for maternity aggregations
• females with young found at 11 bridges, parturition occurs 1st-3rd week of June
Cave Myotis
• found in 10 bridges, in numbers ranging from 1-225 bats in a bridge
• all primary use bridges in outlying areas of metropolitan area (Figure 14)
• present April-October, dramatic decrease in July, peaks in May and September (Figure 15) 84
85
0100200300400500JanFebMarAprMayJunJulAugSepOctNovDec# Bats
Figure 13. Monthly survey results for western pipistrelles for all bridges (n=15) combined in Tucson, Arizona, 1999.
86
87
0100200300400500JanFebMarAprMayJunJulAugSepOctNovDec# of bats
Figure 15. Monthly survey results for cave myotis for all bridges (n=10) combined, Tucson, Arizona, 1999.
• strong evidence that presence of bats is associated with bridges adjacent to undeveloped land (natural open space or agricultural)
• only species to occupy all 3 bridges over flowing water 88
• evidence that sexes segregate- 3 of 4 captured groups were >85% 1 sex
Big Brown Bat
• found in 9 bridges in numbers ranging from 1-88
• found throughout study area (Figure 16)
• present May-October with peak in September (Figure 17)
• females with young found at 5 bridges
• parturition begins at the end of June
• occupied bridges adjacent to both developed and undeveloped land with different degrees and types of human activity
• evidence that presence of bats is associated with bridges with more meters of crevice within the preferred width range
• largest colony roosted in the center of a bridge where temperatures were very stable and several degrees cooler than ambient daily highs
Pallid Bat
• found at only 1 bridge in outlying part of metropolitan area (Figure 18)
• maternity colony, present spring through fall (Figure 19)
• bridge surrounded by large areas of natural open space
89
90
050100150200250JanFebMarAprMayJunJulAugSepOctNovDec# bats
Figure 17. Monthly survey results for big brown bats for all bridges (n=9) combined in Tucson, Arizona, 1999.
91
92
00011458175754737110102030405060708090JanFebMarAprMayJunJulAugSepOctNovDec# bats
Figure 19. Monthly trends for pallid bats in one bridge in Tucson, Arizona, 1999.
• bats roosted over areas with least amount of human disturbance- rare pedestrian traffic only
• parturition occurred about the 3rd week of June
Yuma myotis
• found in 2 bridges (Figure 18) in numbers ranging from 1-3 93
• almost all male, present March-October
California myotis
• found in 2 bridges, 1 or 2 individuals at a time (Figure 18)
• almost all male, present most frequently November-March
Silver-haired bat
• 1 female found in November (Figure 18)
• bridge in small riparian area, in outlying, undeveloped area of metropolitan area
DISCUSSION
CENSUS TECHNIQUES AND METHODS
One important result of this study is that it illustrates the difficulty of obtaining accurate estimates of abundance and the high variability of bats using bridges as roost sites. Both of these factors have important implications for efforts to monitor bat populations and to make decisions that may affect the use of a bridge by bats. 94
Estimating abundance
Estimates of abundance of bats during monthly surveys differed in accuracy, depending on species and roosting patterns. Counts for single bats and small groups of all species were accurate. For large colonies of free-tails, which often roosted in multiple layers, counts became less accurate as colony size and density of bats increased. Because I counted or estimated only the bats I could see, the estimate for the colony was its minimum size, and became more underestimated as size increased. Counts of a species that roosted interspersed among free-tails were also problematic because some bats were hidden from view. For instance, I counted 48 pallids that were mixed in with Mexican free-tails, but a short time later counted 66 pallids. Despite these problems, I am confident that surveys during the day, with binoculars and a spotlight, is the best method for estimating numbers of bats in bridges. Live counts of emerging bats in the evening are not feasible for large bridges, especially with only one person, because the exit area is large, bats exit from both sides of the bridge, and it is often not possible to distinguish species. In addition, there are many problems with observer bias for live counts. Even experienced observers over and underestimate numbers of bats exiting depending on the number and speed of the exit (D. Dalton, unpub. data). Videotaping, which eliminates observer bias, does not provide an adequate field of view to cover exit areas for large bridges, and supplemental infrared light, which is required, would not reach far enough to illuminate the entire area.
For each species, I combined results for all bridges to show monthly fluctuation of abundance over the year; however, the monthly combined totals may not be accurate, 95
especially for Mexican free-tails, because this species occupied so many bridges. I assumed there was no movement among bridges within the 2-week survey time each month, but I have no way to verify this. If bats moved from one bridge to another within the survey period, it is possible I double counted or missed some individuals. In Texas, some banded Mexican free-tails moved from one building to another within a season (Short et al. 1960). Although I surveyed the bridges each month in the shortest time period possible (≤ 2 weeks) to minimize the likelihood of double counting or missing bats, I do not know if local roost switching occurred within the survey period.
Survey frequency
For all species, there was surprising variation in month to month abundance at bridges. Monthly surveys appear to be the minimum frequency required to determine gross seasonal changes in abundance. Surveying only once a season is likely to produce misleading data. For example, surveys in July, when one would expect bats to be present if they used a bridge, would have found no bats in 4 bridges. These same bridges held colonies of 200-650 bats the rest of the year. Surveys only once in the spring and fall could miss species using a bridge as a migratory stopover roost because bats roost there for a relatively brief time. For example, Mexican free-tailed bats arrive at Carlsbad Caverns at different times every year; changes in local weather patterns seem to affect their migration (Constantine 1967). I found some Tucson bridges occupied only in April, and would have thought they were never occupied had I surveyed only once in March or May. 96
Results for the few bridges at which I conducted frequent surveys indicate, that for pipistrelles, even monthly surveys are insufficient to provide an accurate picture of presence-absence and changes in abundance. Pipistrelle numbers at a bridge fluctuated almost daily, and although most changes in abundance were small, there were sometimes dramatic changes within a few days. Presence of a species that uses a bridge in low numbers, such as the cave, Yuma, and California myotis I observed at River Rd (9557) bridge, might also be undetected by surveys only once a month. Depending on the date of the survey, I would have missed the presence of both Yuma and California myotis in several months at River Rd bridge with only one survey per month. For a species such as cave myotis, which used bridges in the spring and fall, more frequent surveys would have more likely captured true peak numbers, and given more information on when and how bats departed in the summer and returned in the fall.
Annual variation in bridge use
Results on the presence of a species, its abundance, and seasonal use of a bridge are based primarily on survey results from 1999. Although these results certainly provide insight into how a bridge is used, they provide only a one-year snap-shot picture. Abundance, seasonal use, and even presence of a species could vary from year to year. For instance, 200+ free-tails roosted in Ft. Lowell (7581) bridge in 1999 from January through April. Numbers decreased in May and June to 50 bats, and no free-tails were present the rest of the year. In March of 2000, there were 50-80 bats and no guano underneath the colony. Because there was little rain in January and February, the absence of guano suggests that the colony had arrived recently, and was not present 97
during January and February (even occasional foraging flights would have resulted in some guano accumulation). The colony decreased to < 20 in April, then increased to 200 in June. Unfortunately, I did not monitor the bridge after June to determine if bats remained during the summer, but different January-June trends between 1999 and 2000 are evident. Variation in annual activity patterns has been observed elsewhere. Mexican free-tailed bats do not use Carlsbad Cavern as a maternity roost every year; the use of bridges near the cavern varies also (Constantine 1967). Some sites may be “spill-over sites” and occupied only in years when abundance at near-by roosts becomes unusually high (Fraze and Wilkins 1990). Other factors could influence the abundance and even presence of a species at a particular site from year to year. Exclusion from buildings in the area and local temperature patterns in Tucson could affect bridge use patterns. For bats migrating through the area, weather patterns at their departure site could also affect their use of Tucson bridges.
PROPORTION OF BRIDGES USED BY BATS IN TUCSON
Of 238 bridges surveyed (every bridge in the study area), 18% (43) were constructed of parallel concrete beams or slabs with vertical crevices between the beams; these bridges were selected for study because of their high potential for use as day-roosts. I found bats day-roosting in 35 (81%) of the 43 bridges. If the 5 bridges with crevices too narrow or shallow for any species to occupy are excluded, occupancy increases to 92% of 38 bridges. 98
Seventeen percent (41) of bridges in the study area were steel. I found no evidence that bats used these bridges; other surveys have also found no use (Keeley and Tuttle 1996).
The remaining bridges (65%) were constructed of concrete, but did not have vertical crevices between beams. These bridges were excluded from study (discussed in the methods section), but a few of these may be occupied by bats. Construction types included single slabs, I-beams, T-beams, and hollow box girders. In some of these bridges, transverse expansion joints provide potential day-roost sites (Figure 3). I found bats (species unknown) roosting in an expansion joint in one bridge over a wash; other researchers have found colonies of several species in this type of roost site (Davis and Cockrum 1963, Sidner 1997). It is possible that other bridges with expansion joints in the study area house bats. However, most of these bridges span heavily traveled streets, and I think they are less likely to be used than those that span washes.
Another potential day-roost site for individual or small numbers of bats is the horizontal crevice between a beam and bridge pier, if the space is not sealed (Figure 3). I found a single western pipistrelle in such a crevice in the same bridge where bats occupied the expansion joint. Individual or small numbers of bats may also use covered drain pipes, 10 cm in diameter (Keeley and Tuttle 1996). I found covered drain pipes in 8 bridges; one pipe was occupied (species unconfirmed).
The exact total proportion of bridges occupied by day-roosting bats in the Tucson area is unknown, but is at least 16% (the proportion of bridges I found occupied). Thorough surveys of the bridges in the study area with the potential roost site structures 99
described above may result in additional day-roost locations. However, I think it is unlikely that the total proportion of occupied bridges would be greater than about 20%, based on results from this study on the characteristics of bridges associated with occupancy by bats. As has been found in other studies in the U.S. (Keeley and Tuttle 1996, 1999), the majority of bridges in the study area are not suitable for use as day- roosts, but there is high occupancy for those bridges that are suitable.
