Mexican wolf recovery: three-year program review and assesment

              1

Recommended Citation: Paquet, P. C., Vucetich, J., Phillips, M. L., and L. Vucetich. 2001. Mexican wolf recovery: three year program review and

assessment. Prepared by the Conservation Breeding Specialist Group for the United States Fish and Wildlife Service. 86 pp.

MEXICAN WOLF RECOVERY: THREE-YEAR

PROGRAM REVIEW AND

ASSESSMENT1

Prepared by the Conservation Breeding Specialist Group (CBSG), Apple Valley, Minnesota; for

the United States Fish and Wildlife Service, Albuquerque, New Mexico:

PAUL C. PAQUET

UNIVERSITY OF CALGARY &

CONSERVATION SCIENCE INC.

JOHN A. VUCETICH

MICHIGAN TECHNOLOGICAL UNIVERSITY

MICHAEL K. PHILLIPS

TURNER ENDANGERED SPECIES FUND

LEAH M. VUCETICH

MICHIGAN TECHNOLOGICAL UNIVERSITY

Conservation Breeding Specialist Group

12101 Johnny Cake Ridge Road

Apple Valley, MN 55124-8151

tel: 1-952-997-9800

e-mail: office@cbsg.org

June 2001

-i-

ACKNOWLEDGMENTS

We thank Onnie Byers and Linda Phillips for their help in preparing this document. Wendy

Brown, Daniel Groebner, and Brian Kelley promptly provided the biological information

necessary for our review. We are particularly grateful to the Interagency Field Team,

Interagency Management Advisory Group, and David Parsons for advising us about the details of

the Mexican wolf reintroduction program. Lastly, we thank the San Carlos Apache Nation for

hosting a meeting with individuals and organizations involved with reintroduction of Mexican

wolves.

-ii-

TABLE OF CONTENTS

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -i-

TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -ii-

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -v-

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -vii-

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 1 of 86

2. ISSUES FOR WHICH ASSESSMENTS WERE REQUESTED . . . . . . Page 2 of 86

3. OUR APPROACH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 3 of 86

4. ECOLOGICAL BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 5 of 86

a. RESILIENCY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 5 of 86

b. THE PERILS OF SMALL POPULATIONS . . . . . . . . . . . . . . . . . . . Page 6 of 86

c. USE OF HABITAT AND PATTERNS OF TRAVEL . . . . . . . . . . . Page 6 of 86

d. INFLUENCE OF WOLVES ON THE BIOLOGICAL COMMUNITY

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 6 of 86

e. RESPONSE OF WOLVES TO HUMAN ACTIVITIES . . . . . . . . . . Page 8 of 86

f. HUMAN INFLUENCE ON HABITAT USE BY WOLVES . . . . . Page 10 of 86

g. RESPONSE OF WOLVES TO LINEAR DEVELOPMENTS . . . . Page 11 of 86

5. HAVE WOLVES SUCCESSFULLY ESTABLISHED HOME RANGES WITHIN

THE DESIGNATED WOLF RECOVERY AREA? . . . . . . . . . . . . . . . Page 14 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 14 of 86

b. DATA SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 14 of 86

c. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 14 of 86

d. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 15 of 86

e. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 23 of 86

6. HAVE REINTRODUCED WOLVES REPRODUCED SUCCESSFULLY IN THE

WILD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 24 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 24 of 86

b. DATA SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 24 of 86

c. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 24 of 86

d. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 24 of 86

e. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 27 of 86

7. IS WOLF MORTALITY SUBSTANTIALLY HIGHER THAN PROJECTED IN

-iii-

THE EIS? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 28 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 28 of 86

b. DATA SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 29 of 86

c. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 29 of 86

d. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 30 of 86

e. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 34 of 86

8. IS POPULATION GROWTH SUBSTANTIALLY LOWER THAN PROJECTED

IN THE EIS? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 35 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 35 of 86

b. DATA SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 35 of 86

c. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 35 of 86

d. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 36 of 86

e. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 45 of 86

9. ARE NUMBERS AND VULNERABILITY OF PREY ADEQUATE TO SUPPORT

WOLVES? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 46 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 46 of 86

b. DATA SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 47 of 86

c. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 47 of 86

d. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 48 of 86

e. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 52 of 86

10. HAS THE LIVESTOCK DEPREDATION CONTROL PROGRAM BEEN

EFFECTIVE? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 53 of 86

a. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 53 of 86

b. DATA SUMMARY AND METHODS . . . . . . . . . . . . . . . . Page 53 of 86

c. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 53 of 86

d. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 55 of 86

11. HAVE DOCUMENTED CASES OF THREATS TO HUMAN SAFETY

OCCURRED? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 56 of 86

a. DATA SUMMARY AND METHODS . . . . . . . . . . . . . . . . Page 56 of 86

b. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 56 of 86

c. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 60 of 86

12. OVERALL CONCLUSIONS AND RECOMMENDATIONS . . . . . . . Page 61 of 86

a. PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 61 of 86

b. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 61 of 86

c. RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 62 of 86

Biological and Technical Aspects . . . . . . . . . . . . . . . . . . . . . . . . . Page 62 of 86

Valuational and Organizational Aspects . . . . . . . . . . . . . . . . . . . Page 65 of 86

-iv-

BIBLIOGRPAHY AND LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . . Page 70 of 86

-v-

LIST OF FIGURES

Figure 1. Summary of Mexican wolf radio telemetry data, 1998-2001. . . . . . . . . Page 17 of 86

Figure 2. Monthly radio-telemetry locations of reintroduced Mexican wolves, Arizona, 1998-

2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 17 of 86

Figure 3. Many telemetry locations resulted from data entry errors. For example, numerous

locations were in the state of California and in the Gulf of California. . . . . Page 18 of 86

Figure 4. Variation among wolf packs in the proportion of telemetry locations within the primary

zone and within the recovery area (Apache/Gila N.F.). These data include all telemetry

locations of reintroduced Mexican wolves from 3 March 1998 to 3 March 2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 19 of 86

Figure 5. Temporal trends in the proportion of telemetry locations (pooled across all packs)

within the primary zone (Apache N.F.) and within the recovery area (Apache/Gila N.F.).

These data include all telemetry locations of reintroduced Mexican wolves from 3 March

1998 to 3 March 2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 19 of 86

Figure 6. Approximate area occupied by free- ranging Mexican wolf population in Arizona and

New Mexico, 1998-2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 20 of 86

Figure 7. Density of free-ranging Mexican wolf population in Arizona and New Mexico, 1998-

2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 20 of 86

Figure 8. Seasonal distribution of free-ranging Mexican wolf population in Arizona and New

Mexico, 1998-2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 21 of 86

Figure 9. Polygons reflecting the spatial extent of pack home ranges in relation to the primary

zone (Apache N.F.) And recovery area (Apache/Gila N.F.). These data include all

telemetry locations of reintroduced Mexican wolves from 03 March 1998 to 03 March

2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 22 of 86

Figure 10. Projected numbers of breeding pairs (in the EIS) and actual numbers of litters for

reintroduced Mexican wolves, 1998-2001. . . . . . . . . . . . . . . . . . . . . . . . . . Page 26 of 86

Figure 11. Actual and projected numbers of recruits for reintroduced Mexican wolves, 1998-

2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 26 of 86

Figure 12. Causes of wolf mortality for Mexican wolves reintroduced to Arizona, 1998-2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 31 of 86

Figure 13. Cause specific wolf mortality for Mexican wolves reintroduced to Arizona, 1998-

2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 31 of 86

Figure 14. Survival analysis of reintroduced Mexican wolf population assuming that recapture

represents a mortality event. Analysis was conducted for the period 1998-2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 33 of 86

Figure 15. Survival analysis of reintroduced Mexican wolf population assuming that recaptures

do not represent a mortality event. Analysis was conducted for the period 1998-2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 33 of 86

Figure 16. Projected and actual annual growth rates of free-ranging Mexican wolf population.

Actual growth rate is strongly influenced by frequent intervention. . . . . . . Page 37 of 86

Figure 17. Projected and actual sizes of free-ranging Mexican wolf population, 1998-2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 38 of 86

-vi-

Figure 18. Number of free-ranging radiocollared Mexican wolves, 1998-2001. The difference

between the max and min accounts for 4 wolves whose signals were lost, and in one case,

a wolf that threw its collar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 39 of 86

Figure 19. Growth rates of other recovering wolf populations. Sources:

<http://www.r6.fws.gov/wolf/annualrpt99/> and unpublished documents from JAV

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 40 of 86

Figure 20. Mean annual growth rate for other recovering populations. . . . . . . . . . Page 41 of 86

Figure 21. Number of wolves over time in other recovering populations. . . . . . . Page 41 of 86

Figure 22. Mean onthly growth rate (r) since March 1998. The expected value of r is 0.02. The

standard error is 0.07.* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 43 of 86

Figure 23. Mean monthly growth rate (r) since December 1998 (when population went

temporarily extinct). The expected value of r is 0.06. The standard error is 0.08.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 44 of 86

Figure 24. Because of frequent interventions the vital rates we derived (survival and population

growth) are unlikley to reflect the population’s future viability. A balance between

intervention and the effects of natural population processes is needed. . . . Page 45 of 86

Figure 25. Prey (n = 55) probably killed by reintroduced Mexican wolves, 1998-2001.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 49 of 86

Figure 26. Potential number of wolves that, in theory, could occupy target objective of 12,950

km² (5,000 mi²) within the Blue River Wolf Recovery Area. Estimates are based on prey

biomass available to wolves and are maximum numbers. The individual contribution of

ungulate prey species is shown for comparison with other studies. . . . . . . Page 50 of 86

Figure 27. The number of livestock-wolf interactions fluctuated seasonally in the primary

recovery zone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 54 of 86

-vii-

LIST OF TABLES

Table 1. Known births and recruitments of reintroduced Mexican wolves recorded from 1998-

2001. Only 1 litter was conceived in the wild. . . . . . . . . . . . . . . . . . . . . . . Page 25 of 86

Table 2. Population events recorded for reintroduced Mexican wolf population between 1998

and 2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 36 of 86

Table 3. Potential wolf numbers (ranges) for recovery areas based on predicted population

densities of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 51 of 86

Table 4. Numbers of domestic animal injuries and deaths due to wolf depredation. The data are

for confirmed, probable and unconfirmed wolf depredations. . . . . . . . . . . . Page 54 of 86

Table 5. Ownership of property where domestic animal injuries and death due to wolves took

place. The data are for confirmed, probable, and unconfirmed wolf depredations.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 54 of 86

Table 6. Summary of wolf-human interactions reported for the Mexican wolf reintroduction

program, 1998-2001. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 57 of 86

Table 7. Biological criteria for measuring project success of Mexican wolf reintroduction at

about 5 and 10 years following completion of reintroduction phase. If the evaluation falls

between failure and success, the viability of the population should be classified as

uncertain. These guidelines follow the Red List Categories (IUCN 1994:

www.iucn.org/themes/ssc/redlists/ssc-rl-c.htm) . . . . . . . . . . . . . . . . . . . . . Page 65 of 86

Mexican wolf review . . . Paquet et al. 2001 Page 1 of 86

1. INTRODUCTION

Herein we assess the progress of efforts to reestablish Mexican wolves (Canis lupus

baileyi) in the Blue Range Wolf Recovery Area (BRWRA). This review is a direct result of an

Environmental Impact Statement (EIS) concluded by the U.S. Fish and Wildlife Service (USFWS)

in 1996 (U.S. Fish and Wildlife Service 1996). The EIS and associated final rule (Parsons 1998)

call for the USFWS to reestablish Mexican wolves to the BRWRA. The recovery area

encompasses 17,752 km2 (6,854 mi2) of the Apache National Forest in southeastern Arizona and

the Gila National Forest in southwestern New Mexico.

Specifically, the U.S. Department of Interior has authorized the USFWS to reintroduce

about 15 wolves every year for 3 to 5 years in the BRWRA primary recovery zone. The primary

recovery zone comprises about 2,664 km2 (1,029 mi2) of the Apache National Forest (Groebner et

al. 1995). The remainder of the Apache National Forest and all the Gila National Forest make up

the secondary recovery zone. The USFWS may conduct re-releases in the secondary recovery

zone and wolves that move from the primary recovery zone can inhabit the secondary zone.

The USFWS began reintroductions with the release of 11 wolves in March 1998. From

then until March 2001 the USFWS released another 45 individuals on 61 occasions. An

Interagency Field Team comprising employees from the USFWS, Wildlife Services (U.S.

Department of Agriculture), Arizona Department of Game and Fish, and New Mexico

Department of Game and Fish carried out the releases and associated fieldwork

The final rule governing the reintroduction project (Parsons 1998) and the 1998 Mexican

Wolf Interagency Management Plan both require the USFWS to conduct a comprehensive review

of the project at the end of the third year (i.e., March 2001). The full evaluation must include

recommendations regarding continuation, modification, or cancellation of the reintroduction

effort. If appropriate, the evaluation may include recommendations on whether and how to use

the White Sands Wolf Recovery Area.

The primary goal of the reintroduction effort is to restore a self-sustaining population of

about 100 wild Mexican wolves distributed over 12,950 km² (5,000 mi²) of the BRWRA. Such

an objective is consistent with the 1982 Mexican Wolf Recovery Plan (U.S. Fish and Wildlife

Service 1982 (EIS). The 1998 Mexican Wolf Interagency Management Plan projects that about 9

years will be required to achieve this objective. Wolves in the BRWRA are to be managed to

reduce negative impacts and maximize positive influences on the lifestyles and economy of local

residents.

The USFWS contacted the Conservation Breeding Specialist Group (CBSG) to conduct

the specified review. CBSG is ideally suited for the task because of extensive worldwide

experience with small population restoration, conservation, and management. On behalf of

CBSG, Paul Paquet assembled an expert review team composed of John Vucetich, Michael

Philips, and Leah Vucetich. The team review is based on data provided by the USFWS data

collected in the first 3 years of the reintroduction project.

Mexican wolf review . . . Paquet et al. 2001 Page 2 of 86

2. ISSUES FOR WHICH ASSESSMENTS WERE REQUESTED

Our assessment addresses the following questions as outlined by the 1998 Mexican Wolf

Interagency Management Plan.

  Have wolves successfully established home ranges within the designated wolf recovery

area?

  Have reintroduced wolves reproduced successfully in the wild?

  Is wolf mortality substantially higher than projected in the EIS?

  Is population substantially growth lower than projected in the EIS?

  Are numbers and vulnerability of prey are adequate to support wolves?

  Is the livestock depredation control program effective?

  Have documented cases of threats to human safety occurred?

We were not asked to address the following 2 additional questions identified in the 1998

Mexican Wolf Interagency Management Plan:

  Is effective cooperation occurring with other agencies and the public?

  Are combined agency funds and staff adequate to carry out needed management,

monitoring and research?

Mexican wolf review . . . Paquet et al. 2001 Page 3 of 86

3. OUR APPROACH

Although a paucity of data compels us to speculate on many biological issues, we do so

using the best available information about wolf ecology. The lack of information reflects the short

time the Program has been underway. Where necessary and appropriate we infer from published

studies conducted elsewhere, our own experiences, and the experience of other researchers and

managers. Throughout the report, we are careful to distinguish fact from inference, speculation,

and professional opinion. Our conclusions and recommendations reflect our current knowledge

and the fundamental principles of Conservation Biology.

Conclusions and recommendations depend on the likelihood of the assumptions underlying

the assessment. Therefore, we reviewed several principles of conservation biology, which apply to

restoring and maintaining a viable population of wolves. Some of these principles are established

generalizations, some are testable hypotheses, and others are practical guides that we assessed as

important in developing our recommendations.

  The fewer data or more uncertainty involved, the more conservative conclusions must be

  To be comprehensive, an assessment must be concerned with multiple levels of biological

organizations and with many different spatial and temporal scales.

  Species well distributed across their native range are less susceptible to extinction than

species confined to small portions of their range.

  Large blocks of habitat containing large populations of a target species are superior to

small blocks containing small populations.

  Maintaining viable ecosystems is usually more efficient, economical, and effective than a

species by species approach.

  Viability of wild populations depends on the maintenance of ecological processes.

  Human disturbances that mimic or simulate natural disturbances are less likely to threaten

restoration efforts than disturbances radically different from the natural regime.

We note that how we measure and perceive the success or failure of wolf recovery is

contextual. Accordingly, our focus is on wolf ecology and how the quality of management affects

the persistence of the reintroduced Mexican wolf population. Specifically, we are concerned with

the viability of the population as affected by habitat quality, population size, population isolation,

and agency management. Although a viable wolf population could affect people’s lives and the

economy, we do not consider social and economic issues in this report.

Finally, our protocol for assessment was to:

  Review pertinent scientific literature

  Use available data provided by the Interagency Field Team

  Review pertinent regulations, polices, and rules

  Evaluate data quality

  Identify data gaps

  Analyze and interpret data

Mexican wolf review . . . Paquet et al. 2001 Page 4 of 86

  Compare progress with program goals

  Evaluate program success & failures

  Develop data collection, data management & conservation recommendations

Mexican wolf review . . . Paquet et al. 2001 Page 5 of 86

4. ECOLOGICAL BACKGROUND

a. RESILIENCY

Resilience has been defined as the ability to absorb disturbance and still maintain the same

relationship between populations or state variables (Holling 1973) and the degree to which an

entity can be changed without altering its minimal structure (Pickett et al. 1989). Thus, resilience

can be thought of as a property of a system, whereas persistence is the outcome (Weaver et al.

1996).

Wolves evolved in environments that included prevailing disturbance regimes with certain

ecological characteristics and boundary conditions. Disturbance varied in frequency, duration,

extent, and intensity, thereby resulting in different spatio-temporal patterns of change. Behaviors

and life history traits conferred resilience that enabled wolves to absorb these intrinsic

disturbances and persist. Modern humans, however, have presented new regimes of disturbance

that could be considered exotic because they are qualitatively novel or quantitatively atypical.

Three mechanisms of resilience at different hierarchical levels are: individual - plasticity in

foraging behavior that ameliorates flux in food availability; population - demographic

compensation that mitigates increased exploitation; and metapopulation dispersal - that provides

functional connectivity among fragmented populations. Accordingly, flexible food habits, high

annual productivity, and dispersal capabilities enable wolves to respond to natural and

human-induced disturbances (Weaver et al. 1996). However, environmental disturbances at

various temporal and spatial scales may exceed the ability of wolves and systems that support

them to absorb disturbance (Weaver et al. 1996).

Wolves display remarkable behavioral plasticity in using different prey and habitats (Mech

1991). They are able to substitute one resource for another in the face of environmental

disturbance (Weaver et al. 1996). Specifically, wolves specialize on vulnerable individuals of

large prey [elk (Cervus elaphus) and moose (Alces alces)] yet readily generalize to common prey

[usually deer (Odocoileus sp.)] (Weaver et al. 1996).

Wolf populations are able to compensate demographically for excessive mortality. Under

certain circumstances this compensation enables wolves to respond to increased rates of juvenile

or adult mortality with increased reproduction and/or survival, thereby mitigating demographic

fluctuations (Weaver et al. 1996). Dominant wolves are able to reproduce at a very young age

and usually reproduce every year thereafter (Weaver et al. 1996). Age at reproductive

senescence has not been well documented but few females survive to reproduce past age 9 (Mech

1988). Wolves also display remarkable ability to recover from exploitation. For example, during

a wolf reduction program in the Yukon, wolves recovered to pre-reduction densities within 5

years (Hayes and Harestad 2000). Wolves immigrated into the study area during early recovery,

followed by increases in pack size from reproduction (Hayes et al. 2000).

The final mechanism that confers resilience to wolf populations is dispersal. When

dispersal is successful, vanishing local populations are rescued from extirpation (Brown and

Kodric-Brown 1977) and functional connectivity of metapopulations is established (Hansson

1991). Wolves have tremendous dispersal capabilities and as a result, “connectivity” of

populations can be high. Dispersing wolves typically establish territories or join packs within 50-

Mexican wolf review . . . Paquet et al. 2001 Page 6 of 86

100 km of the pack in which they were born (Fritts and Mech 1981, Fuller 1989, Gese and Mech

1991, Wydeven et al. 1995, Boyd et al. 1996). Some wolves, however, move longer distances.

For example, Fritts (1983) observed a wolf that traveled at least 917 km.

b. THE PERILS OF SMALL POPULATIONS

Small populations, because of random normal variability in demographics, are more likely

to become extinct than larger populations (Schonewald-Cox et al. 1983). Moreover, these small

populations are thought to be vulnerable because of deleterious effects of inbreeding (Wright

1977) and chance environmental disturbances such as forest fires, disease or infestations that

affect a species or its prey (Franklin 1980). In theory, the interaction of these factors increases

the probability of extinction (Shafer 1987).

Small insular populations may have a restriction of genetic variation because they

represent a very small subset of the total population (i.e., a few individuals). As populations

become smaller a further reduction in genetic variation results in decreased survival (i.e.,

increased mortality). Increased mortality leads to additional reduction in genetic variation

resulting in an "extinction vortex.” Biologists theorize that because of this self-amplifying cycle

the rate of extinction for small populations is higher than predicted from the population size alone

(cf. Caro and Laurenson 1994).

c. USE OF HABITAT AND PATTERNS OF TRAVEL

Throughout its broad geographical distribution the gray wolf is considered an ecosystem

and prey generalist. However, populations are adapted to local conditions and are, therefore,

specialized concerning den site use, foraging habitats, and prey selection. In mountain regions,

the effects of physiography, weather, prey distribution, and prey abundance combine to

concentrate activities of wolves into forested valley bottoms (Paquet 1993, Paquet et al. 1996,

Paquet et al. 1996, Weaver 1994, Singleton 1995, and others).

Elevation can also govern seasonal movements of wolves. In mountainous areas with high

snowfall, use of low elevation valleys increases during winter, where frozen rivers and lakes,

shorelines, and ridges are preferred because of ease of travel. Ski trails, snowmobile trails, graded

roads, and packed roads can artificially enhance the range and efficiency of winter forays (Paquet

1993). Singleton (1995) has suggested that variation in pack size, variation in home range size,

and interactions with sympatric predators may influence habitat use and travel patterns. He

further speculated that turning frequency or travel route complexity are likely to vary depending

on whether an animal is within a patch of concentrated resource availability (e.g., deer winter

ranges), moving between known patches, or exploring new areas.

d. INFLUENCE OF WOLVES ON THE BIOLOGICAL COMMUNITY

Generally we understand that the ecology of predators, prey, and scavengers, is

intertwined. However, the details of these relationships, and the general role of predation in

shaping the structure of ecological communities is poorly understood. Changes in predator-prey

relationships may affect species other than wolves and their prey. Disruption of top predators can

affect interspecific associations by disrupting relationships within food webs. This, in turn, may

cause unanticipated ripple effects in populations of other species (Paine 1966, 1969, 1980;

Mexican wolf review . . . Paquet et al. 2001 Page 7 of 86

Terborgh and Winter 1980, Frankel and Soulé 1981, Wilcox and Murphy 1985, Wilcove et al.

1986, Valone and Brown 1995), which markedly alter the diversity and composition of a

community (Paine 1966). Multi species effects often occur when changes in a third species

mediate the effect of one species on a second species (or analogous higher-order interactions).

For example, a wolf can affect a grizzly bear (Ursus arctos) by reducing the availability of a

limiting resource (possibly an ungulate). Also a secondary carnivore such as a coyote (C. latrans)

can affect the degree to which a herbivore's lifestyle is influenced by a primary carnivore such as a

wolf. Ecologists have only begun to develop theory that attempts to explain the coexistence of

prey in terms of predator-influenced niches ("enemy-free space").

Terborgh and Winter (1980) noted that we know little about the loss of top carnivores in

terrestrial environments, and predicted a wave of extinctions following the loss of any key species.

For example, if species interact as competitors, as predator and prey, or as facilitators in

successional processes, then the presence of one species may influence the extinction probability

of another "linked" species.