Unoccupied bridges
I never observed bats in 8 of the 43 bridges I studied. For most of these bridges, the explanation appears straightforward and simple. Based on the crevice dimensions used by western pipistrelles, the smallest bat species in the U.S., it appears that the crevices in 5 bridges (8449, 8460, 9029, 9513, JW44) are too narrow or too shallow for even this species to use. One of the remaining unoccupied bridges (577) spans heavy street traffic, which bats seem to avoid (see discussion on ‘feature under bridge’ ).
The absence of bats from 2 bridges is puzzling. Grant (9017) has suitable crevice dimensions for pipistrelles and cave myotis. However, it is adjacent to industrial properties and may be avoided for that reason, or because of some other factor not considered in the study. Both pipistrelles and cave myotis occur in the Tucson Mountains, <10 km to the west; however, there is no natural corridor, such as a wash or undeveloped land, from the foothills to the bridge. Why I never found bats in Airport east bridge (7844), is also unclear. It is <200 m from Airport west (7843) bridge, which was used incidently by cave myotis in 1998 and 1999. Crevices appear too narrow for big browns and too low for free-tails, but suitable for pipistrelles and cave myotis. It is 100
possible that the bridge was used occasionally and briefly by transient bats, and that surveys once a month were insufficient to detect such use. Crevices in the east bridge are somewhat shallower and lower than in the west bridge; perhaps this is why only the west bridge was occupied. It is also possible that the area surrounding the airport bridges is too highly urbanized to provide adequate foraging areas. Like Grant, there is no corridor to outlying areas, and in addition, the airport is farther to the nearest mountains. Perhaps, considering these factors, the anomaly is that one of the bridges was occupied at all.
BAT SPECIES FOUND IN BRIDGES
Of the 10 crevice-dwelling species commonly found in Sonoran desertscrub (Hoffmeister 1986) and likely to be found in the Tucson basin area, 7 used bridges as day-roosts: Mexican free-tails, western pipistrelles, cave myotis, big browns, pallids, Yuma myotis, and California myotis. A silver-haired bat, usually found at higher elevations, also roosted in a bridge (presumably while migrating). Pocketed free-tailed bats (Nyctinomops femorosaccus), big free-tailed bats (Nyctinomops macrotis), and western mastiff bats (Eumops perotis) were not found during surveys. It is possible that pocketed free-tailed bats were present and I misidentified them as Mexican free-tails; I would not be able to distinguish the 2 species in crevices even with binoculars. Although every captured free-tail was a Mexican free-tail, I did not capture bats at every bridge occupied by free-tails. Pocketed free-tails usually form small colonies of <100 bats (Barbour and Davis 1969) and would be easy to miss among the thousands of Mexican free-tails. I am fairly certain I would have noticed a big free-tailed bat, and very sure I would not have misidentified a mastiff bat. It is probable that the latter 2 species are too 101
large to use bridge crevices in Tucson; Mexican free-tails are much smaller and occupy the widest available crevices in most bridges. Western mastiff bats have been reported in Tucson, but they typically occupy cliff crevices ≥ 50 mm wide (Barbour and Davis 1969). Only 3 bridges have crevices > 50 mm wide, and among them, only 22 meters of crevice that wide exist.
Abundance of bats
As expected, Mexican free-tailed bats were the most abundant and widespread species found roosting in Tucson bridges, occurring in numbers from a single bat to over 11,000 in a bridge, and found in 32 bridges. They are the most commonly found and most abundant species in bridges across the country in the southern part of the U.S. (Keely and Tuttle 1999). The number of pipistrelles found roosting in bridges was unexpected; this was the second most frequently found (15 bridges) and abundant species, occurring in numbers from a single bat to > 160 in a bridge. Pipistrelles are easily overlooked because they roost singly, switch roosting places frequently, are quiet, and leave little or no guano to draw attention to roosting places. Before this study, the species was documented at only one bridge in Tucson (S. Ruther, AGFD, pers.comm.). Although pipistrelles have been found roosting in bridges in other geographic areas (Keely and Tuttle 1999), Tucson may be a hot spot in the country for bridge-roosting pipistrelles.
Cave myotis was the third most abundant and frequently found (10 bridges) bat, roosting in bridges in numbers ranging from a single bat to >200 bats. Keeley and Tuttle (1999) also found this species to be the third most abundant bat in bridges across the 102
southern and western U.S.. Most cave myotis in Tucson bridges roosted singly or in small groups of less than 10 bats. Keeley and Tuttle (1999) also found this pattern, although they also found large nursery colonies.
I had expected to find many more big brown bats roosting in bridges; this species is commonly found in buildings in Tucson (Dave Purwin, Scott Richardson, pers. comm.) and was the second most abundant species found in Keeley and Tuttle’s 1999 study. I found big browns in only 7 bridges in rather low numbers from 1-88 in a bridge.
Small numbers of Yuma and California myotis were found; each species used only 2 bridges. It is possible that more frequent surveys would have resulted in more detections of these species, but their use of bridges in the study area seems limited. More frequent surveys may also have detected more transient bats, such as the single silver-haired bat observed.
Colony type
Bridges were used as maternity roosts by big browns, pallids, and pipistrelles. I did not capture any male big browns in June or July at maternity sites. Although they are known occasionally to roost in maternity colonies (Davis et al. 1968), male big browns typically choose cooler areas of the roost (Hamilton and Barclay 1994) or choose cooler locations or altitudes in summer (Easterla 1973). I also did not capture any adult male pallid bats, except for one at the end of July. Segregation of males and females in the summer is common for this species during the summer (Hall 1946, Beck and Rudd 1960, Vaughan and O’Shea 1976, O’Shea and Vaughan 1977), although it apparently does not 103
always occur (Orr 1954). Male pipistrelles were present in bridges all summer, but rarely roosted near female aggregations.
Both male and female Mexican free-tailed bats were present at bridges during maternity season, although sexes segregated within at least some bridges. This species uses bridges as maternity roosts across the southern U.S. from Florida to California (B. Keeley, R. Miller, pers. comm.). However, the extent to which Mexican free-tailed bats used Tucson bridges as maternity roosts is unknown. I found only small numbers of young; it is possible I simply failed to see young that were present because they are hard to find (B. Keeley, pers. comm). Further research is necessary to determine whether the decrease in free-tail abundance at many bridges during July was due to females leaving to give birth elsewhere, as has been found in Texas (Fraze and Wilkins 1990), males leaving for cooler sites (Constantine 1967), or both. Cockrum (1969) documented a large decrease in both males and females at a bridge in southern Arizona in late June, 1959, and July, 1960. It is also possible that Tucson bridges are used in some years for maternity sites and not used in other years, as is the case in Carlsbad Caverns (Constantine 1967).
Winter colonies of free-tails appear to comprise both sexes. I found some evidence of segregation within a bridge and a greater abundance of males but sampling was limited to a few bridges and sample sizes were small. Further research, therefore, is necessary to determine to what extent winter colonies of free-tails exhibit sexual segregation and if more males than females over-winter in Tucson bridges. 104
Cave myotis was the only species (except for the single silver-haired bat) that used bridges only for spring and fall transient roost sites. My observations seem to reflect patterns observed by others in Arizona that males migrate north first in the spring (Hayward 1961) and that females in August leave maternity colonies to join males (Hoffmeister 1986). Bats sampled in April at a bridge were 89% male; in May 60% were male. At 2 bridges in August, females accounted for >85% of cave myotis sampled.
Seasonal patterns
Mexican free-tails were present throughout the year, and although I have no conclusive evidence that different populations of free-tails occupied bridges in the winter and summer, it seems likely. Abundance dropped dramatically in April and November throughout the study area. During these 2 months, free-tails roosted in small groups or as individuals scattered throughout a bridge, and several bridges were occupied that were not used the rest of the year. These observations suggest migration to and from the Tucson area.
Pipistrelles also occupied bridges year-round although abundance was lowest in the winter. Many researchers have netted pipistrelles during the winter at low temperatures (e.g. O’Farrell et al. 1967, Ruffner et al. 1979), but few hibernating bats have been found (Cross 1965, Hoffmeister 1986). Although Cross (1965) observed only male activity in winter in Sabino Canyon northeast of Tucson, I observed 2 banded females as well as 6 males roosting in a bridge in winter. Because these bats were periodically absent from the bridge and individuals were observed using different crevices, they must have flown occasionally. This observation supports other 105
researchers’ results finding both sexes active, but more males than females (O’Farrell et al. 1967, Ruffner et al. 1979).
There is an interesting inverse correlation between seasonal trends May-September of a colony of cave myotis residing in a mine outside of Tucson and cave myotis roosting in the bridge nearest the mine (Figure 20). The build up at both roosts in April and May suggests an influx into the area, which is typical for this species. The bats at the bridge make up only a fraction of the mine population (maximum number at mine = 6100 in July, maximum number at bridge = 120 in September), and therefore cannot account for the entire increase in the mine population in the summer. Further research is necessary to determine whether bats in the bridge moved on to a summer roost site farther away, or contributed to the nearby mine population.
VARIATION AMONG INDIVDUAL BRIDGES IN USE BY BATS
Species diversity differed among bridges, although I found only one or 2 species in most (74%) of the 35 occupied bridges. Only 3 bridges occupied by bats were not used by free-tails; these 3 bridges were used primarily by cave myotis.
A few bridges stand out for their high diversity. River Rd bridge (9557) is small in size (199 meters of crevice) but supports 5 species. It is the only bridge at which I found Yuma and California myotis regularly, and ranks first in pipistrelle density. Its crevices are too shallow and narrow for large colonies of free-tails, but small numbers roost there occasionally. Cave myotis are rare visitors. This bridge is located on a small wash lined with native vegetation, adjacent to low density housing and natural open 106
space. One of the 3 major washes in Tucson is <1 km to the south. The Santa Catalina Mountains are
~ 6 km to the north, but the foothills extend to just north of the bridge.