Recent evidence suggests the importance of cascading trophic interactions on terrestrial

ecosystem function and processes. Research has documented differences within systems from

which large predators have been removed or are missing (Glanz 1982, Emmons 1984, Terborgh

1988, Leigh et al. 1993, Terborgh et al. 1999). Accordingly, the ecosystem impacts of wolves

may be more profound than previously expected. For example, on Isle Royale, Michigan wolf

predation on moose has been shown to influence positively biomass production in trees of boreal

forest (McLaren and Peterson 1994). Growth rates of balsam fir (Abies balsamea) were

regulated by moose (Alces alces) density, which in turn was controlled by wolf predation

(McLaren & Peterson 1994). When the wolf population declined for any reason, moose reached

high densities and suppressed fir growth. This top-down “trophic cascade” regulation is

apparently replaced by bottom-up influences only when stand-replacing disturbances such as fire

or large windstorms occur at times when moose density is already low (McLaren & Peterson

1994). This is strong evidence of top-down control of a food chain by wolves (Terborgh et al.

1999). Research elsewhere suggests elk (Cervus elaphus) populations not regulated by large

predators affect negatively the growth of aspen (Populus tremuloides) (Kay 1990, Kay and

Wagner 1994, White et al. 1992, D. Smith pers. comm.), though information remains equivocal

(L. Morgantini pers. comm.).

In addition to the obvious interactions between wolves and prey, wolves provide a regular

supply of carrion to scavengers. Less obvious community dynamics might include the

relationships between different predators, and how wolves influence these relationships. For

example, how do wolves modify the relationships between coyotes and foxes?

Interest in the role of wolves in the broader ecosystem is not new. From 1939-1944 Adolf

Murie (1944) conducted field studies in Denali Park Alaska to determine "...the ecological picture

centering about the wolf of Mount McKinley National Park". Here, he entertained questions

about the relationships between park wolves and other wolves, between wolves and their prey,

and between wolves and other predators. Few studies, however, are available to yield insights

into many of the relationships between wolves and other ecosystem components.

Mexican wolf review . . . Paquet et al. 2001 Page 8 of 86

e. RESPONSE OF WOLVES TO HUMAN ACTIVITIES

The seriousness of human disturbance is ultimately a human judgement and, as such, some

may consider any alteration of the normal activities of wolves to be undesirable. The ecological

issue is how the probability of persistence changes with habitat degradation, small population size,

and population isolation. The management issue is what probability of persistence and

environmental quality is compatible with legislation and acceptable to society. Interpretation of

the wolf-human interaction is confounded by multiple factors that influence how wolves use the

landscape and react to people (Mladenoff et al. 1995, L. Boitani pers. comm., L. Carbyn pers.

comm., E. Zimen pers. comm.). Because of the wolf’s inherent behavioural variability, it is

unlikely that all wolves react equally to human induced change. Moreover, many extraneous

factors contribute to variance in behaviour of individual wolves. Because we have developed no

reasonable expression of those differences, assessments are usually applied at the pack and

population levels.

The specific conditions in which wolves are 'disturbed' (i.e., distribution, movements,

survival, or fecundity are impaired) are believed to be highly variable. The extent and intensity of

disturbance appear to vary with environmental and social context, and the individual animal (L.

Boitani pers. comm.). Though wolves are sensitive to human predation and harassment (Thiel

1985, Jensen et al. 1986, Mech et al. 1988, Fuller 1989, Mech 1989, Purves et al. 1992, Fuller

et al. 1992, Mech 1993, Mech 1995, Thurber et al. 1994, Mladenoff et al. 1995. Paquet et al.

1996), we have limited empirical information on tolerance to indirect human disturbance. Several

studies suggest the main factor limiting wolves where they are present and tolerated by humans is

adequate prey density (Fuller et al. 1992). Although human activities have been shown to

influence the distribution (Thiel 1985, Fuller et al. 1992, Paquet 1993, Mladenoff et al. 1995) and

survival of wolves (Mech et al. 1995, Mladenoff et al. 1995, Paquet 1993. Paquet et al. 1996),

human-caused mortality is consistently cited as the major cause of displacement (Fuller et al.

1992, Mech and Goyal 1993, and others).

Studies that have quantified wolf/human interactions have shown wolves avoid humans or

are displaced via human induced mortality (Paquet et al. 1996). Avoidance is temporal (Boitani

1982) and spatial (Mladenoff et al. 1995, Paquet et al. 1996). Several studies that used road

densities as an index of human influence concluded that human activities associated with roads

affect the survival and behaviour of wolves. Interpretation, however, was confounded because

many human activities associated with roads result in the death of wolves. Thus, absence of

wolves in an area may not be the result of behavioural avoidance per se. Data from Ontario,

Wisconsin, Michigan, and Minnesota suggest that wolf survival is usually assured at road densities

below 0.58 and 0.70 km/km² (Thiel 1985, Jensen et al. 1986, Mech et al. 1988, Fuller 1989,

Mech 1989, Fuller et al. 1992). A study in Alaska concluded that wolves avoid heavily used

roads and areas inhabited by humans, despite low human caused wolf mortality (Thurber et al.

1994). Landscape level analysis in Wisconsin found mean road density was much lower in pack

territories (0.23 km/km² in 80% use area) than in random non pack areas (0.74) or the region

overall (0.71). Few areas of use exceeded a road density of >0.45 km/km² (Mladenoff et al.

1995).

Recent reports suggest wolves in Minnesota tolerate higher levels of disturbance than

previously thought possible. Wolves, for example, are now occupying ranges formerly assumed

Mexican wolf review . . . Paquet et al. 2001 Page 9 of 86

2Wolves from the Midwestern United States have hybridized with coyotes (Canis latrans)

(Wayne et al. 1991, Wayne et al. 1992, Lehman et al. 1991), or red wolf hybrids (Wilson et al.

2001), which may affect their behaviour (Fox 1971) and their relationship with humans.

Consequently, extrapolating information from Minnesota, Michigan, Minnesota, and Ontario may

be inappropriate for the Rocky Mountains. Wolves in the Rocky Mountains show no

introgression of coyote genes (Forbes and Boyd 1996).

to be marginal because of prohibitive road densities and high human populations (Mech 1993,

Mech 1995). Legal protection and changing human attitudes are cited as the critical factor in the

wolf’s ability to use areas that have not been wolf-habitat for decades. If wolves are not killed,

they seem able to occupy areas of greater human activity than previously assumed (Mech 1993,

Fuller et al. 1992). Based on these observations, Mech (1995, p. 275) comments that

misconceptions about the wolf’s inherent ability to tolerate human activity encourage unwarranted

protection.

Nonetheless, wolves in Minnesota continue to avoid populated areas, occurring most

often where road density and human population are low (Fuller et al. 1992).2 Moreover, the fact

that wolves are using areas of greater human activity suggests dispersers or marginalised

individuals are being pushed into suboptimal habitat. More suitable and safe habitat may be

saturated by dominant animals or packs. This supports the idea that wolves occupy habitat closer

to humans only if necessary. A similar phenomenon has been shown in grizzly bears (D. Mattson

et al. 1987, Mattson pers. Comm.) and many avian species.

We are aware of only 4 studies that have systematically and explicitly examined human

population density and wolf distribution. In all studies, the absence of wolves in human

dominated areas may have reflected high levels of human caused mortality, displacement resulting

from behavioural avoidance, or some combination of both. All were conducted at a landscape

scale and assessed population or pack level responses of wolves to humans. In Wisconsin, human

population density was much lower in pack territories than in non pack areas. Wolf pack

territories also had more public land, forested areas with at least some evergreens, and lower

proportions of agricultural land. Notably, no difference was detected between white-tailed deer

(Odocoileus virginianus) densities in pack territories and non pack areas. Overall, wolves

selected those areas that were most remote from human influence (Mladenoff et al. 1995) using

areas with fewer than 1.54 humans/km² and less than 0.15 km roads/km². Most wolves in

Minnesota (88%) were in townships with <0.70 km roads/km² and <4 humans/km² or with <0.50

km² and <8 humans/km². High human or road densities likely precluded the presence of wolf

packs in several localities within contiguous, occupied wolf range (Fuller et al. 1992). In Italy,

wolf absence was related to human density, road density, urban areas, cultivated areas, and cattle

and pig density. However, because human density, road density, and urbanized areas were highly

inter correlated no specific human effect was established (Duprè et al. in press).

In the Bow River Valley, Alberta the selection or avoidance of particular habitat types was

related to human use levels and habitat potential (Paquet et al. 1996). Wolves used disturbed

habitats less than expected, which suggests the presence of humans altered their behaviour. Very

low intensity disturbance (<100 people/month) did not have a significant influence on wolves, nor

Mexican wolf review . . . Paquet et al. 2001 Page 10 of 86

did it seriously affect the ecological relationships between wolves and their prey. At low to

intermediate levels of human activity (100-1,000 people/month) wolves were dislocated from

suboptimal habitats. Higher levels of activity resulted in partial displacement but not complete

abandonment of preferred habitats. As disturbance increased, wolves avoided using some most

favourable habitats. In portions of the Valley where high elk abundance was associated with high

road and/or human population density, wolves were completely absent. Overall, habitat alienation

resulted in altered predator/prey relationships.

The observed patterns of displacement suggest the presence of humans repulses wolves,

although a strong attraction to highly preferred habitats increases a wolf’s tolerance for

disturbance. As conditions become less favorable, the quality of habitat likely takes on greater

importance. Tolerance thresholds are unknown but, as noted, in the Bow River Valley changes in

patterns of habitat use were evident when human activity exceeded 100 people/month. Nearly

complete alienation of wolves occurred when more than 10,000 people/month used an area.

f. HUMAN INFLUENCE ON HABITAT USE BY WOLVES

The degree of human influence probably varies according to the environmental context. If

a particular habitat is highly attractive, wolves appear willing to risk exposure to humans, at least

within some limits (Chapman 1977). As levels of disturbance increase, favorableness of habitat

likely takes on greater importance. For example, we know that wolves select home sites near

intense human activity when denning areas are limited, or where innocuous human activity occurs

(Chapman 1977). The presence of artificial food sources (e.g., carrion pits, garbage dumps) also

attracts wolves and reduces avoidance of human activity (Chapman 1977, L.D. Mech pers.

comm., Paquet 1996, Krizan 1998). In the Bow River Valley, wolves denned within 500 m of the

Trans Canada highway when Parks Canada was dumping carrion in the area. Wolves abandoned

the home site after Parks stopped dumping of the carrion.

The tension between attraction and repulsion is probably expressed differently by

individuals, packs, and populations. Attraction to an area is a complex sum of physiography,

security from harassment, positive reinforcement (e.g., easily obtained food), population density,

and available choice. Moreover, the response to a particular disturbance seems to depend on

disturbance-history (E. Zimen pers. comm.); a critical concept in understanding the behaviour of

long-lived animals that learn through social transmission (Curatolo and Murphy 1986, S. Minta

pers. comm.).

We can group human influence into effects on wolf habitat and populations. Habitat

disturbance can be short or long term and can include direct loss of habitat (i.e., vegetation

removal, vegetation change, or isolation and removal of prey). Direct habitat loss does not

include the loss of habitat due to temporal or spatial alienation (sensory disturbance) or from

fragmentation of habitat. Indirect losses will occur due to habitat alienation, where wolves

abandon habitat because of nearby disturbances or are spatially isolated from using them because

of impediments to movements. Changes in population can occur directly through alterations in

habitat and indirectly because of disturbing activities.

The major impacts of human induced changes are, in order of decreasing importance,

physical loss of habitat, loss of prey species, fragmentation of habitat, isolation of habitat,

alienation of habitat, alteration of habitat, changes in original ratios of habitat, and changes in

Mexican wolf review . . . Paquet et al. 2001 Page 11 of 86

juxtaposition of habitats. These effects combine to have local and population level influences by

altering the composition of biological communities upon which wolves are dependent, restricting

movements, reducing foraging opportunities, and limiting access to prey. Obstructing movements

also increases the vulnerability of wolves to other disturbances as they attempt to learn new travel

routes.

The degree to which human activities disrupt wildlife reflects the type and extent of

disturbance, which interacts with the natural environment to affect environmental quality. In

mountainous landscapes wildlife often responds markedly to disturbances that occur at small

spatial scales. This is because the topography amplifies the effects of disturbances by

concentrating activities of humans and wildlife into valley bottoms. The forced convergence of

activities limits spatially the range of options wildlife have for coping with disruption, reducing

resilience to anthropogenic disturbance (Weaver et al. 1996, Alaska Department of Fish and

Game unpublished data).

Indirect human influences can affect an animal’s chance to survive and reproduce. As

wolves approach their limits of tolerance, they become increasingly susceptible to what would

otherwise be minor influences. In the mountainous terrain, natural landforms and the condensed

arrangement of habitats make wolves highly susceptible to the adverse effects of human

disturbance. Because most development occurs in areas preferred by wolves, human activities

unavoidably increase the risk of death and injury for wolves, decrease opportunities for wolves to

move freely about, displace or alienates wolves from preferred ranges, and interrupt normal

periods of activity. In less physiographically complex environments multiple travel routes link

blocks of wolf habitat. Destruction or degradation of one or 2 routes is not usually critical,

because safe alternative routes are available. In contrast, wolves living in mountains cannot avoid

valley bottoms or use other travel routes without affecting their fitness. Therefore, tolerance of

disturbance is probably lower than in other human dominated environments where wolves can

avoid disturbed sites without seriously jeopardizing survival.

g. RESPONSE OF WOLVES TO LINEAR DEVELOPMENTS

The security of wolf populations in the many regions may be tenuous, because linear

developments heavily dissect wolf ranges (i.e., highways, secondary roads, railways, and power

line corridors). Highway mortality has become a primary cause of wolf mortality and there is

accumulating evidence of habitat loss, fragmentation, and degradation related to roads (Purves et

al. 1992, Paquet 1993). Ensured connectivity of quality habitats is important for survival of large

carnivores (Beier 1993, Paquet and Hackman 1995, Doak 1995, Noss et al. in press), especially

for those that face a high risk of mortality from humans or vehicles when travelling across settled

landscapes (Noss 1992, Beier 1993).

There are several plausible explanations for the absence of wolves in densely roaded areas.

Wolves may behaviourally avoid densely roaded areas depending on the type of use the road

receives (Thurber et al. 1994). In other instances, their absence may be a direct result of

mortality associated with roads (Van Ballenberhe et al. 1975, Mech 1977, Berg and Kuehn 1982).

Besides fragmenting and consuming critical habitat, linear developments provide access to remote

regions, which allows humans to deliberately, accidentally, or incidentally kill wolves (Van

Ballenberghe et al. 1975, Mech 1977, Berg and Kuehn 1982). Despite legal protection, 80% of

Mexican wolf review . . . Paquet et al. 2001 Page 12 of 86

known wolf mortality in a Minnesota study was human-caused (30% shot, 12% snared, 11% hit

by vehicles, 6% killed by government trappers, and 21% killed by humans in some undetermined

manner) (Fuller 1989). Mech (1989) reported 60% of human-caused mortality in a roaded area

(even after full protection), whereas human caused mortality was absent in an adjoining region

without roads. On the east side of the Central Rockies between 1986 and 1993, human caused

mortality was 95% of known wolf death. Thirty-six percent (36%) of mortality was related to

roads (Paquet 1993).

Wolves also experience higher mortality in areas with higher road density. On Prince of

Wales Island, Alaska, researchers report a significant jump in wolf mortality (kill/259 km²) in

areas where road densities exceeds.25 km/km². While wolf mortality in the category of most

densely roaded areas is highest, the variance is also high. The authors suggest that at some

threshold of road density or human activity, wolves may abandon an area, resulting in decreased

trapping and hunting mortality (Alaska Department of Fish and Game, unpublished data).

Linear developments may also be physical and/or psychological impediments to wolf

movement. Road density and human density have been inversely correlated with viable

populations of wolves in several areas. Along the Ontario-Michigan border, distribution of

breeding packs occurred only in Ontario. Except for Cockburn Island, only lone wolves were

found in areas close to the border or in Michigan. In Ontario, the density of roads in areas not

occupied by wolves was greater than in areas occupied by wolves. Mean road density in

Michigan, where no wolves resided, was also greater than in wolf-occupied areas of Ontario.

High human densities, represented by road densities of > 0.6 km/km², were believed to be a

barrier to wolf dispersal into Michigan (Jensen et al. 1986).

Studies in Wisconsin, Michigan, Ontario, and Minnesota have shown a strong relationship

between road density and the absence of wolves (Thiel 1985, Jensen et al. 1986, Mech et al.

1988, Fuller 1989). Wolves generally are not present where the density of roads exceeds 0.58

km/km² (Thiel 1985 and Jensen et al. 1986, cf. Fuller 1989). Landscape level analysis in

Wisconsin, Minnesota, and Michigan found mean road density was much lower in pack territories

(0.23 km/km² in 80% use area) than in random nonpack areas (0.74) or the region overall (0.71).

Road density was the strongest predictor of wolf habitat favorability out of 5 habitat

characteristics and 6 indices of landscape complexity (Mladenoff et al. 1995). Few areas of use

exceeded a road density of >0.45 km/km² (Mladenoff et al. 1995). Notably, radio collared packs

were not bisected by any major federal or state highway. In Minnesota, densities of roads for the

primary range, peripheral range, and disjunct range of wolves were all below a threshold of 0.58

km/km². These results, however, probably do not apply to areas on which public access is

restricted. Mech (1989), for example, reported wolves using an area with a road density of 0.76

km/km², but it was next to a large, roadless area. He speculated that excessive mortality

experienced by wolves in the roaded area was compensated for by individuals that dispersed from

the adjacent roadless area. Wolves on Prince of Wales Island, Alaska currently use areas with

road densities greater than 0.58 km/km². Core areas, however, are generally in the least densely

roaded areas of the home range, and wolf activity that does occur in densely roaded areas occurs

primarily at night. This behavioral response may reflect the limited options wolves have to

relocate when they live on islands or insularized landscapes.

Mexican wolf review . . . Paquet et al. 2001 Page 13 of 86

The response of wolves to different road types and human presence at the boundaries of

Kenai National Wildlife Refuge, Alaska, was examined in a study of radio-collared wolves

(Thurber et al. 1994). Wolves avoided oilfield access roads open to public use, yet were attracted

to a gated pipeline access road and secondary gravel roads with limited human use. Thurber et al.

speculated that roads with low human activity provide easy travel corridors for wolves. The

response of wolves to a major public highway was equivocal. They thought wolf absence from

settled areas and some roads were caused by behavioral avoidance rather than direct attrition

resulting from killing of animals. In Montana, Singleton (1995) found that wolves preferred areas

0.5-1 km from open roads for travel routes. He speculated that wolves did not select locations

more distant from open roads because of the distribution patterns of wintering ungulates and the

barrier provided by the river. Overall, wolves preferred areas with 0.01-2 mi/mi² for travel routes.

Mexican wolf review . . . Paquet et al. 2001 Page 14 of 86

5. HAVE WOLVES SUCCESSFULLY ESTABLISHED HOME RANGES

WITHIN THE DESIGNATED WOLF RECOVERY AREA?

a. BACKGROUND

Biologists usually define the home range of a wolf as an area within which it can meet all

of its annual biological requirements. Seasonal feeding habitat, thermal and security needs, travel,

denning, the bearing and raising of young, are all essential life requirements. The manner in

which habitats for these requirements are used and distributed influences home range size and

local and regional population distributions. Generally, wolves locate their home ranges in areas

where adequate prey are available and human disturbance minimal (Mladenoff et al. 1995, 1997,

Mladenoff and Sickley 1998). Wolves use areas within those home ranges in ways that maximize

encounters with prey (Huggard 1993a, b).

Newly colonizing wolf pack might shift home ranges in response to climate, food

availability, human disturbance, and other factors. A colonizing pack might have a larger, more

fluid, home range than a pack surrounded by other wolf packs (Boyd et al. 1996). Some evidence

suggests that wolf packs colonize areas that were first “pioneered”by dispersing lone wolves

(Ream et al. 1991)

In mountainous areas, topographic position influences selection of home ranges and travel

routes (Paquet et al. 1996). Wolf use of valley bottoms and lower slopes correspond to the

presence of wintering ungulate prey and snow depth in these areas (Singer 1979, Jenkins and

Wright 1988, Paquet et al. 1996). In areas of higher prey density pack sizes increase (Messier

1985) and home range size is closely correlated with pack size (Messier 1985, Peterson et al.

1984).

b. DATA SUMMARY

We assessed home ranges using locations from radio-collared animals. Radio-telemetry

data (>7000 locations) were provided in an Excel database (Monitor). These data include all

telemetry locations from 3 March 1998 to 3 March 2001. Each location was appended by wolf

identification, date, time, and pack membership. Although locations were qualitatively ranked for

accuracy, no quantitative assessment of telemetry error was available. Thus, we classified

locations into 4 categories, which corresponded to the database provided. Class 1, 2, 3, and 4

locations were those within 100 m, 100-250 m, 250-450 m, and greater than 450 m from the true

location, respectively. Only class 1 aerial and ground locations were used in the home range

analysis.

c. METHODS

Our objective was to quantitatively describe areal distribution of reintroduced Mexican

wolves within the recovery region. In a few cases, however, subjective determination of the home

range was more appropriate.

Using ArcView Spatial Analyst, we plotted all class 1 locations. We discarded locations

deemed to be recording errors, extraterritorial forays, and dispersals. We assumed a wolf

Mexican wolf review . . . Paquet et al. 2001 Page 15 of 86

dispersed if it permanently left its original pack and formed a new pack or joined an existing one

(Messier 1985b).

Locations of individual wolves were grouped by pack affiliation. We defined a pack as 2

or more wolves that traveled together more than 1 month (Messier 1984). For each pack we used

one wolf/year to represent the annual home range of the pack. This is a reasonable assumption if

a high degree of association exists between pack members (Kolenosky and Johnston 1967, Fuller

and Keith 1980, Fritts and Mech 1981, Ciucci et al. 1997). We confirmed pack affiliations by

examining telemetry locations of wolves believed to be associating and through visual

observations of the wolves by the field crew.

We used Home Range and Ranges V® software (Kenward and Hodder 1996) to calculate

annual (1 Apr–31 Mar) and seasonal 95% minimum convex polygons (Mohr 1947) for individual

packs and the entire free ranging wolf population within the primary zone and recovery area

(Apache/Gila N.F.). Home range is an extension of ArcView Spatial Analyst. We assumed home

ranges were defined when the observation-area curve formed an asymptote (Kenward and Hodder

1996) and locations were obtained throughout the year.

Accuracy of aerial and ground locations for the entire study was estimated to be 250 m,

which is the highest mean error of telemetry obtained by researchers on other wolf projects. To

account for the 250-m error, we changed the fix resolution from the RangesV® software default of

1 m to 250 m. This resolution is used to set the width of the boundary strip that is included in

polygon edges and areas (Kenward and Hodder 1996, R. Kenward, pers. comm.). We left the

scaling parameter at the software default of 1 m, which means that each coordinate unit was 1 m

from the next.

d. RESULTS

From 1998 through 2001, 9 wolf packs were identified by name in the telemetry database.

However, the criteria for specifying packs were not always biological. Release sites, geographic

locations, and affiliations with other wolves influenced pack designation. Packs, pack

compositions, and configurations of home ranges changed as reintroduced wolves encountered

other wolves, and established new territories. In addition, the frequent removal and

reintroduction of wolves confounded the assignment of individual wolves to specific packs.

The number of recorded aerial and ground locations varied among wolf packs (Figure 1).

For the most part, the frequency of locations reflected the time that radio collared wolves were

free-ranging, rather than differential effort by the field crew. Time of year, however, affected the

number of locations acquired (Figure 2). Discussions with the field team confirmed that for

logistic reasons they reduced monitoring activities in winter. We identified some locations that

were far outside the reintroduction area. Many of these were recording or data entry errors

(Figure 3). Several, however, were from wandering or dispersing wolves.

The proportion of telemetry locations within the primary recovery zone (Apache N.F.) and

within the Blue Range wolf recovery area (Apache/Gila N.F.) varied among packs (Figure 4).

Temporal trends in the proportion of telemetry locations (pooled across all packs) within the

primary zone and within the recovery area also varied (Figure 5). The approximate area occupied

by free-ranging Mexican wolf population changed over time as did the density of wolves. This

was partially a reflection of periodic releases and recaptures of wolves, and also free-ranging

Mexican wolf review . . . Paquet et al. 2001 Page 16 of 86

wolves shifting centers of activity as they established pack affiliations and home ranges (Figures 6,

7, 8).