00.51JanFebMarAprMayJunJulAugSepOctNovDec# bats normalizedMineBridge
107
Figure 20. Comparison of monthly trends of cave myotis in a mine and in the bridge nearest to the mine in Tucson, Arizona, 1999.
Tanque Verde Ranch bridge (TVR) is also small (79 meters of crevice). I found 4 species here; it ranked first in both free-tail and cave myotis density, as well as third in cave myotis abundance. The bridge spans a small running creek, is near Tanque Verde River and ~1 km from the Rincon Mt. foothills, and is near the boundary of Saguaro National Park. I found only one big brown and a couple of pipistrelles in the bridge, but both species forage in the area (they have been mist-netted nearby). The bridge is only about 4 km from a residential area with big brown colonies (pers. obs.). I believe the reason the bridge is not occupied by more big browns is that the crevices are too narrow.
Free-tails may prevent more use of the bridge by pipistrelles. Free-tails usually take up most or all the crevices, with cave myotis roosting among them in the spring and fall. Because pipistrelles do not roost near large groups of bats (based on my observations), the presence of free-tails seems the most likely explanation for the low use of this bridge by pipistrelles. Based on the bridge’s location and its adjacent land use, I also think it would probably be occupied by Yuma and California myotis if free-tails were absent.
Houghton RR bridge (9563) is much larger (725 meters of crevice) than the other 2 bridges, but is far from the largest of the 43 bridges studied (10 bridges have >1000 meters of crevice). This bridge spans heavily used railroad tracks and has one of the 108
highest vibration levels (D. Zaleski, Pima County Dept. of Transportation, pers. comm.) of bridges in the study area. Yet, 4 species roost here; it is the only bridge with pallid bats, has the most cave myotis, and is second and fourth in big brown and free-tail abundance, respectively. Although it is presently in an outlying area surrounded by natural open space composed primarily of creosote and mixed cholla, high density subdivisions are being built within a couple of kilometers. The land around the bridge is flat but the Rincon Mts. are ~ 9 km to the east. It is interesting that pipistrelles do not occupy the bridge. Perhaps it is because there is no wash or other natural corridor to lead foraging bats from the foothills to the bridge.
BRIDGE CHARACTERISTICS
Crevice dimensions
Assuming the macrohabitat features of the area surrounding the bridge are attractive to a species, whether bats occupy the bridge is at least partially determined by the dimensions of the crevices. For example, a bridge that is ideal in every other factor can not be occupied if the bat cannot fit in the crevice; width of the crevice is therefore important. The minimum width of a potentially occupied crevice is determined by a bat’s body size, and therefore differs among species. Minimum width for colonial species is determined by the preferred packing density of a cluster of bats. Free-tails often roost on either side of the same crevice, back to back, and would therefore need a minimum width of at least twice an individual’s body size. Minimum width of crevices used by individual free-tails, as expected, was narrower than widths used by colonies (Table 7). 109
The maximum width of an occupied crevice may be determined by how much extra space an individual bat prefers or the arrangement of bats in a cluster.
The depth of a crevice needs to be at least as deep as the bat’s body length for minimum protection and probably deeper according to a species’ need for protection. For colonial bats, the depth of a crevice determines if the bats can roost in multiple vertical layers. Depth may also determine the temperature gradient available from the top to bottom of the crevice. Height of the crevice above the ground determines the distance to human disturbance under the bridge, protection from terrestrial predators, and the vertical distance available for bats to enter the crevice from below and to drop when exiting.
I measured places where bats roosted for free-tail colonies, free-tail individuals, pipistrelles, cave myotis, and big brown bats. Reported dimensions of crevices used by each species (Table 7) may or may not reflect the true preference of the species, and preferences may in turn depend on the type of bat colony or other characteristics of the roost site, such as human activity or thermal characteristics. For example, big brown maternity colonies used crevices shallower than the minimum depth (36 cm) I observed in bridge roosts; I found colonies roosting in 23-cm deep crevices in porches and carports. For other species, reported minimum depth used may be closer to the actual minimum depth used because I had larger sample sizes and measured roosting places in different bridge structure types.
Reported minimum height for pipistrelles, cave myotis, and free-tails is probably an accurate measure of the minimum height these bats will tolerate in bridge roosts. These species roosted in bridges with slopes at either end and therefore had a choice of 110
heights within a bridge; they also used bridges of different average heights. Sample size for big browns was too small to determine this species’ preferences for height.
Depth
All of the most abundant species (Mexican free-tail, pipistrelle, cave myotis, big brown) occupied channel beam bridges, which have the deepest crevices of the bridge types studied. The minimum depth bats used is therefore what is important and different among species. Minimum depths of roosting places varied from 16 cm for pipistrelles to 36 cm for big browns (Table 7).
Although average depths of crevices differed widely among the 43 bridges, ranging from 4-69 cm, crevice depths within a bridge were fairly uniform, rarely varying by more than a few centimeters. The few bridges with wide variation in depth provide additional insight into species preferences. In Trico (8262) bridge, flood debris in many crevices made depth an almost constantly changing variable; depths ranged from 0-28 cm. Although many areas were 8-15 cm deep, I found cave myotis only in places where there was no debris, at the maximum depth available of 28 cm. Shallower places provided rough enough surfaces to hang from, therefore, it appears the bats preferred the deepest areas. Although the minimum depth used by cave myotis in this study was 18 cm, this species will roost in shallower crevices; I found cave myotis in a long round concrete culvert in the shallow (< 10 cm) seams between sections.
Differences in the amount of expansion joint material at the top of crevices resulted in rather wide variation in crevice depths in 22nd (9034) bridge. Crevices in the north part of the bridge (40% of total crevices) are very shallow or completely sealed; the rest 111
of the crevices are an average of 52 cm deep. Up to half the pipistrelles I observed during monthly surveys roosted in the north section, in crevices 10-14 cm deep, as well as the deepest crevices in the south section. Cross (1965) also observed pipistrelles in cliff crevices as shallow as 13 cm and as deep as 92 cm. My measurements of places where pipistrelles roosted (Table 7) were biased toward 2 channel beam bridges (deep crevices) because these bridges had the most bats. It is possible that the reported minimum depth of 16 cm (the 2.5 percentile of the distribution of measurements) for this species is deeper than would be observed in more evenly sampled bridge structure types.
Of the factors I measured, crevice depth was the most important bridge characteristic for predicting the presence of large colonies of Mexican free-tails in bridges. Colonies roosted only in crevices at least 34 cm deep (Table 7). Free-tails in Texas bridges used crevices of similar depth, ≥ 30 cm deep (Keeley and Tuttle 1996). There seems to be sufficient evidence to conclude that use of deeper crevices is a true preference. Preference for the deepest crevices could be due to temperature profiles or greater protection from disturbance and predators. A comparison of temperatures in a shallow (21cm) crevice compared to a deep (59 cm) crevice showed very little difference in temperatures at the top of the crevice. A deeper crevice may have a greater thermal gradient available for bats to move vertically throughout the day, but from my observations, free-tails were almost always at the top of the crevice. Preference for deep crevices could be for protection from predators. Bats that congregate in groups of millions, such as free-tails, do not have the defense mechanisms of visual acuity and speed that help protect herd animals inhabiting large open spaces. Natural roost sites for 112
free-tails are caves; selecting roosts in bridges that minimize detection, such as the deepest crevices available, would be the closest available approximation to a cave. Free-tails also seem rather sensitive to some types of human disturbance (see discussion section on ‘human activity under bridges’). Their greater sensitivity to human activity relative to other species (Licht and Leitner 1967) provides additional evidence that preference for the deepest crevices is for protection.
In contrast to free-tail colonies, free-tailed bats roosting individually or in small groups did not select the deepest crevices and used shallower slab bridges in addition to channel and box beam bridges. Perhaps, because free-tails roosting singly or in small groups are not as vocal and do not move around as much as large colonies, they are not as vulnerable to detection and do not need the greater protection that the deepest crevices afford. Humans, which are bats’ most common predators, are unlikely to notice bats of any species roosting alone or in small groups (pers. obs.).
In summary, crevices of sufficient depth for pipistrelles, cave myotis, and individual free-tailed bats are widely available among bridges in the Tucson metropolitan area and therefore do not appear to be a limiting factor in roost selection of bridges for these species. Crevice depth may be a limiting factor for large colonies of free-tailed bats. Inserting spacers in crevices occupied by free-tail colonies to make the crevices shallower, and adding bat boxes with deep crevices to slab bridges, would provide additional evidence as to the importance of depth compared to other factors in roost selection for free-tail colonies.
Height 113
The 4 most abundant species that I observed (free-tails, pipistrelles, cave myotis, big browns) roosted in the highest bridges; therefore, differences among species lie in the minimum heights of occupied bridges. The minimum heights observed for pipistrelles, cave myotis, and big brown maternity colonies in bridges were 1.0 m, 1.2 m, and 2.3 m high, respectively. Bridges of sufficient height for these species are widely available throughout the study area and do not seem to limit the presence or distribution of these bats.
Mexican free-tail colonies occupied bridges with a higher minimum height (3.5 m) than bridges occupied by other species roosting in Tucson bridges. Although the presence of free-tails was not associated with height after accounting for depth and width of crevices, presence was associated with higher bridges when height was tested as an independent variable; therefore, height may be important in bridge selection. Further evidence of the importance of height comes from a comparison of 4 bridges along Pantano Wash and 4 along Rillito Creek. The Pantano bridges have a greater number of bats and, although all 8 bridges are >3.5 m high, the Pantano bridges are higher than the Rillito bridges.
In addition to their predilection for higher bridges, free-tail colonies seemed to prefer the highest available crevices within a bridge. They used lower crevices above slopes in the end spans, but only during months of greatest abundance and only when the rest of the crevice over the wash was filled with bats. In other months, only the higher parts of 114
these crevices were used. Keeley and Tuttle (1996) also found this species selected the highest available crevices in a bridge that had heights ranging from 0.3-7.6 m.