Many individuals and packs showed home range fidelity typical of wolves with established

territories (Figure 9). However, frequent social disruption via mortality, recaptures, and re-releases

may have altered the natural territorial behavior of packs. Wolves are long-lived social

carnivores that transmit information between generations and among individual pack members. In

this regard, the establishment, location, and maintenance of home ranges likely depend on a stable

pack structure and the persistence of traditional pack knowledge. The home range behavior of

reintroduced wolves may be highly susceptible to social disruption because they lack a cognitive

map of the area. Moreover, lack of familiarity with the landscape may have a stronger influence

on captive reared animals than wild born.

Mexican wolf review . . . Paquet et al. 2001 Page 17 of 86

Cienega (258)

Campbell Blue (310)

Mule/Pipestem (7)

Pipestem (327)

Saddle (59)

Turkey Creek (15)

Francisco (121)

Gavilan (178)

Mule (271)

Loner (41)

Hawk's Nest (447)

Aerial

Cienega (771)

Campbell Blue (1071)

Mule (758)

Pipestem (604)

Saddle (177)

Turkey Creek (29)

Francisco (539)

Gavilan (311)

Loner (127)

Hawk's Nest (1837)

Ground

Figure 1. Summary of Mexican wolf radio telemetry data, 1998-2001.

Numbers in parentheses are telemetry locations recorded.

200

300

400

500

600

700

800

900

1000

Locations

0 2 4 6 8 10 12

Month

Figure 2. Monthly radio-telemetry locations of reintroduced Mexican

wolves, Arizona, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 18 of 86

Figure 3. Many telemetry locations resulted from data entry errors. For example, numerous

locations were in the state of California and in the Gulf of California.

Mexican wolf review . . . Paquet et al. 2001 Page 19 of 86

0.00

0.20

0.40

0.60

0.80

1.00

PACK

97%

78%

Primary zone

Recovery Area

Figure 4. Variation among wolf packs in the proportion of telemetry

locations within the primary zone and within the recovery area (Apache/Gila

N.F.). These data include all telemetry locations of reintroduced Mexican

wolves from 3 March 1998 to 3 March 2001.

0

0.2

0.4

0.6

0.8

1

Season-Year

Recovery Area

Primary Zone

Figure 5. Temporal trends in the proportion of telemetry locations (pooled

across all packs) within the primary zone (Apache N.F.) and within the

recovery area (Apache/Gila N.F.). These data include all telemetry locations

of reintroduced Mexican wolves from 3 March 1998 to 3 March 2001.

Mexican wolf review . . . Paquet et al. 2001 Page 20 of 86

0

5000

10000

15000

20000

Season - Year

0

2000

4000

6000

8000

Figure 6. Approximate area occupied by free- ranging Mexican wolf population in

Arizona and New Mexico, 1998-2001.

0

3

6

9

12

15

Summer

98

Winter

98-99

Summer

99

Winter

99-00

Summer

00

Winter

00-01

Season-Year

Figure 7. Density of free-ranging Mexican wolf population in Arizona and New

Mexico, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 21 of 86

Figure 8. Seasonal distribution of free-ranging Mexican wolf population in Arizona and

New Mexico, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 22 of 86

Figure 9. Polygons reflecting the spatial extent of pack home ranges in relation to the primary

zone (Apache N.F.) And recovery area (Apache/Gila N.F.). These data include all telemetry

locations of reintroduced Mexican wolves from 03 March 1998 to 03 March 2001.

Mexican wolf review . . . Paquet et al. 2001 Page 23 of 86

e. CONCLUSIONS

We conclude that some wolves have successfully established home ranges and possibly

pack territories within the designated wolf recovery area. We caution, however, that frequent

recaptures and re-releases confounded our analysis. These manipulations may also be interfering

with pack formation and establishment and maintenance of home ranges. Lastly, individual

wolves have shown some indication of dispersing outside the recovery area. This is to be

expected and required if the regional population is to be viable.

Mexican wolf review . . . Paquet et al. 2001 Page 24 of 86

6. HAVE REINTRODUCED WOLVES REPRODUCED

SUCCESSFULLY IN THE WILD?

a. BACKGROUND

i. Births versus recruitment

(1) Compared with adults, pups have relatively low survival rates

during the first year of life.

(2) In a sense, pups do not really contribute to the viability of a

population until they have survived a period of high mortality rate

associated with being a pup.

(3) Although the EIS refers to projected numbers of pups, the

projections seem to treat pups as though they have been recruited

into the adult population (i.e., with survival rates like adults).

b. DATA SUMMARY

We used information recorded in the telemetry and events databases. Additional

information on reproduction was garnered from discussions with the Field Team. Dense

vegetation and the secretive nature of wolves precluded regular and accurate visuals of wolves.

Consequently, the Interagency Field Team did not routinely observe wolves during spring and

summer when pups are easiest to distinguish from adults. We assumed the presence of dens and

rendezvous sites when movements became localized in April through July or when lactating

females or pups were captured. Sometimes, ground checks confirmed potential denning and

rendezvous areas.

c. METHODS

We determined natality directly from field observations of dens, rendezvous sites (pup

rearing and resting areas), and packs. We ascertained successful year-specific reproduction using

changes in pack size from March to the following December. We assumed unsuccessful

reproduction (i.e., no or failed reproduction) when a pack did not display focal activities in the

summer. Annual recruitment was derived from winter pack sizes recorded in February.

d. RESULTS

Births have taken place in the wild (Table 1). Births and recruitment rates, however are

lower than projected in the EIS (Figures 10 and 11).

Mexican wolf review . . . Paquet et al. 2001 Page 25 of 86

Table 1. Known births and recruitments of reintroduced Mexican wolves recorded from 1998-

2001. Only 1 litter was conceived in the wild.

PARENTS

Female

Male

ESTIMATED

DATE OF

BIRTH

(M/D/Y)

CONCEIVED

IN WILD? WILD BIRTHS

174 166 35915 No Litter of 5 pups (known number due

to necropsy report showing 5

placental scars); one survived to ~ 4

months., then disappeared after 174

was killed.

191 208 5/1/99 No Litter of unknown number (6

confirmed).

482 166 5/1/99 No Litter of 6 pups (known number due

to necropsy report showing 6

placental scars); pups were never

documented for this pair by the field

team --pair never settled in an area so

likely pups were lost immediately.

486 131 5/1/00 Yes Litter of unknown number (one

confirmed).

191 208 5/1/00 No Litter of unknown number (one

confirmed).

189 190 4/15/00 No Litter of unknown number

Mexican wolf review . . . Paquet et al. 2001 Page 26 of 86

1 2 3

0

3

6

9

12

15

18

1998 2000 2002 2004 2006

Projected # of breeding pairs

Actual numbers of litters

4

6

2

1

Year

Figure 10. Projected numbers of breeding pairs (in the EIS) and actual

numbers of litters for reintroduced Mexican wolves, 1998-2001.

0

4

8

12

16

20

1998 2000 2002 2004 2006

Year

Actual number of

recruits

Projected number of

recruits

0?

3?

1?

10

5

20

Figure 11. Actual and projected numbers of recruits for reintroduced

Mexican wolves, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 27 of 86

e. CONCLUSIONS

The number of free-ranging Mexican wolves at the end of third year is similar to that

projected in the EIS. Survival and recruitment rates, however are far too low to ensure

population growth or persistence. Without dramatic improvement in theses vital rates, the wolf

population will fall short of predictions for upcoming years.

Mexican wolf review . . . Paquet et al. 2001 Page 28 of 86

7. IS WOLF MORTALITY SUBSTANTIALLY HIGHER THAN

PROJECTED IN THE EIS?

a. BACKGROUND

Researchers do not agree on the annual rate of mortality that causes a population decline

in wolves. However, Keith (1983) and Fuller (1989) reviewed several wolf studies across North

America and concluded that harvests exceeding 28-30% of fall populations resulted in declines.

Fuller (1989) further concluded that populations would stabilize with an overall annual mortality

rate of 35%. He felt, however, the effects of harvest could vary with time and population

structure. Specifically, a population containing many pups could withstand much higher

mortality.

Various researchers have suggested different rates of annual mortality they believe control

growth of wolf populations. However, the annual rate of mortality that causes a population

decline in wolves is unknown. Furthermore, many researchers consider only harvest (hunting or

trapping) when they calculate mortality rates that cause wolf population declines. For instance

Mech (1970) concluded an annual harvest of 50% or more was necessary to control wolf

populations based on pup-adult ratios but did not distinguish between harvest and natural

mortality. Keith (1983) reviewed studies of 13 exploited populations and determined that

harvests exceeding 30% of fall populations resulted in population declines. Similarly Fuller

(1989) found annual rates of wolf increase vary in direct response to rates of mortality and where

humans kill wolves, harvests exceeding 28% of autumn or early winter populations might result in

a population decline. He concluded a population would stabilize with an overall rate of annual

mortality of 0.35 or rate of human-caused mortality of 0.28. Consequently, the exact relationship

between the annual rate of mortality from all human causes (harvest, collisions with cars and

trains) and population limitation or decline in wolves is uncertain.

In areas where ungulate biomass is low, researchers have noted that starvation and

intraspecific aggression are common. For instance, in southwestern Quebec, Messier (1985a)

noted wolves with fewer prey available incurred more deaths from natural causes, namely

starvation and intraspecific aggression. Similarly, Mech (1977a) noted occurrence of starvation

and intraspecific aggression increased as prey availability declined in Minnesota. Disease cannot

be linked with certainty to low availability of food but the relationship makes sense intuitively. A

population of wolves lacking food should be more vulnerable to disease than one with more food

available. Furthermore, food shortage leading to nutritional stress could combine with disease

factors to increase the significance of otherwise innocuous or sub-lethal conditions (Brand et al.

1995).

In most studies, no disease-related mortality has been reported (VanBallenberghe et al.

1975, Mech 1977a, Fritts and Mech 1981, Messier 1985a, Potvin 1987, Ballard et al. 1989,

Hayes et al. 1991, Meier et al. 1995, Pletscher et al. 1997). In other studies, from 2-21% of wolf

mortality has been attributed to disease (Carbyn 1982, Peterson et al. 1984, Fuller 1989, Ballard

et al. 1997). Ballard et al. (1997) concluded that occurrence of rabies was a significant factor in a

decline of wolves from Alaska. In that study, rabies-caused mortality was 21%.

Mexican wolf review . . . Paquet et al. 2001 Page 29 of 86

Quantifying the importance of food in limiting population growth based on cause of death

alone is difficult. In the literature, results vary among studies. On Isle Royale, annual mortality

from starvation and intraspecific strife (both related to low food availability) ranged from 18-57%

during a 20-year period (Peterson and Page 1988). In populations where some human-caused

mortality occurs, and thus compensates for natural mortality (starvation, accidents, disease and

intraspecific strife), about 8% of individuals greater than 6 months-of-age can be lost each year

(Ballard et al. 1987, Fuller 1989). Some researchers have accepted this variability and decided

any sign of starvation among adult wolves means food is limiting population growth (Fritts and

Mech 1981, Ballard et al. 1997, P. Paquet, pers. comm.). This assumption is reasonable given

adults typically are the last members of the population affected by food shortage (Eberhardt 1977)

and as such, may be the most sensitive indicators of a shortage of food.

Human-caused mortality can also be an important limiting factor (Peterson et al. 1984;

Ballard et al. 1989, 1997). However, quantifying the importance of human-caused mortality as a

limiting factor is difficult. These causes include legal harvest (Fuller and Keith 1980, Keith 1983,

Gasaway et al. 1983, Messier 1985a, Ballard et al. 1987, 1997, Peterson et al. 1984, Potvin

1987, Bjorge and Gunson 1989, Fuller 1989, Hayes et al. 1991, Pletscher et al. 1997), illegal

harvest (Fritts and Mech 1981, Fuller 1989, Pletscher et al. 1997), vehicles on highways (Berg

and Kuehn 1982, Potvin 1987, Fuller 1989, Paquet 1993, Parks Canada 1994, Forbes and

Theberge 1995, Paquet and Hackman 1995, Thiel and Valen 1995, Bangs and Fritts 1996), and

trains (Paquet 1993, Parks Canada 1994, Paquet and Hackman 1995, Paquet et al. 1996).

b. DATA SUMMARY

We used information recorded in the telemetry and events databases. Additional

information, clarification of events, and interpretation of events was provided by the Interagency

Field Team. All free-ranging Mexican wolves were radio-collared from time of release.