Why do Mexican free-tails roost in higher bridges? Members of the free-tailed bat family, Molossidae, have long, narrow wings compared to vespertilionids. When leaving a roost, such as a cave ceiling or bridge crevice, molossids need to drop a greater distance before achieving enough lift to fly horizontally. This may explain why free-tails roosted in higher crevices compared to the other species I observed in bridges. However, the Mexican free-tailed bat is the smallest molossid in Arizona and therefore needs less drop space than the larger molossids. Although Schmidly (1991) states Mexican free-tails use roosts > 3 m high to have sufficient room to drop for flight, my observations indicated they drop < 1 m. The drop for the slightly larger N. femorosaccus has been observed to be 1.2-1.5 m (Krutzsch 1944). Why, then, did Mexican free-tails roost in Tucson bridges at least 3.5 m high, and use the highest available crevices? It may be that higher roost sites afford greater protection against predators and greater distance to human disturbance under the bridge. Individual and small groups of bats roosted in lower crevices (minimum height 3.0 m) than large colonies; perhaps, as may be true of crevice depth differences, small numbers of bats are less easily detected and can afford to roost in lower, shallower crevices than large colonies. Perhaps, although free-tails need only about a meter to drop from the crevice, they prefer at least 2 meters of space below the drop distance, possibly to avoid ground-dwelling predators. Another hypothesis is that free-tails need a greater vertical distance to enter a crevice than is needed to achieve lift on departure. 115
Width
Measurement and analysis: Measurements of widths and width interpolations provide a basic description of bridge crevices. However, width can vary from the bottom to the top of a crevice; in most of these cases the crevice tapers at the top. Because I could only measure width at the bottom of the crevice, the width available to a bat at the top of a tapered crevice would not be accurately represented.
To make comparisons among bridges, I calculated a mean crevice width for each bridge from measurements and interpolations for each meter of each crevice. Bridges differ greatly; mean widths ranged from 4-33 mm (Appendix B). Only one (9565) of the 5 bridges with a mean crevice width < 10 mm was occupied by bats of any species.
Unlike depth and height, which are generally uniform, widths can differ greatly among crevices within a bridge, and usually even along the length of a single crevice. The usefulness of mean crevice width is therefore limited to comparisons among bridges on a gross scale (as described in the previous paragraph). It is not a valid descriptor of crevices that are available or not available to bats for 3 reasons. First, bridges with the same mean crevice width can have different ranges of widths. River Rd (9557) and Pima (9790) bridges both have a mean width of 11 mm (Figure 21), however, River Rd has crevices up to 35 mm wide and Pima has crevices up to only 17 mm wide. Second, distributions of widths are rarely Gaussian and differ among bridges with similar mean widths and ranges. For example, in 2 bridges with mean widths of ~ 10.5 mm and widths ranging from 0-23 mm, 10% of total crevice length is ≥ 19 mm wide in Ft. Lowell (7581) bridge, in contrast to only 2.5 % of crevice length ≥ 19 mm wide at Airport east 116
(7844). Finally, no descriptive statistic describes the spatial configuration of different widths among the crevices in the bridge. If an individual crevice is fairly uniform in width, the entire crevice will be available to a particular species, or not, depending on its width. If the crevice is wedge shaped, all, part, or none of the crevice may be available. If all the crevices in the bridge are similar, availability throughout the bridge is uniform; however, in most bridges, crevices are dissimilar. It is not uncommon for a bridge to have some crevices entirely < 5 mm wide, some with the widest widths available, and some in-between. Availability to bats may depend on the spatial arrangement of widths. For example, of 14 crevices in
117
Ft. Lowell, 10 are <15 mm wide and unavailable to large groups of free-tails. However, the bridge was occupied by ~200 bats because one crevice was almost entirely ≥ 20 mm in width.
To determine if an association existed between width of crevices and use of the bridge by a species, mean crevice width of the bridge as an explanatory variable was unsatisfactory for the reasons discussed above. Instead, I used a binomial variable that indicated if the bridge had > 25 meters of crevice length within the width range (obtained by measurements of roosting places) used by each species. The choice of 25 meters was somewhat arbitrary because I do not know the minimum number of meters of a certain width range necessary for a species to occupy a bridge. It probably depends on the species, and for colonial bats, perhaps the size of the colony. However, 25 meters, even if not contiguous, seems adequate for occupation, and it is well below the number of meters of total crevice length for the bridges studied, which varied from 70-2525 m. Unfortunately, I was only able to test this variable with free-tails. The presence of free-118
tails was associated with bridges with > 25 m of crevice length 15-34 mm wide, but not significantly after accounting for crevice depth and height (Appendix D).
For another explanatory variable, I used the number of meters of crevice in the bridge that were within the width range used by a species. Using this value was more representative of what was available to bats than mean crevice width. For instance, both River Rd (9557) and 6th (1542) bridges have the same mean width and are about the same size in area, but have 23 m and 1 m of crevice 21-45 mm wide (the width range used by big brown maternity colonies), respectively. A disadvantage of using this variable was that the number of meters of crevice in a certain width range in a bridge is correlated to varying degrees (from 52% for big brown maternity colony range to 97% for pipistrelle range) with the size of the bridge (the total number of meters of crevice). The presence of big brown maternity colonies was associated with bridges with more meters of crevice 21-45 mm wide (Appendix G). I believe that crevice width is indeed a factor in roost selection by big brown maternity colonies because of the lower correlation between the explanatory variable and bridge size; many large bridges do not have wide crevices. In contrast, although there was conclusive evidence that the number of free-tails increased in bridges with more meters of crevice 15-34 mm wide (Appendix D), this result is not biologically significant because of the high correlation (80%) between the variable and bridge size.
Neither variable, ‘>25 meters available within a given width range’ or ‘number of meters within a given width range’, was entirely satisfactory for exploring the association between width of crevices and use by bats. In addition to the problems discussed above, 119
neither accounted for the spatial arrangement of crevice widths throughout the bridge, and the implications for roost selection due to spatial configuration are unknown. For all of these reasons, it was difficult to determine the relative importance of crevice width compared to crevice depth or height, or to other explanatory variables.
Widths used by bats: Unlike depth and height of crevices, which are fairly uniform within a bridge, width of crevices varies; therefore, not all crevices are available to a particular species. Bats of the species I observed appear to have a preferred maximum width as well as the minimum width that was determined by size of the species. The range of widths I observed for Mexican free-tails, big browns, cave myotis, and western pipistrelle were similar to those Keeley and Tuttle (1999) observed (Table 19).
My results illustrate the differences in widths used by individuals and groups of Mexican free-tails (Table 7). The difference in minimum widths used by big brown maternity colonies and all big browns in September may be due to the addition of individuals in the September measurements. It is also possible that pregnant bats require a wider crevice than those bats roosting in bridges temporarily in late summer. The difference may also simply be due to low sample size for maternity colonies. Reported widths for cave myotis are representative of individuals or small groups of 2-4 bats that roosted side by side, although I included the measurements for the few, densely packed larger groups I observed.
The crevices that free-tail colonies occupied were among the widest crevices available in every bridge; for all occupied bridges, mean width of occupied crevices was larger than mean width for all crevices in the bridge. For example, in Tanque Verde 120
(9569) bridge, mean crevice width for all crevices was 10 mm (± 0.1 SE); mean width for areas of crevices used by free-tail colonies was 17 mm (± 0.2 SE).
The availability of crevices of sufficient width may be a limiting factor for big brown maternity colonies; the probability that a bridge would be occupied increased with increased number of meters of crevice 21-45 mm wide. All 5 bridges that housed
maternity colonies had > 25 m of crevice 21-45 mm wide; about half the bridges studied
had < 25 m.
Table 19. Crevice widths used by bat species roosting in bridges: a comparison of results from 2 studies. Measurements are in millimeters.
Big brown
Mexican free-tail
Cave myotis
Western pipistrelle
This study
21-45
15-34
11-31
8-29
Keeley and Tuttlea
19-38
13-32
13-25
13-25
a Keeley, B. W., and M. D. Tuttle. 1999. Bats in American bridges. Bat Conservation International, Austin TX www.batcon.org/bridge/ambatsbridges 2/18/2000
Conclusions Widths of crevices vary widely within and among Tucson bridges. Narrow crevices, within the width ranges selected by pipistrelles and cave myotis, are widely available and do not appear to limit the number or distribution of bats of these species among bridges in the study area. The availability of wider crevices in and among 121
bridges decreases as width of crevices increases. The availability of crevices of sufficient width may be limiting to some extent for large free-tail colonies and to a greater extent for big brown bats.
Structure type
Crevice dimensions are time consuming and difficult to measure. If structure type of the bridge determined crevice dimensions, structure type could be used to provide an estimate of crevice dimensions. The structure type of the bridge, or the type of beam used, can be easily determined by a quick site inspection or by contacting a bridge inspector. I investigated if and how beam types were correlated with crevice dimensions, and if structure types could be used to identify potential roost sites for a particular species.
The 43 bridges in the study were made of concrete beams of 4 types: channel, box, slab, and double-stem (Figure 5). Almost half the bridges were channel beam and about a third were box beam. Crevice dimensions differed among the structure types. Crevices between channel beams were wider and deeper than crevices between other beam types. Box beam crevices were deeper than slab or double-stem beam crevices. Although there was no significant difference, box beams tended to be wider than slabs and double-stem beams. Slab and double-stem beam crevices did not differ statistically, but slab beam crevices tended to be both wider and deeper than double-stem beams. Channel beam bridges were higher than box beams, though not significantly, and both were higher than slab bridges. Variance was too great in the 2 double-stem beam bridges to differ in height from the other 3 types. In summary, channel beam crevices were the widest, 122
deepest, and highest, followed by box beam, slab, and double-stem beam crevices in descending order.
Other studies have found structure type significant in predicting use by bats (Keeley and Tuttle 1996, Gore and Hovis 1997, McDonnell and Clark 1999). These studies have included wood and metal structures, and also concrete culverts, as well as types of concrete structures. Unfortunately, in this study, the probability of a species occupying concrete bridges of a certain structure type could not be determined statistically because of low sample sizes; however, patterns are evident. Free-tails in varying numbers occupied every channel beam bridge; 75% of the large colonies roosted in channel beams. Pipistrelles, cave myotis, big browns, and pallids also used this structure type. Although it was the most common type, which may partly account for its frequent use, crevices in channel beam bridges had the widest range of widths available among the structure types, and therefore all species could find areas of appropriate width. Crevices in channel beam bridges were also the deepest and highest of all 4 structure types, well above the minimum requirements for each species.
Box beam bridges were also occupied by free-tails, although colonies of free-tails used only 33% of box beam bridges studied. Big browns occupied only one box beam bridge, which was constructed differently and had much wider crevices than any other box beam bridge. Although crevice depth is sufficient, it appears that narrower width prevents greater use of box beam crevices by both free-tail colonies and big browns. Pipistrelles and cave myotis, which used narrower widths than the 2 larger species, did not use box beams less than channel beams. 123
Of the 8 species I found roosting in bridges, all but big browns and pallids occupied slab bridges. Because slab crevices are narrower than box and channel beam crevices, big browns may not be able to use this structure type. Although I found many individuals and small groups of free-tails roosting in slab bridges, I found no large colonies; crevices may be too shallow for large colonies to use. In addition, crevice widths may be on the threshold for this structure type to support even individuals; the occupied slab bridges had wider crevices than the unoccupied.
I found Yuma myotis, California myotis, and the silver-haired bat only in slab bridges. Crevices appear to be appropriate for individuals of these species because they were present. However, because I found only small numbers of individuals, I cannot infer that the crevice dimensions they used were optimal. I have no data for how structure type, compared to other factors, affects use of bridges by these species.
I never saw bats of any species in the 2 double-stem beam bridges. With mean crevice widths of 6 and 7 mm, and mean crevice depths of 9 and 4 cm, crevices were just too narrow and shallow to be adequate roost sites. I believe they are unsuitable even for pipistrelles, however, neither bridge was in the part of town where pipistrelles roosted. If the bridges were in the part of the study area inhabited by pipistrelles, it would be interesting to see if, although almost all crevices were unavailable, these tiny, solitary bats would use the few places in the bridge where they could fit.
The results from this study cannot be inferred to bridges of these structure types in other areas, even in Arizona, where construction practices may differ. More research in other geographic areas is necessary before generalizations can be made for use of a 124
particular structure type by a species. However, if crevice dimensions are similar to those in this study, bridge structure type could be used to help plan surveys for an area (although the Bridge Record is not always helpful in denoting structure type). For example, double-stem beam bridges could be eliminated entirely (unless in likely pipistrelle habitat). If free-tails were present in the area, it would be most efficient to find large colonies by surveying channel beam bridges first, then box beam bridges. Slab bridges could be eliminated unless one were interested in use by small numbers free-tails. Initial surveys for big browns would concentrate on channel beams. If the objective was to locate the 3 myotis species or pipistrelles, all channel, box, and slab beam bridges would need to be surveyed.
Feature under bridge
Because the vast majority (86%) of bridges spanned washes, statistically determining a relationship between the feature under the bridge and the presence or absence of bats was not possible. However, from my observations, it appears bats avoided the 2 bridges over heavy street traffic. Although appropriate crevice dimensions were available in these 2 bridges for all species I encountered, I found only 3 transient free-tails in one of the bridges. Two of these bats were probably migrating and possibly desperate for a roost; I found them in April on the morning of a rare Tucson snowstorm. Keeley and Tuttle (1999) also found that bats typically avoided roosting over busy roadways.
It is interesting that cave myotis roosted in all 3 bridges that spanned running water. Water does not appear to be a requirement, however, because the largest number 125
of cave myotis were found in the bridge spanning railroad tracks. The 3 bridges over water were also in outlying areas, and undeveloped adjacent land use was significantly associated with bridges occupied by this species. Further research is necessary to determine if cave myotis select bridge roosts with higher humidity, for nearby foraging areas, with less human disturbance, or a combination of these, or other, factors. Many, but not all, summer nursery roosts of cave myotis are associated with high humidity (Hayward 1961, Kunz 1973) and winter hibernacula found in southern Arizona were all wet mine tunnels (Hayward 1961).
Human activity under bridges
The effect on bats of human activity under bridges is difficult to determine. How a bat responds may depend on the type, intensity, duration, and proximity of the activity, the species of bat, and also the season (rearing young vs. winter). Human disturbance is believed to be a major cause in the decline of populations of many species throughout the country (McCracken 1986, McCracken 1988, Stihler and Hall 1993, O’Shea and Vaughan 1999). However, human activity does not necessarily cause disturbance to bats. Bats may ignore human activity or become habituated, or they may be temporarily disturbed but not suffer detrimental long-term effects. Bats that live in bridges in an urban environment seem to have tolerance to some degree of human activity.
I was unable to rank human activities on an ordinal scale of disturbance because I did not know how bats perceive the many different combinations of types and intensities of activities I observed. Low sample sizes prevented statistical testing of data placed in nominal categories based on my perceptions of disturbance types and levels. Because my 126
study was not designed to focus on this factor, there may be effects on species’ presence that are not evident; however, I found bats roosting in bridges I considered to have the most frequent and disturbing types of human activity observed in the 43 bridges, with one exception. As discussed in the previous section, bats avoided bridges over street traffic.
It appears that the many bats living in Tucson bridges have become habituated to the usual activities under their roosts. Although I did not mark any free-tails, big browns, or pallids, it is unlikely that the thousands of free-tails that roost in the same bridges year after year are different individuals every year, and big brown and pallid maternity colonies are known to show site fidelity, particularly in man-made structures (Brigham and Fenton 1986, Sidner 1997). Banded pipistrelles, Yuma myotis, and California myotis were present multiple years at the same bridge where they were banded.
Houghton Railroad (9563) bridge is intriguing because of high species diversity and abundance in spite of high disturbance. It has a high volume of train traffic (1-3 per hour), daily (but low volume) vehicle use of the dirt road right of ways, and frequent recreational use by off-road vehicles and bicyclists. It also serves occasionally as a party place. Vibration levels of the bridge are among the highest of bridges in the area, and truck traffic over the bridge has weight restrictions. Trains always blow their whistles when passing under the bridge. Although the center span over the tracks is high (8 m), the beams above are black from train exhaust fumes. In spite of these circumstances, free-tails roost in the bridge year-round. Although the vast majority of free-tails avoided the center span over the tracks, there were usually a few roosting at the ends of the crevices in this span. I saw more big brown bats directly above the tracks than in other 127
spans, and cave myotis occasionally roosted there. In contrast, pallid bats roosted in 5 places in the bridge; all of these places were at or near the ends of the bridge over vegetation, and had the least human activity underneath. Trains did affect bats’ behavior, but only briefly. During an evening outflight I observed, 2 trains went by. Each time, the outflight stopped in all spans, then resumed several seconds after the last car passed under. What makes this bridge so appealing to bats is just speculation. For free-tails, it provides one of the highest roosts (among bridges) in the study area. For the other 3 species, perhaps the surrounding natural open space provides good foraging areas. Although bats are known to roost in many bridges spanning railroads, they do not roost where there are black fume stains (B. Keeley, pers. comm.) Why some bats tolerate crevices with exhaust at this bridge is unknown. It may indicate that roosts are limiting in the Tucson area.
My observations of the roosting patterns of bats within a bridge suggest there are differences among species’ tolerances of human activity. Pipistrelles appear to be rather tolerant of at least certain types of human activity. In the bridges with very frequent activity (Table 12), pipistrelles roosted in areas above the bicycle-walkways in proportion to availability in total bridge area (the walkways were closer to the crevices and received more activity than any other part of the bridge). In other bridges, pipistrelles also roosted over ledges and slopes that were much closer to crevices than the floor of the wash and were therefore closer to people underneath the bridge. This species is commonly found at private residences in Tucson, where they roost between fascia 128
boards and walls of porches and are exposed to the daily activities of families (pers.obs., D. Purwin, pers. comm.).
Big brown bats also seem tolerant of many types of human activity. One colony tolerated the construction of a new bridge about 30 meters from their bridge. Heavy equipment drove in the wash between the 2 bridges constantly, and also drove under their bridge. Construction lasted from spring through fall and the colony returned the next year. In Houghton Railroad (9568) bridge, a late summer roost, big browns roosted directly above the tracks more than in the other spans. It is possible that they were limited to this area because of the large number of free-tails at the bridge, but it is unlikely. I often saw big browns interspersed among free-tails at other bridges; they could have done so at this bridge. It is also possible, that although the disturbance was tolerated by these transient big brown groups, a maternity colony would not choose to raise young under the same conditions.
Big brown bats are commonly found at residences in Tucson (pers. obs.) and appear to coexist successfully, if allowed to, with nearby humans. However, as with every species, there are limits to tolerance. The colony at 22nd (9034) bridge roosted over a ledge only a couple of meters high and left the bridge between July 20 and August 8, 1999. I observed people directly under the roost on more than one occasion, heard reports of children throwing rocks up at the crevice, and found a dead baby big brown on the ground. I have no direct evidence, but a correlation between disturbance and abandonment appears to exist. 129
Mexican free-tails may be more sensitive to disturbance than some other species. Licht and Leitner (1967) found that free-tails in a barn loft retreated when the researchers approached, whereas pallids and Yuma myotis showed little disturbance. At Houghton Railroad (9563) bridge, free-tails avoided the span over the tracks, and used crevices in spans with less frequent disturbance. I never observed free-tails roosting over the bicycle-walkways in bridges; they roosted in crevices over the wash that had less activity underneath. As discussed in the section on crevice dimensions, the association found between the presence of free-tails and the deepest and highest crevices may be explained by the protection offered by these crevices.