Moreover, each radio-collared Mexican wolf was and continues to be relocated regularly and

frequently via ground and aerial telemetry. Frequent monitoring reveals whether each wolf is

alive or dead at the time of relocation

c. METHODS

We were not able to address the question of annual mortality directly because removals

and re-releases precluded calculating annual rates of mortality. Thus, we estimated survival rates

for the Mexican wolf population and then compared these estimated values with the survival rates

projected in the EIS. Survival rate is the chance (or probability) of surviving some specified time.

Survival rates are typically expressed as values between zero and one. For example, if the annual

survival rate of an individual is 0.82, we would say that individual has an 82% chance of surviving

during the next year. Survival is a critical population process and estimating survival rates is an

important part of measuring viability of populations. Management of protected wolf populations

requires quantitative survival measurements to identify factors that drive population change.

From the survival rate one can also understand the mortality rate. The mortality rate of an

individual or population is one minus the survival rate.

Using the telemetry data we compiled a table showing the number of wolves that were

alive each month, died each month, and recaptured each month. The table provided the

Mexican wolf review . . . Paquet et al. 2001 Page 30 of 86

foundation for formal analysis of survival rates. We estimated survival rates of radio-collared

wolves using the Kaplan-Meier (K-M) product limit estimator (Kaplan EL and Meier 1958). We

carried out this analysis using the programs MARK and Minitab (Version 12). Conceptually, the

analysis uses the relationships between the number of wolves that die each month and the number

monitored each month. Although estimating a rate of survival for each month is possible, the data

show that annual survival rates do not vary substantially across longer periods. Thus, we

estimated survival rates using an information-theoretic approach (Buhrnam and Anderson 1999)

that determines the most appropriate time scale (e.g., monthly, seasonally, or annually).

From the perspective of a free-ranging population, returning a wolf to captivity (from now

on, recapture event) is equivalent to a mortality event. Thus, we conducted 2 survival analyses.

One analysis considered only true biological deaths, and the other treated biological deaths and

recapture events as mortality events. In both analyses, we reincluded wolves from time of release

until “mortality” or disappearance of the radio-signal occurred.

Sample sizes were too small to use Cox's proportional hazards model and determine the

influence of important covariates (such as age and sex) on survival. We did not calculate cause-specific

mortality. Mortality was described, however, using percents. We assumed that the

proximate cause of death was the ultimate cause of death. We were unable to assess the relative

importance of other factors that may have been involved.

The starting date of the survival study was March 1998 and the end date was March 2001.

For known deaths we estimated the date of mortality to the nearest day using evidence from the

telemetry and events data bases. When information was unavailable, we deemed day of mortality

the midpoint of the interval between the last day the wolf was known alive and the day it was

discovered dead. The cause of mortality was often identified on site and when possible,

confirmed by necropsy (Interagency Field Team pers. comm.)

d. RESULTS

Forty-seven (47) wolves were monitored From March 1998 (when Mexican wolves were

first released) to March 2001. Twenty-three (23) wolves are currently being monitored. Four (4)

wolves are unaccounted for. Twenty (20) wolves were recaptured following release. Nine (9) of

these were re-released and are known to be alive. Two (2) wolves were re-released but contact

was lost and their fate is unknown. One of the re-released wolves died. Eight (8) of the

recaptured wolves were not re-released and some died in captivity. Seventeen (17) wolves are

known to have died, 10 in the wild (Figure 12). Human caused mortality was the most common

cause of death. Of the human related deaths, most were caused by gunshots (Figure 13). Wolves

also died from distemper and parvovirus. Both these diseases are contracted or originally spread

from domestic animals. Death by disease was higher than projected in the EIS.

When recaptures were included as mortalities, survival rates were lower than projected in

the EIS (Figure 14). Excluding recaptures as mortalities resulted in survival rates exceeding the

EIS projections in 1999 and 2000 (Figure 15). Survival rates from either method, however, were

lower than for wolves in the Flathead region of Montana and British Columbia (Pletscher et al.

1997), lower than for wolves in the central Canadian Rocky Mountains, lower than a recovering

wolf population in the Yukon (Hayes and Harestad 2000), and higher than an exploited

population in Alaska (Ballard et al. 1987).

Mexican wolf review . . . Paquet et al. 2001 Page 31 of 86

Other natural causes

13% (n=2)

Disease

31% (n=5)

Human-caused

56% (n=9)

Figure 12. Causes of wolf mortality for Mexican wolves reintroduced to Arizona, 1998-

2001.

Vehicle

19% (n=3)

Parvovirus

19% (n=3)

Gunshot

37% (n=6)

Distemper

13% (n=2)

Mt. lion

Brain tumor 6% (n=1)

6% (n=1)

Figure 13. Cause specific wolf mortality for Mexican wolves reintroduced to Arizona,

1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 32 of 86

0.21

0.02

0.53

0.08

0.19

0.76

0.0

0.2

0.4

0.6

0.8

1.0

SEASON & YEAR ± 1 S.E.

Projected

survival

Figure 14. Survival analysis of reintroduced Mexican wolf population assuming that

recapture represents a mortality event. Analysis was conducted for the period 1998-

2001.

0.34

0.81

0.88

0.0

0.2

0.4

0.6

0.8

1.0

1998 1999 2000

YEAR ± 1 S.E.

Projected

survival

Figure 15. Survival analysis of reintroduced Mexican wolf population assuming that

recaptures do not represent a mortality event. Analysis was conducted for the period

1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 33 of 86

e. CONCLUSIONS

Frequent removals and re-releases of wolves confounded our analysis of rates and causes

of mortality. However, if recaptured wolves were at high risk of being killed, then survival is

much lower than projected in the EIS. Human-related deaths were the greatest source of

mortality for reintroduced Mexican wolves. Shooting was the major source of death. Numerous

other studies have reported human-caused deaths as the major cause of wolf mortality (Fuller and

Keith 1980, Berg and Kuehn 1982, Boitani 1982, Carbyn 1982, Ballard et al. 1987, Fuller 1989,

Mech 1989, Pletscher et al. 1997, and many others).

Mexican wolf review . . . Paquet et al. 2001 Page 34 of 86

8. IS POPULATION GROWTH SUBSTANTIALLY LOWER THAN

PROJECTED IN THE EIS?

a. BACKGROUND

Rates of increase in wild wolf populations have varied between 0.93 and 2.40 (Fuller and

Keith 1980, Fritts and Mech 1981, Ballard et al. 1987, Hayes et al. 1991, Messier 1991, Pletscher

et al. 1997). Several factors limit growth of wolf populations; those reported most commonly

include ungulate biomass (Van Ballenberghe et al. 1975, Mech 1973, 1977a, 1977b, Fuller and

Keith 1980, Packard and Mech 1980, Keith 1983, Messier 1985a, 1987, Peterson and Page 1988)

and human-caused mortality (Van Ballenberghe 1981, Gasaway et al. 1983, Keith 1983, Peterson

et al. 1984, Fuller 1989, Paquet et al. 1996, Noss et al. 1996). Keith calculated the maximum rate

of increase for wolves (r = 0.304, ~ = 1.36) (1983) based on the highest reproductive and survival

rates reported from studies on wild wolves. He corroborated the results by comparing the

estimate with data from wolves that colonized Isle Royale National Park, 1952-1959 (r = 0.304,  

= 1.39). These were likely maximum rates of increase because the population was initiated by few

individuals with abundant food (Keith 1983). However, both rates are still much lower than a

theoretical exponential rate of 0.833 (  = 2.30) given maximum reproduction (Rausch 1967), a

stable age distribution and no deaths.

Keith (1983) suggested the amount of food available and age structure of the population

affect rates of growth of wolf populations. VanBallenberghe (1981), Gasaway et al. (1983),

Keith (1983), Peterson et al. (1984), Ballard et al. (1987), and Fuller (1989) found that wolf

populations can be limited by harvest levels of 20-40%, but that the lower rate has a more

significant effect in an area with low ungulate biomass (Gasaway et al.1983). Another factor to

consider is that effects of harvest vary with time and population structure (Peterson et al. 1984,

Fuller 1989). If productivity is high, and consequently the ratio of pups to adults is high, the

population can withstand a higher overall mortality because pups (non-producers) make up a

disproportionate amount of the harvest (Fuller 1989). Furthermore, net immigration or

emigration may mitigate the effects of harvest (Fuller 1989).

b. DATA SUMMARY

We assessed the density of the wolf population, size of established packs, and population

growth using radiotelemetry data and direct observation by the Interagency Field Team. Most of

these data are contained in the Monitoring and Events databases.

c. METHODS

We calculated density of wolves/1000 km2 by determining intra-pack densities (home

range size/number of wolves in pack) of radio-collared wolves and averaging these densities per

year (Potvin 1987, Bjorge and Gunson 1989, Okarma et al.1998). The size of packs was based

on numbers of wolves observed during midwinter aerial locations (15 Jan-15 Feb). We estimated

population growth using finite rates of increase ( ) based on the ratio of successive yearly

estimates of density. Mean annual finite rate of increase was calculated by taking the

antilogarithm of the mean exponential rate of increase (r = ln  ) for the population (Fuller 1989).

Mexican wolf review . . . Paquet et al. 2001 Page 35 of 86

The fundamental equation of population demography for a closed population is:

Nt = Nt – 1+ Bt – Dt

where Nt = population size at time t, Bt = number of recruits at time t, Dt = number of deaths at

time t,

For a wild population, removals are similar to mortality and re-releases similar to

recruitment. Therefore, the equation that best describes the reintroduced Mexican wolf

population is:

Nt = Nt – 1+ Bt – Dt +  t –  t

where  t = (unpredictable) removals of ‘naughty’ wolves,  t = subsequent re-releases of those

‘naughty’ wolves,  t >> Bt,  t >> Dt

d. RESULTS

From available databases and discussions with the Interagency Field Team, we identified a

number of events relevant to assessment of population dynamics (Table 2). Using this

information, we calculated population growth rates (Figures 16, 17) and the varying number of

free-ranging wolves over time (Figures 18 and 19). Growth rates and numbers of wolves were

close to projections, although frequent re-releases and removals obscured comparisons. To

provide context for interpreting these results, we also generated mean growth rates for other

reintroduced and recovering wolf populations (Figures 20, 21, 22). To date, the growth rate of

the reintroduced Mexican wolf population is comparable with similar reintroduction and recovery

efforts.

Table 2. Population events recorded for reintroduced Mexican wolf population between 1998

and 2001..

POPULATION

EVENT

NUMBER

Recruits 3 - 5

Re-releases 21

Deaths 10 - 16

Removals 31

Mexican wolf review . . . Paquet et al. 2001 Page 36 of 86

0

0.2

0.4

0.6

0.8

1

1998 2000 2002 2004 2006 2008

Year

Projected

Actual

Figure 16. Projected and actual annual growth rates of free-ranging Mexican wolf population.

Actual growth rate is strongly influenced by frequent intervention.

Mexican wolf review . . . Paquet et al. 2001 Page 37 of 86

0

20

40

60

80

100

120

1998 2000 2002 2004 2006 2008

Year

Projected

Actual

Figure 17. Projected and actual sizes of free-ranging Mexican wolf population, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 38 of 86

0

5

10

15

20

25

30

Mar-98 Sep-98 Mar-99 Sep-99 Mar-00 Sep-00 Mar-01

Date

Minimum

Maximum

Figure 18. Number of free-ranging radiocollared Mexican wolves, 1998-2001. The

difference between the max and min accounts for 4 wolves whose signals were lost, and in

one case, a wolf that threw its collar.

Mexican wolf review . . . Paquet et al. 2001 Page 39 of 86

Scandinavia

10

30

50

70

1975 1985 1995

Michigan

0

50

100

150

200

250

1985 1990 1995 2000

0

20

40

60

80

1975 1985 1995

0

40

80

120

160

1996 1998 2000

Year

Yellowstone

Idaho

NW Montana

Figure 19. Growth rates of other recovering wolf populations. Sources:

<http://www.r6.fws.gov/wolf/annualrpt99/> and unpublished documents from JAV

Mexican wolf review . . . Paquet et al. 2001 Page 40 of 86

0

0.3

0.6

0.9

Figure 20. Mean annual growth rate for other recovering populations.

0

50

100

150

200

250

0 3 6 9 12 15 18 21

Year

NW Mont

Yellowstone

Idaho

Michigan

Scandinaiva

Figure 21. Number of wolves over time in other recovering populations.

Mexican wolf review . . . Paquet et al. 2001 Page 41 of 86

Assessing the average growth rate only tells part of the story. Fluctuations in growth rates

are also critical. The more fluctuation the greater the extinction risk. In this case, to assess

fluctuations, we need to examine the population trajectory on a different time scale.

Using data collected since March 1998, we calculated a 39% chance that the annual

growth rate is < 0.0; a 43% chance the annual growth rate is  0.10; and a 50% chance the annual

growth rate  0.20 (Figure 22). Using data collected since December 1998, we calculated a 23%

chance that the annual growth rate is < 0.0; a 26% chance the annual growth rate is  0.10; and a

29% chance annual growth rate  0.20 (Figure 23).

Mexican wolf review . . . Paquet et al. 2001 Page 42 of 86

-0.3 -0.2 -0.1 0 0.1 0.2 0.3

Growth Rate (monthly)

r = 0.02

Figure 22. Mean onthly growth rate (r) since March 1998. The expected value of r is 0.02.

The standard error is 0.07.*

*A monthly growth rate of 0.083 corresponds to an annual growth rate of 0.1. A monthly growth

rate of 0.0166 corresponds to an annual growth rate of ~0.2.

Mexican wolf review . . . Paquet et al. 2001 Page 43 of 86

Growth Rate (monthly)

-0.3 -0.2 -0.1 0 0.1 0.2 0.3

r = 0.06

Figure 23. Mean monthly growth rate (r) since December 1998 (when population went

temporarily extinct). The expected value of r is 0.06. The standard error is 0.08.

Mexican wolf review . . . Paquet et al. 2001 Page 44 of 86

Need to find a balance

Figure 24. Because of frequent interventions the vital rates we derived (survival and

population growth) are unlikley to reflect the population’s future viability. A balance between

intervention and the effects of natural population processes is needed.

e. CONCLUSIONS

To date, intervention has dominated natural processes. So, determining if the growth rate

is lower than predicted in the EIS is not possible. If the current rate of intervention continues,

restoration of a population of 100 wolves would require 28 re-releases annually and 41 removals

annually. Although the current population size is similar to that projected in the EIS, we suspect

that population growth would have fallen far short of expectations without intervention. Clearly,

managers must balance future introductions, recaptures, and re-releases with the need to establish

and maintain natural population processes (Figure 24).

Mexican wolf review . . . Paquet et al. 2001 Page 45 of 86

3 Territory and home range size is more closely correlated with pack size than with prey

density (Messier 1985, Peterson et al. 1984). In areas of higher prey density pack sizes increase

(Messier 1985). Messier’s (1985) data indicate that between 0.2 and 0.4 moose/km², territory

area per wolf is independent of moose abundance.

9. ARE NUMBERS AND VULNERABILITY OF PREY ADEQUATE TO

SUPPORT WOLVES?

a. BACKGROUND

Without human disturbance, densities reflect the wolf’s dependency on ungulate prey

species (Keith 1983). Wolf population dynamics are believed to be largely dictated by the per

capita amount of prey and its vulnerability to predation, and the degree of human exploitation

(Keith 1983; Fuller 1989). The effect of food on wolf demography is mediated by social factors,

including pack formation, territorial behavior, exclusive breeding, deferred reproduction,

intraspecific aggression, dispersal, and by primary prey shifts (Keith 1983).

Wolf populations are closely linked to population levels of their ungulate prey (Keith

1983, Messier 1985a, Fuller 1989). Maintaining viable, well-distributed wolf populations depends

on maintaining an abundant, available, and stable ungulate population. Packard and Mech (1980)

concluded that intrinsic social factors and the influence of food supply are interrelated in

determining population levels of wolves. In situations where other factors reduce prey

populations (e.g., winter weather), predation by wolves can inhibit the recovery of prey

populations for long periods (Gasaway et al. 1983). In a multiprey system, the stability, or

equilibrium, of ungulate prey and wolf populations seems to depend on a variety of factors,

including the wolf predation rate, the number of ungulates killed by hunters, the ratio of ungulates

to wolves, and the population growth rate of different ungulate species (Carbyn 1982, Huggard

1992, Paquet 1993, Paquet et al. 1996, Paquet 1989).

Changes in habitat composition and distribution can have a significant effect on prey

densities and distributions, and therefore wolf spatial distribution. Wolf packs may react to

changing conditions in varying ways, depending on the location of their territories in relation to

other packs and prey distribution. If packs have lower prey densities within their territories, they

may exploit territories more intensely.3 This may be achieved by 1) persevering in each attack, 2)

using carcasses thoroughly, 3) feeding on alternative and possibly second -choice food resources

such as beaver (Castor canadensis) (Messier and Crete 1985), and 4) patrolling their territory

more intensely (Messier 1985). Messier, in his study area in southeastern Quebec, found daily

distances of Low Prey packs were on average either greater (summer) or equal (winter) to daily

distances of High Prey packs. The territory size, however, was approximately 35% smaller in the

Low Prey area, supporting the fact that wolves were searching each unit area with greater

intensity in both seasons.

Many studies emphasize the direct effects (e.g., prey mortality) wolves have on the

population dynamics of their ungulate prey (Carbyn 1974, Mech and Karns 1977, Carbyn 1983,

Gasaway et al. 1983, Messier 1994, Messier and Crete 1985, Peterson et al. 1984, Gunson 1983,

Mexican wolf review . . . Paquet et al. 2001 Page 46 of 86

Ballard et al. 1987, Boutin 1992, and others). However, predation can also profoundly affect the

behaviour of prey, including use of habitat, time of activity, foraging mode, diet, mating systems,

and life histories (Sih et al. 1985). Accordingly, several studies describe the influence wolves

have on movements, distribution, and habitat selection of caribou (Rangifer tarandus), moose,

and white-tailed deer (Mech 1977, Stephens and Peterson 1987, Ballard et al. 1987, Nelson and

Mech 1981, Messier and Barrette 1985, Messier 1994). Wolves can increase the rate at which

they accrue resources by seeking out areas with dense concentrations of prey (Huggard 1991,

Weaver 1994). Prey, in turn, can lower their expected mortality rate by preferentially residing in

areas with few or no wolves. Several studies have suggested that ungulate prey seek out

predator-free refugia to avoid predation by wolves (Mech 1977, Holt 1987, Paquet 1993). Wolf

predation in the Superior National Forest of northern Minnesota was found to affect deer

distributions within wolf territories (Mech 1977). Densities were greater along edges of

territories where predation was thought to be less.

Unusually mild or severe winter weather can result in ungulate populations that are

temporarily higher or lower than predicted habitat capability (which reflects long-term average

maximum). Where predation is a factor, ungulates may exist at levels well below carrying

capacity for relatively long periods. The interactions of ungulates and their predators (in our case

wolves, coyotes, foxes, black bears, and cougars) may, under some circumstances, overshadow

habitat capability as a controlling factor for ungulate populations. Ungulate populations may be

more strongly influenced by the frequency and depth of population lows, than by habitat

capability.

Ungulate biomass can affect rates of population increase and resulting densities of wolves.

Building on work of Keith (1983), Fuller (1989) reviewed 25 studies of North American wolf and

prey populations and found rates of increase of wolf populations are most affected by relative

availability of ungulate biomass (directly influencing survival of pups <6 months old) and

human-caused mortality. He concluded that regardless of prey type or stability of wolf

populations, average wolf densities are clearly correlated with the biomass of ungulates available

per wolf. Furthermore, he found the index of ungulate biomass per wolf is highest for heavily

exploited (Ballard et al. 1987) or newly protected (Fritts and Mech 1981) wolf populations and

lowest for unexploited wolf populations (Oosenbrug and Carbyn 1982, Mech 1986) or those

where ungulates are heavily harvested (Kolenosky 1972).

b. DATA SUMMARY

We used information in the carcasses database to assess wolf use of prey species. Prey

densities and the weights of prey were derived from Groebner et al. (1995).

c. METHODS

We estimated potential wolf numbers using regression equations that relate wolf numbers

to ungulate biomass (Keith 1983, Fuller 1989). The equations were modified to reflect prey

Mexican wolf review . . . Paquet et al. 2001 Page 47 of 86

4Y = 0.041X

where Y = wolf numbers, X = prey biomass

species available to wolves in Arizona and New Mexico.4 Accordingly, biomass was calculated by

multiplying population densities of elk, white-tailed deer, and mule deer (O. hemionus) by average

edible weights of elk, white-tailed deer, and mule deer. We used weights of 159 kg (350 lb.) for

elk, 36 kg (80 lb.) for white-tailed deer, and 55 kg (122 lb.) for mule deer (Groebner et al. 1995).

We used prey densities of 1.1 km² for elk, 0.9 km² for white-tailed deer, and 2.8 km² for mule

deer (Groebner et al. 1995). Assuming that ungulate populations would decline slightly in the

presence of wolf predation, prey densities were reduced 10% in our final calculations. We

assumed prey were evenly distributed and equally available throughout the primary and secondary

release sites. Bighorn Sheep (Ovis canadensis), pronghorn (Antilocapra americana, javelina

(Tayassu tayacu), and beaver (Castor canadensis) were not included in our analyses because no

population data were available.

d. RESULTS

The Interagency Field Team recorded 55 probable wolf kills. Elk constituted 85%, mule

deer 7%, and deer of unknown species about 4% of recorded kills. The predominance of elk in

the diet was consistent among packs (Figure 25). Based on numbers of prey available and

biomass available within the primary release site, elk were used disproportionately. Note,

however, that observational bias may skew collection of kill data. Elk are easier to find because

they are larger than deer and not consumed as rapidly. In addition, the seasonal movements of

wolves and their prey can affect spatial overlap and thus availability. Lack of data and time

prevented us from assessing this possibility.

Based on ungulate biomass, the Blue Range Wolf Recovery Area (6,854 mi² or 17,751

km²) can, in theory, support a an estimated 468 wolves (range 292-821). The target recovery

area of 12,950 km² (5,000 km²) could support between-212 and 599 wolves (Figure 26) (Table

3). We believe these estimates are high because they assumes all prey are equal and will be

consumed in proportion with availability. Given our experience with multiple prey systems

elsewhere this is unlikely to occur. We therefore calculated wolf population estimates for

individual prey species. Accordingly, elk in the Blue Range Wolf Recovery Area could support

about 213 wolves, and the combined deer species about 255 wolves.

Mexican wolf review . . . Paquet et al. 2001 Page 48 of 86

Pack

Campbell Cienega Francisco Hawk's Mule Pipestem

0

5

10

15

20

W.T. Deer Mule Deer

Elk Cow

Deer

Figure 25. Prey (n = 55) probably killed by reintroduced Mexican wolves, 1998-2001.

Mexican wolf review . . . Paquet et al. 2001 Page 49 of 86

Biomass Index (100 km²)

0 200 400 600 800 1000

0

200

400

600

Elk WT Deer Mule Deer All Prey

Potential Wolf Numbers (Mean)

Y = 0.041X + 0

Figure 26. Potential number of wolves that, in theory, could occupy target objective of 12,950

km² (5,000 mi²) within the Blue Range Wolf Recovery Area. Estimates are based on prey

biomass available to wolves and are maximum numbers. The individual contribution of

ungulate prey species is shown for comparison with other studies.

Mexican wolf review . . . Paquet et al. 2001 Page 50 of 86

Table 3. Potential wolf numbers (ranges) for recovery areas based on predicted population

densities of ungulates 5 years post restoration of Mexican wolf population. We partitioned the

table to show the contributions of different ungulate species.