Bats that use Tucson bridges as temporary roosts, such as cave myotis and bats migrating through the area, may be more sensitive to the disturbances to which resident bats have become habituated. Conversely, transients may use bridges with even higher levels of disturbance because they are present for only a short time. My observations are inconclusive. Of 6 primary use bridges occupied by transient cave myotis, 4 were categorized as receiving only rare disturbance. Concluding that cave myotis tend to avoid bridges with the most disturbance may not be the correct interpretation of this observation. Presence of this species in Tucson is associated with bridges in outlying areas adjacent to undeveloped land, 3 of which span running water. These bridges are less likely to have activity under them than bridges in more populated areas or over dry washes. The factors influencing bridge choice, therefore, could be undeveloped land, perhaps for foraging reasons, roosts over water to gain higher humidity, less human disturbance, or a shorter distance to summer and winter roosts than in-town bridges. 130
Researchers must attempt to prevent or minimize disturbance to the study animal and yet obtain good quality data if the objectives of the study are to be fulfilled. I was aware that surveying with a spotlight could be disturbing, as bats are sensitive to bright, white light (Mann 1999). However, a red light, though less disturbing, was inadequate for counting bats in crevices. I minimized disturbance by lighting bats for as brief a time as possible to obtain necessary data, often 1-2 seconds, and by being quiet. Although I cannot assume the bats that showed no physical reaction were not affected, the majority of bats did not react overtly. Bats that reacted did so by opening their mouths and baring their teeth (Myotis spp.), moving a wing or leg, and occasionally crawling away (free-tails), but none ever flew. I found no evidence of anything but momentary irritation, as would be expected if a bright light was suddenly shone in one’s eyes. It is possible that my disturbance was minimized because bats roosting in bridges are used to higher ambient light levels than bats dwelling in caves or mines. Bats in bridge crevices do not ever experience complete darkness during the day. Ambient light levels vary but can be quite high, particularly under a small or long, narrow bridge.
Adjacent land use
Most (79%) of the 43 bridges were adjacent to developed land; therefore I could not determine if presence of bats in bridges adjacent to developed land was associated with adjacent land use. The presence of cave myotis, however, is associated with bridges adjacent to undeveloped land. As discussed in the previous section on human activity, there are several untested factors that may contribute to this statistical association. 131
I expected to find more pallid bats roosting in bridges in the study area; however, I found them in only one bridge that was adjacent to natural open space. This species eats large insects which are flightless, low flying, or that often light on vegetation (O’Shea and Vaughan 1977). Pallid bats forage by flying < 2.5 m above the ground (O’Shea and Vaughan 1977), or by lighting and then crawling on the ground, which they do well (Orr 1954). It is not surprising, then, to find pallids roosting in an undeveloped area. However, there are 6 other bridges in the study area also adjacent to agricultural land or natural open space. In addition, bridges that may be adjacent to developed land, but within commuting distance to foraging areas, should also be appropriate roost sites. All but one of the other bridges in undeveloped areas are slabs and box beams, and probably do not have crevices of sufficient width for colonies of these rather large bats. However, CDO Railroad (CDO) bridge, which does have wide crevices and supports a colony of big browns, is 30 meters from the site of a bridge (replaced in 1999) that had a maternity colony of pallids 1959-1960, but which has not been seen in recent years (R. Davis, pers comm.). Although CDO is adjacent to abandoned agricultural land, it is also now partly adjacent to commercially developed land. The state road of 1960 is now an interstate highway, and the area surrounding the bridge has become much more urbanized in the last 50 years. It is possible there are insufficient foraging areas because of increased urbanization, or the level of disturbance in the area is too high for a colony to persist.
Bats occupied bridges adjacent to commercial and high density residential properties, however, they apparently avoided industrial areas. I never saw any bats in one of the 6 bridges adjacent to industrial land; the other 5 were used only incidentally by 132
free-tails and a couple pipistrelles. Of 15 bridges along a concrete channelized wash, free-tails occupied 9 of 10 of the western bridges. Six of these 9 were primary use bridges. On the eastern end, although all 5 bridges were occupied, only 2 bridges had > 3 bats at one time, and that occurred only once at each bridge. All 15 bridges are channel or box beam, have no vegetation, and receive about the same amount of disturbance. The bridges on the east side are adjacent to industrial properties, those on the west side are adjacent to commercial or high density residential properties. Because free-tails can fly so far to forage, lack of vegetation and insects in industrial areas would not prevent this species from roosting there. Further research is needed to determine why bats did not use the eastern bridges.
Vegetation obstructing access
Vegetation along the sides of a bridge could obstruct access and prevent bats, particularly less maneuverable species, from roosting in the bridge. Mexican free-tailed bats (one of the less maneuverable species) in Texas were more likely to roost in bridges spanning bare vs. vegetated ground (Keeley and Tuttle 1996). Bat Conservation International found greater occupancy in bat houses that were placed > 6 meters from obstacles to flight (Tuttle and Hensley 1997).
In the bridges I studied, the greatest amount of vegetative obstruction to access along the sides of the bridge was about 30% of the bridge. Not surprisingly, I found no relationship between the absence of a species and bridges having the greatest amount of obstruction. Although I did not have bridges with the full continuum of 0-100% vegetation, and therefore do not know if lack of samples precluded finding an effect, 133
common sense dictates that because Tucson bridges all had low amounts of vegetation, bats were not prevented from access to any of them.
There is suggestive, but inconclusive evidence, that the presence of pipistrelles is associated with bridges with more vegetation after accounting for adjacent land use. Further research is necessary to find the factors important to pipistrelles in bridge selection. Vegetation alongside the bridge may be a covariate for the amount of vegetation in the area of the bridge; increased vegetation may provide better foraging habitat. Vegetation may also be associated with vegetated washes that provide a natural corridor from the natural roosts in cliff crevices in the mountains to the bridges in town. Finally, because the strongest factor associated with pipistrelles seems to be the location of the bridge within a certain geographical area of town, a possible statistical relationship to the amount of vegetation along the bridge could be biologically meaningless.
I found a strong positive relationship between the amount of vegetation and the number of free-tails at the bridges occupied by this species. Free-tails forage in open areas at high altitudes and can travel far from the roost (McCracken 1996); therefore, for foraging reasons, vegetation near the bridge is not likely to be an important factor in roost selection. However, although vegetation may not in itself be important, it may affect other attributes of a roost site, such as temperature and humidity, to make it more or less desirable.
It is also possible, and perhaps more likely, that the statistical relationship between increased vegetation and increased abundance has no biological significance. It could be simply an artifact of analysis; I grouped bridges into 3 categories: 0-9%, 10-19%, and 20-134
30% vegetation. Compared to a bridge with, for example, 75% vegetation, the difference between bridges with 10% and 20% vegetation is small; because there are no samples >30%, my 3 ordinal categories statistically maximized small biological differences. Bridges along the concrete channelized wash had 0% vegetation and housed many free-tails. The stadium at the University of Arizona has a large colony of free-tails and is surrounded by parking lots and buildings. When I compared the 4 bridges on the Pantano Wash, which support large colonies, to the 4 bridges on the Rillito River, which support very small numbers, both height of the bridge and the amount of vegetation were significant when tested independently. When tested together, neither was significant. Because the Pantano bridges are both higher and have more vegetation, and height has been found to be a significant factor for Texas free-tails (Keeley and Tuttle 1996), it seems likely that height is the important variable and not vegetation.
Temperature
Bridges
The most interesting pattern I observed in temperature profiles was that, in the summer, temperatures at the tops of crevices in bridges decked with asphalt approached or equaled ambient daily highs. In bridges with a more insulated deck (e.g. gravel under railroad ties, soil and trees, covered bus stop), daily high ambient temperatures were moderated by at least a few degrees at the top of the crevices. The range of daily temperatures in the more insulated crevices was also smaller than the range in crevices decked with asphalt. From the preliminary evidence obtained in this study, it appears that the type and thickness of material on a bridge deck may influence crevice 135
temperatures to a greater degree than factors such as crevice depth, width, or the position of a crevice in the bridge. The thickness and type of material on the surface of the bridge are not factors I investigated, but further research on how these factors affect temperatures in bridge crevices is warranted. There may even be differences between bridges decked only with concrete or with both concrete and asphalt. Typically, a concrete deck about 10 cm thick is laid over the bridge beams, and then about 5 cm of asphalt is laid over the concrete, although sometimes the asphalt is omitted (J. Ebert, Tucson Dept. of Transportation, pers. comm.). When repairs are needed, an additional layer of asphalt is added, or the original layer can be scraped off first; the depth of asphalt, therefore, can vary among bridges, or over time at the same bridge. The differences in reflected heat between white concrete and black asphalt, and differences in thickness of concrete and asphalt decks are factors that may affect crevice temperatures.
Bats
Temperature regime in the roost is an important attribute of the bats’ microhabitat because of its influence on energy expenditure (McNab 1982). Warmer temperatures in maternity colonies increase the rate of growth and development of young (McNab 1982). For adults, ideal roost temperatures are those that minimize a species’ metabolic rate and therefore conserve water and energy reserves. Microchiropteran bats are heterothermic, but like all mammals, have thermoneutral zones and physiological tolerances that vary with species. As each species has a unique set of physiological and natural history requirements (Hill and Smith 1984), physiological differences related to temperature may explain some of the differences among species’ use of bridges in the Tucson area. 136
In 3 of the 5 bridges at which I monitored temperatures, daily high temperatures at the top of the crevice approached or equaled ambient highs in the summer. Normal average highs are 37.5° C (monthly average) in Tucson for the months of June and July (the hottest months); however, daily highs of 41-42° C throughout the summer are not uncommon. The average number of days per year during the 1990’s with highs ≥ 37.8° C was 71 (www.srh.noaa.gov). Bats living in bridges in Tucson must therefore be able to tolerate high temperatures for up to several hours a day throughout the summer.
Mexican free-tailed bats appear to be more tolerant of high temperatures than pallid bats or big browns and may be one reason free-tails are more numerous and widespread in Tucson bridges. In a barn loft, pallids moved to a cooler spot when ambient temperature reached 38° C; free-tails moved at 41-42° C (Licht and Leitner 1967). Big brown bats have a lower thermoneutral zone and higher evaporative water loss (ambient temperatures 30-40° C) than free-tails (Herreid and Schmidt-Nielson 1966). High ambient temperatures and low humidity seem a reasonable explanation for limited use of bridges in Tucson, especially those decked with asphalt, by big brown bats (Appendix G).
It is possible that cave myotis are also less tolerant of heat than Mexican free-tails; cave myotis are a temperate species, whereas free-tails are sub-tropical (Herreid 1963a). Cave myotis left bridges in Arizona when temperatures exceeded 33° C (Hayward 1961), however a colony inhabited an attic in Arizona where temperatures reached 37° C in July (Constantine 1958). High temperatures may also be a reason that cave myotis abundance decreased in Tucson bridges in the summer in this study. 137
Even though free-tails seem more heat-tolerant than some other species, temperatures in bridges in Tucson may become too hot for them. Although free-tails have been observed in the wild in temperatures up to 43° C, the greatest populations in caves were observed at 25-38° C (Herreid 1963b). Individual Mexican free-tails in a lab avoided temperatures >35° C when a temperature gradient was available to them (Herreid 1967). Most adults did not survive temperatures >40° C (80% humidity) for 6 hours; however, tolerance at 40° C was higher in October than in May, indicating that the bats become acclimated to high temperatures over time (Herreid 1967). My observations suggest summer temperatures in Tucson may be near the threshold of their tolerance. Temperatures in crevices are about the same as ambient, which, as discussed above, are quite high in June and July. At Park/Ajo (9042) bridge, where crevice temperatures in the center of the bridge are several degrees cooler than at either end, free-tails roosted in the ends during cooler months, but moved to the center of the bridge in summer. Further research is necessary to determine if high temperatures contributed to the decrease in abundance of free-tails in most bridges in July.
Caves and mines have very stable temperatures compared to bridges; in a cool cave, a maternity colony of bats can create a suitable microclimate by raising the temperature through radiated body heat (McNab 1982). These favorable conditions can last for weeks before the bats must move because ambient temperatures, or the bats’ requirements, change. Because bridges and buildings are exposed to solar radiation, summer temperatures in bridges are higher than in caves or mines. Also, temperatures in buildings and bridges are much less moderated, and fluctuate on a 24-hour cycle. Bats 138
move much more frequently in unstable environments than in stable ones as they attempt to minimize energy expenditure through behavioral thermoregulation (Valenciuc 1989). Thus, bats are more likely to select a bridge because of the existing temperature regime in crevices, switching roosts as temperatures change over the seasons, rather than attempting to modify an existing temperature regime through radiated body heat. In Tucson, radiated body heat from large colonies may increase temperatures in crevices (the effect of radiated heat would depend on the amount of body heat conducted to the concrete substrate), but the effect of clustering on crevice temperatures is unknown. It seems likely that bridges are selected in the spring because they offer warm temperatures, and as ambient temperatures increase in June and July, temperatures in crevices become too hot for some species, either because they have lower tolerance for heat, or because clustering increases temperatures too much. Observations of a decrease in free-tails and cave myotis in the summer, and short-term decreases in pipistrelles with temporary increases in daily high temperatures, seem to support the hypothesis that temperatures become too high; however, further research is necessary to determine to what extent Tucson’s high summer temperatures affect roost selection in bridges by bats.
Factors not measured
I did not measure distance to the nearest permanent water source, though Bat Conservation International found permanent water within ¼ mile to be an important factor for successful bat houses (Tuttle and Hensley 1997). Because there are abundant water sources in the Tucson metropolitan area, I assumed this was not a limiting factor in roost selection. In addition, it was not feasible to locate and measure all possible sources 139
of drinking water such as birdbaths and swimming pools. Furthermore, whether a particular water source is available to a bat species depends on the size of the water source and the area of open space around it.
Light, noise, and vibrations hypothetically could deter bats from roosting in a bridge. I did not measure ambient light under bridges. The narrowest bridges, which had higher light levels under the center of the bridge than wider bridges, were among those occupied by bats. Noise and vibration do not seem to deter bats from roosting in bridges in California (G. Erickson, California DOT, email on Batline listserve). Although I could not measure the ultrasonic frequency component that bats are most sensitive to, the loudest noise levels I observed under bridges were from trains and low military planes directly overhead; bats occupied bridges in both situations. The bridge with the highest vibration levels was also occupied. Traffic volume on bridges, correlated with noise and vibration levels, was not measured. Based on my observations, traffic volume levels on some bridges occupied by bats were among the highest traffic levels in my study area.
Although some species of bats may be attracted to roost sites in areas where artificial light at night attracts insect prey, some species may avoid lighted areas. The relationship between street lights and bat activity is rather complex. The degree of urbanization around street lights, season, and foraging style are a few factors that may affect bat behavior (Racey 1998). Furthermore, only incandescent, mercury, or sodium-mercury lamps attract insects, not the newer monochromatic or orange sodium lamps (Rydell and Baagøe 1996). Street lights, and other light sources, are abundant at almost all bridges, therefore I did not include their presence or absence as a factor in this study. 140
DOES COMPETITION AMONG SPECIES EXIST IN BRIDGE ROOSTS?
The bridges that ranked first in abundance (maximum number bats in a monthly survey) for each of the 4 commonly found species were different from each other; Broadway (9033) had the most Mexican free-tails, Campbell (9618) the most pipistrelles, Houghton RR (9563) the most cave myotis, and Park/Ajo (9042) the largest big brown maternity colony. Although the difference in use of bridges among species could be coincidence, it seems more likely to be the reflection of species’ different requirements. For example, bridges of appropriate crevice dimensions are widely available to pipistrelles and cave myotis throughout the study area, and crevice dimensions used by the 2 species largely overlap; yet the distribution of these 2 species across the study area is very different (Figures 12, 14). As bats probably select roosts based on a set of conditions rather than for specific attributes of a site (Sherwin et al. 2001), requirements and preferences concerning adjacent land use, human activity, distance to and location of alternate roost sites, or location of foraging areas may differ between pipistrelles and cave myotis and therefore affect their choice of bridge roost.
It is also possible that at some bridges, differences in bridge use is a result of competition among species. Although different roost requirements may result in the selection of different bridges by species whether or not roost sites are limited, competition can occur only if roosts are limiting. Unfortunately, whether roosts are limiting in the Tucson area for any or all species is unknown. However, the high occupancy (92%) of the study bridges with crevices large enough to be occupied may indicate that roosts are in high demand. 141
Competitive exclusion from a bridge of one species by another is probably only possible by Mexican free-tails because they form such large colonies. Large groups of free-tails, which inhabit many bridges year-round, could exclude solitary bats, such as pipistrelles. In my observations, pipistrelles rarely roosted in the same crevices as large groups of free-tails, and if they did, maintained a distance of > 6 m. Although I have no evidence, I believe the reason I found only a couple pipistrelles, and no Yuma or California myotis, at Tanque Verde Ranch bridge is that the free-tails took up almost all the available space. Large free-tail colonies may also limit the number of pipistrelles roosting in the bridges along Pantano Wash. In Pantano Wash bridges, free-tails roosted primarily in the center of each span and avoided crevices over slopes and ledges. Pipistrelles roosted at or near the ends of each span. In 2 bridges, 46% and 64% of pipistrelles I observed in 1999 roosted over the lower ledges and slopes, which accounted for 17% and !8% of total bridge area. In 2 bridges with only small numbers of free-tails and the greatest abundance of pipistrelles, pipistrelles roosted throughout the entire bridge and used areas over ledges and slopes in proportion to their availability.
It is possible, but unlikely, that Mexican free-tails could exclude other colonial species from bridges because other species frequently roost with free-tails. I observed colonies of big browns, pallids, and cave myotis intermingled with free-tails in bridges; whether the presence of large free-tail colonies limited the size of these colonies is unknown.
At the 3 bridges where free-tails, pipistrelles, and big browns reside, my observations suggest there may be niche partitioning, due to the large numbers of free-142
tails at each bridge. Within each bridge, height in the end spans varies greatly due to slopes and ledges at either end of the bridge. All 3 species use the highest areas, but minimum heights used differ. Crevice widths also vary greatly throughout each bridge. Many crevices fall within width ranges used by all 3 species, but some do not. Width range preferences differ among the species but overlap (Table 7). In Broadway bridge (9033), big browns occupied the widest crevice, which was never occupied by another species. In 22nd (9034) bridge, big browns used a crevice among the widest available. Free-tails also used crevices of the same width in different areas of the bridge; however, the big browns roosted above a ledge only 2.3 m high, which was too low for free-tails. In Speedway (9568) bridge, big browns again roosted in one of the widest crevices, which was only occasionally used by individual free-tails. As discussed above, pipistrelles roosted in areas too low for free-tails, and in crevices too narrow for free-tails. It appears that both pipistrelles and big browns roosted in places where crevice width and height were within their respective preferences, but in the part of the preferred range that did not overlap with preferences of the more numerous free-tails.
SPECIES DIVERSITY AND DEGREE OF URBANIZATION
The Sonoran Desert supports a rich diversity of bat species. As the Tucson basin was converted from desert to an urban area, both roosting and foraging habitat for bats changed. Roost sites, such as trees and caves, have been destroyed or abandoned due to human disturbance. Perennial streams no longer flow in the Santa Cruz and Rillito rivers, and most of the riparian canopy forests and large areas of mesquite bosques are gone. 143
Even with the addition of anthropogenic roosts, such as bridges and buildings, the diversity of bat species in Tucson appears to decline with increasing urbanization. The 3 bridges in Tucson with the highest species diversity are located at the edge of the urban area or in outlying areas, and are adjacent to natural open space. The majority of bridges in the most urban part of the study area supported only one or 2 species, usually free-tails, and big browns or pipistrelles. It is possible that free-tails and big browns are the only true urban bats, and that, with the possible exception of pipistrelles, other species are restricted to the fringe of the urban area or outlying areas. If this is true, there are implications for the future for which species can use bridges, and probably other man-made structures, in areas that are now on the fringe of development but are becoming more urbanized. Monitoring Houghton RR (9563) bridge would provide insight into this subject by observing if species diversity declines as the area around the bridge changes from natural open space to mass-graded, high density single family homes.
Decreasing species diversity in urban areas compared to non-urban areas has been documented for other taxa, such as native species of birds (Emlen 1974), and reptiles and amphibians (Schlauch 1976), as well as for bats (Brosset et al. 1996). In Avon, England, most of the 12 bat species present in the area roosted in buildings; however only one species was found in the center of cities, whereas 7 species roosted in suburban areas and parklands close to city centers (Jones and Jaynes 1988). Lowered bat diversity and abundance occurred even in vegetated areas, such as an urban river park, compared to a rural river corridor (Kurta and Teramino 1992). 144
Many bat researchers, citing Humphrey (1975), state that roosts are the ultimate limiting factor for species distribution and abundance. However, Humphrey was referring to distribution over a large scale (temperate North America); at smaller scales, other factors may limit diversity. For example, water may limit the distribution of some species in desert areas. For species that can persist in urban environments, lack of foraging areas may be more limiting than roost sites (Pierson 1998, Geggie and Fenton 1985). Loss of natural roost sites and the subsequent use of man-made structures in an area seems to support the hypothesis that roosts are limiting (Humphrey 1975, Brigham 1991). However, there must be adequate foraging habitat within the nightly commuting distance, which differs among species. Close foraging areas are especially important for lactating females who return during the night to feed young, because long distances are energetically inefficient. In urban areas, some loss of bat diversity is probably due to lowered insect abundance and diversity. Lower activity levels of foraging big brown bats in urban areas was attributed to lower insect availability (Geggie and Fenton 1985, Furlonger et al. 1987). Fewer trees in high density development sections in a Czech city may have caused lowered bat diversity and activity (Gaisler et al. 1998).
That big brown and Mexican free-tailed bats were found in bridges throughout the most urban areas of Tucson is not surprising. Free-tails can commute long distances to foraging areas. Big browns, closely associated with humans, are generalists in diet (Ross 1967) and thus able to exploit insect resources in urban areas, although they have been observed commuting up to 18 km from Denver’s urban core to foraging areas outside the 145
city (O’Shea et al. 1999). Further research is necessary to determine what limits the distribution of pipistrelles to some highly developed areas of Tucson, but not others.
MANAGEMENT IMPLICATIONS AND RECOMMENDATIONS
THE IMPORTANCE OF BRIDGES TO BATS IN TUCSON
Results of this study indicate that bridges are important roost sites for bats in Tucson. They provide year-round roosts for resident bats, transitory stop-over sites for migrating bats, and roosts for summer and winter residents. Bats occupied 92% of the bridges in my study that had crevice dimensions appropriate for use by at least one species (35 bridges).
Relative importance of individual bridges
146
Most of the bridges I monitored were used year-round, and over half the bridges were used by more than one species. The bridges at which I found only a few bats were probably used more than my monthly surveys indicate, especially during seasons of migration and dispersal of young. Use of bridges by bats in Tucson is widespread, and perhaps even greater than my results indicate.
Although I documented large (< 11,000) groups of Mexican free-tailed bats using individual bridges, this species was also found in smaller numbers at many bridges. Individuals and small groups used bridges during migration, and throughout the year, that do not appear conducive for use by large colonies. The largest number of bridges used by free-tails occurred in April during migration. Although any one particular bridge may not be crucial to the population, collectively, bridges could be an important source of temporary roost sites for bats migrating through the Tucson area.
We do not know how important one particular roost site is to a population or species. Obviously, those bridges occupied by the largest number of bats should receive close monitoring and protection. Using abundance as the only criterion, however, would be a mistake. For example, a bridge used by 50 pipistrelles or 100 cave myotis may be much more important to populations of those species than a bridge used by 1000 free-tails is to that species. Based on species diversity, and the abundance of each species using bridges in Tucson, there are several bridges that warrant the closest monitoring and protection (not listed in any order): Houghton RR (9563), Broadway (9033), Tanque Verde (9569), Speedway (9568), 22nd (9034), First Ave (9617), Campbell (9618), Park/Ajo (9042), River Rd (9557), Ina (9566), and Lee Moore (1635). I am not implying 147
that bridges excluded from the list should not be monitored or protected, only, that because agency resources are limited, listed bridges should receive priority.
The importance of bridges compared to buildings and natural roost sties in Tucson
The relative importance of bridges as roost sites compared to buildings and natural roosts sites in the Tucson area is unknown. Building roosts may be more abundant, but more ephemeral, because property owners can exclude bats and seal up roost sites. The importance of bridges as roost sites may also be species specific. Anecdotal evidence indicates that big brown bats use buildings more than bridges in Tucson whereas Mexican free-tails appear to be more abundant in bridges. The relative importance of any one roost site may change through time as alternative roost sites become more or less available. Protected natural areas adjacent to the Tucson metropolitan area, such as Saguaro National Park, Coronado National Forest, and county and state parks, provide natural roost sites. The extent to which roost sites have been lost because of human disturbance is unknown, but some loss has been documented.
LOCATING BRIDGES WITH BATS--SUGGESTED PROTOCOL
Identify bridges with the greatest potential for roost sites
The Bridge Records of the Arizona Department of Transportation list all bridges and culverts under the jurisdiction of city, county, state, and federal governments. The Record provides locations and structural information and can be used to identify which bridges are concrete. Metal and timber bridges have low potential for bat roosts compared to concrete bridges (Keeley and Tuttle1996, 1999). The Bridge Record will also differentiate bridges and culverts. Unfortunately, I could not use the Bridge Record 148
to correctly classify bridges according to the type of concrete beam used. The codes for structure type were often inconsistent. For example, both box beams (with multiple vertical crevices) and I-beams (without vertical crevices) were listed as structure type 505.
Bridges carrying railroad tracks are not listed in the Bridge Record. Railroad companies can provide the location and structure type of bridges under their jurisdiction.
Initial site inspection
Identify structure type of bridge and potential roost sites
A quick survey of the bridge will identify the type of construction and potential roost sites. Look for multiple vertical crevices between beams, horizontal crevices where beams or slabs rest on bridge piers and abutments, vertical crevices (expansion joints) between the spans of the bridge above piers, drain pipes that are open at the top or lead into a hollow girder, and recessed, open areas between I-beams or similar type beams.
Crevices and expansion joints can be open, sealed at the top only, or partially or completely sealed. Sealing may not be consistent throughout the bridge.
Survey for bats and evidence of bats
Survey all potential roost sites with a spotlight and binoculars. Use the maximum light available (≥ 1 million cp) and binoculars with a short focal distance. Take a ladder to check expansion joints. For bridges with crevices between beams, do not overlook the very end of the crevice over the pier. Look for guano under potential roost sites. For I-beam type bridges, look for guano between the beams under the cave-like recesses, 149
especially at the abutments where it is darkest. Urine crusts or stains may be visible in crevices or on walls if bats have roosted there for some time.
Follow-up surveys
Surveying with a spotlight and binoculars is necessary to detect presence or absence. If I had looked only for guano instead of up in the crevices, I probably would have missed pipistrelles entirely, even the more than 150 pipistrelles roosting in one bridge. Although huge free-tail colonies are usually easy to detect, in winter even large groups of free-tails produce practically no guano and are silent; looking in the crevices is the only way to find them.
Multiple surveys are also important. Bats come and go, often within a season. Most people think summer is the best time to survey because hibernating species are active. However, a single summer survey is not sufficient. In Tucson bridges, Mexican free-tails and cave myotis were at their lowest abundance in July. One bridge had hundreds of free-tails all year but 0 during the July survey. Some bridges are only used in winter or only during migration. Furthermore, in many cases, numbers can change dramatically within a short time period. Four surveys, one per season, would be the absolute minimum necessary to develop a picture of bat use of a bridge. Surveys, however, determine presence only, not absence; finding no bats by surveying once a season is not evidence the bridge is never occupied .
Minimizing disturbance to bats
During my surveys, I intentionally minimized disturbance by lighting bats for as brief a time as possible to identify the species and count individuals, usually a few 150
seconds at most. I obviously irritated many bats, but I believe that this brief and occasional disturbance did not adversely affect them in the long-term. I would advise against using a spotlight to observe bats frequently or for extended periods of time, unless it can be determined that for a given species, there is no negative effect.
ADDITIONAL SURVEY NEEDS
I monitored the 43 bridges (of >200) in my study area that had the greatest potential for roost sites and that were feasible to monitor frequently. I found bats in 2 additional bridges that were not monitored for logistical or study design reasons; surveys should be conducted on these 2 bridges to determine species identification and seasonal patterns of use. In addition, bats may use several bridges that provide potential roost sites (other than the crevices I studied) such as expansion joints, horizontal crevices, covered drain pipes, or hollow box girders. Although it is possible more roosts could be located, most bridges in the Tucson area are not suitable for use as day-roosts; it is therefore important to locate all existing day-roosts.
In addition to the day-roost sites described above, bats also use bridges as night roosts (Pierson and Rainey 1997, Perlmeter 1999, Adam and Hayes 2000). I have observed guano from night-roosting bats under I-beam bridges in Tucson. Bridges that are not suitable for use as day-roosts can provide appropriate night-roosting conditions. Bridges used as night roosts in Tucson, therefore, should be located.
Finally, concrete box, or pipe, culverts are also potential roost sites. A maternity colony of 35,000 cave myotis was found in a Texas culvert (Keeley and Tuttle 1996). 151
Culverts are numerous and widespread throughout Tucson and I have observed bats or guano on opportunistic surveys of several culverts.
The results of my study are a large, but partial, piece of the picture on the use of highway structures by bats in the Tucson area. Locating bridges used as night-roosts and locating day- and night-roosts in culverts is necessary to complete the picture.
MAINTENANCE, REPAIR, AND REPLACEMENT OF BRIDGES
Advance planning
Departments of transportation have long-range plans for the bridges slated for repair, modification, or