PREY

SPECIES

Primary

Zone

(2,664 km²)

Recovery

Objective*

(12,950 km²)

BRWRA*

Low

(17,563 km²)

BRWRA*

High

(17,563 km²)

White-tailed

Deer

10-13

Mule Deer 46-63

White-tailed

and Mule Deer

118-323 162-245 293-443

Elk 50-67 94-276 129-195 250-378

All Prey 106-143 212-599 292-441 543-821

*For white-tailed and mule deer, we used an average biomass.to derive wolf estimates.

Mexican wolf review . . . Paquet et al. 2001 Page 51 of 86

e. CONCLUSIONS

Given the current ratio of wolves to ungulate prey, we conclude the reintroduced Mexican

wolf population is not limited by food. Adequate prey are available to support and sustain a

growing wolf population. Estimated wolf numbers derived from ungulate biomass were similar to

numbers projected in the EIS. Because wolves depend primarily on ungulates for food, long-term

survival of wolves in the study region depends primarily on protection of habitat for elk and deer.

Mexican wolf review . . . Paquet et al. 2001 Page 52 of 86

10. HAS THE LIVESTOCK DEPREDATION CONTROL PROGRAM

BEEN EFFECTIVE?

a. BACKGROUND

Although an effective livestock depredation program is critical for wolf recovery, effective

assessment of such a program requires more specific guidance and data than we were provided.

b. DATA SUMMARY AND METHODS

Our analysis is based on interpreting records in the Events and Incidences databases.

c. RESULTS

Forty-two (42) reports of possible wolf-livestock interactions were recorded between

March 1998 and March 2001. Of these, the Interagency Field Team concluded that 5 events were

accidents, 9 were non-wolf predators [e.g., bear (Ursus americanus), lion (Felis concolor),

coyote (C. latrans)], 18 were wolf related, and 10 were probably wolf related. That is, 28 events

involved wolves or probably involved wolves. These included uninjured livestock, injured

livestock, and killed livestock (Table 4). The Interagency Field Team recorded 10 confirmed

livestock-wolf interactions where no injury or death occurred. At a minimum, 55% (26) of all

free-ranging wolves have interacted with livestock. Thirty-six percent (17) have interacted with

livestock 3 or more times. Approximately 10% have interacted with livestock 5 or more times.

Approximately three-quarters of the livestock injuries or deaths occurred on National Forests.

The number of reported livestock-wolf interactions varied seasonally (Figure 27). The

interactions reported annually since the first reintroduction of Mexican wolves were; 5 from

March 1998 to March 1999, 17 from March 1999 to March 2000, and 6 from Mar 2000 to Mar

2001.

Seventeen (17) reports of wolf interactions with cats or dogs were recorded between

March 1998 and March 2001. These 17 reports included uninjured dogs, injured dogs, and killed

dogs or cats. Of these, we concluded that; 13 interactions involved wolves; 1 interaction probably

involved a wolf, and; 3 interactions cannot be classified using the data provided. The Interagency

Field Team recorded 8 dog-wolf interactions where no injury or death occurred. Of the 13

interactions that definitely involved wolves, 5 resulted in the cat or dog being killed or injured

(Table 5).

The average response time for all reported domestic animal-wolf interactions was less than

24 hours. The longest response time was 3 days, which occurred once.

Mexican wolf review . . . Paquet et al. 2001 Page 53 of 86

Table 4. Numbers of domestic animal injuries and deaths due to wolf depredation. The data are

for confirmed, probable and unconfirmed wolf depredations.

SPECIES

OUTCOME OF

INTERACTION

Injured Killed

Cow 1 5

Calf 2 8

Bull 1 1

Mini Colt 0 1

Lamb 0 1

Dog 3 1

Cat 0 1

Total 7 18

Table 5. Ownership of property where domestic animal injuries and death due to wolves took

place. The data are for confirmed, probable, and unconfirmed wolf depredations.

OWNERSHIP LIVESTOCK

INJURIES

OR DEATHS

CAT/DOG

INJURIES

OR

DEATHS

National Forest 14 1

Private 3 2

Other or not

recorded

2 2

Mexican wolf review . . . Paquet et al. 2001 Page 54 of 86

0

2

4

6

8

10

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Month

Number of confimred livestock-wolf

interactions

Figure 27. The number of livestock-wolf interactions fluctuated seasonally in the primary

recovery zone.

d. CONCLUSIONS

Livestock are omnipresent in the Blue Wolf reintroduction area. Because of the extensive

temporal and spatial distribution of livestock, interactions with wolves are unavoidable. From the

information made available to us, we believe the Service has been responsive to wolf-livestock

and wolf-domestic animal conflicts. An equivalent level of responsive will be necessary in the

future. Similarly, livestock producers using public lands can make a substantive contribution to

reducing conflicts with wolves through improved husbandry and better management of carcasses.

Mexican wolf review . . . Paquet et al. 2001 Page 55 of 86

11. HAVE DOCUMENTED CASES OF THREATS TO HUMAN SAFETY

OCCURRED?

Although no injuries or deaths have occurred, several wolf-human interactions have been

reported. Consequently, evaluation of these incidences is largely qualitative based on our

experiences with wolves in other parts of North America. We note that captive reared wolves

released to the wild may behave differently than wild born wolves (Breitenemoser et al. in press).

a. DATA SUMMARY AND METHODS

Our analysis of this issue is based on interpreting records in the Events and Incidences database.

b. RESULTS

The Interagency Field Team reported eleven interactions between March 1998 and March

2001 (Table 6). On average, they reported one event every 3 months. However, the rate may be

increasing (3 events from Mar 1998 to Mar 1999, 1 event from Mar 1999 to Mar 2000, 7 events

from Mar 2000 to Mar 2001). If the rate is increasing, it is probably due to more wolves rather

than an increased propensity for wolves to interact with humans. On average, one interaction was

reported every 7 weeks from Mar 2000 to Mar 2001. Although data are too few to be certain,

interactions do not seem to predominate in any particular time of the year.

Seven (of 11) interactions involved something that would be expected to attract wolves

(e.g., dogs, deer carcass, livestock). Specifically, 5 (of these 7) involved dogs. One (of 11)

interaction was instigated by the people involved (event #10). In 2 (of 11) events, the people

involved felt as though their lives were threatened. In 4 (of the 11) events, an official response

(i.e., from reintroduction personnel) occurred within 24 hours. In the other 7 events, no response

date or time is reported. In 9 (of the 11) events, response involved an inspection of the site.

In 2 events (# 1 and #7), the people involved reported being fearful for their safety.

However, experience suggests that because the people of event #7 responded appropriately, they

were probably never in danger. In event #1, the wolf was shot. Event #8 is similar to cases in

Ontario, British Columbia, and Alaska where wolves have injured people. In these all these cases,

the people responded inappropriately to curious wolves or wolves attracted to food.

Twelve (12) different wolves have been involved with human interactions. Approximately

25% of all the wolves that have been released into the wild have been involved in a reported wolf-human

ineraction. Eight (of these 12) wolves were involved in only a single event. One (of these

12) wolves (i.e., 590) was involved in 4 events. All these events took place in August and

September of 2000. Since then, wolf 590 has not been involved in any human interactions. Three

(of the 12) wolves (i.e., 587, 509, 511) were involved in 3 events. All 3 events included wolf 590.

The ‘immediate’ fate of the 12 wolves was: 1 shot, 2 brought into captivity, 1 brought into

an acclimation pen, and in 8 cases no attempt was made to capture the wolf. The ‘ultimate’ fate

of of the 12 wolves was: 2 shot, 3 permanently brought into captivity, 6 either are still free-ranging

or died of natural causes, and for 1 wolf (i.e., #298, the potential data entry error) no data

were available.

Mexican wolf review . . . Paquet et al. 2001 Page 56 of 86

Table 6. Summary of wolf-human interactions reported for the Mexican wolf reintroduction

program, 1998-2001.

E

V

E

N

T DATE

WOLVES

INVOLVED MEMO

1 April 28, 1998 156 Wolf 156 was shot by a camper who feared for his

family's safety when the wolf came into their camp

and attacked their dog.

2 May 8, 1998 494 494 became a nuisance frequenting the town of

Alpine from 5/8/98 through 5/28/98 and was

permanently removed from the wild.

3 January 6, 1999 166, 482 Campbell Blue pair jerked down a deer carcass

hanging in some archery hunter's camp.

4 January 5, 2000 522 Female 522 hanging around hunters camp

interacting with dogs. Trapped and put in

acclimation pen to hold through hunting season.

5 April 14, 2000 166, 518 Dean Warren reported very aggressive encounter

with Campbell Blue pair with the female, 518

bumping his horse and passing under it. Wolves

also attacked one of his dogs. They followed him to

cabin and he held up in it until the wolves left.

6 May 16, 2000 298, 191 Renee Dupree jogging with 2 dogs when 2 wolves

approached -- wolves clearly interested in dogs.

Renee scares wolves away.

Mexican wolf review . . . Paquet et al. 2001 Page 57 of 86

7 August 20, 2000 511, 509, 587,

590

Don and his cocker spaniel were out in the middle

of the meadow behind his trailer when 4 wolves (

most likely Francisco) came tearing out of the

woods towards them. Don fired 1 hot in front of

the wolves but they kept coming ("one with a look

of fierce determination"). He fired a second shot as

they got closer and they reared away. He was very

upset at the situation and felt that they were a

danger to both people and animals/pets. Later that

week, people camped nearby observed several

wolves and pups resting in the shade under and

around Don's trailer. At the time, he was inside

watching golf with his dog, unaware that the wolves

were outside. He was irrate when he learned of the

incident, stating that this was not the behavior of

wild animals and concerned about what would have

happened had he or his dog come out of the trailer.

8 August 24, 2000 511, 509, 587,

590

Scott observed Francisco (and Cienega) on multiple

occasions during his time camping at Double

Cienega. Sometimes they came right through cmp <

5 ft of him taking pictures, although the pups

seemed more skittish, other times farther away

within the campground or out in the meadow. He

also saw them once farther up Double Cienega and

"the shaggy one" (yearling male 590) laid down w/in

10 ft and just looked at him while he took pictures.

9 September 25,

2000

590 Yearling male 590 hanging around Double Cienega

Campground for the majority of the day.

Mexican wolf review . . . Paquet et al. 2001 Page 58 of 86

10 September 29,

2000

511, 509, 587,

590

5-6 people camped in Double Cienega from about

8/21-8/30/00. Throughout the week they interacted

with Francisco. On multiple occasions they howled

the pack in, chased them on ATVs, left food out,

and shot blunt arrows at them. The wolves also

chased their horses, mules, and the people in the

ATVs. They were informed that this behavior was

not acceptable, and we explained that what they

were doing may possibly have negative effects on

the wolves behavior. On 8/30/00, while speaking

with the hunters, N. Sanchez observed the wolves

chasing the mules. He then hazed the wolves by

running at them and throwing rocks. They ignored

him. We first spoke with the group on about

8/23/00. We informed them about the Mexican

Wolf Recovery Project, the presence of wolves in

the area, and proper behavior with respect to the

wolves (ie. Do not leave out food; keep an eye on

mules/ horses; if you see the wolves, yell and throw

rocks at them.) We also told them to let us know if

they had any interactions with the wolves.

11 October 1, 2000 Unknown At about 0440 Cole went out the front door on the

porch and observed an animal in the driveway. At

first he thought it was a German Shepard, then by

the color and size he realized it was a wolf. He

shewed it away and it headed west down the road.

He tried to follow it in his truck but lost track of it.

When he got back to the house it was by the back

door eating out of the dog dish. He shewed it away

again and it ran behind the house between the

animal pens and the barn. He checked the dog dish

and it was empty. He was not sure if there had been

food in it or not. Stark and Grant responded to the

call made by Ms. Leona Brown (the landowners

sister). We looked at the area where the report was

taken and observed large canid tracks in the

driveway and yard. (track size=5x3 1/2", in sand

and gravel). No other tracks were found in area.

Stark and Armistead returned on 10/2 at about

0500.

Mexican wolf review . . . Paquet et al. 2001 Page 59 of 86

c. CONCLUSIONS

Wolf-human interactions have been reported consistently and regularly since the beginning

of the program. Approximately 25% of the individuals in the free-ranging population have been

involved with wolf-human interactions. As the wolf population grows, the Program should be

prepared for steadily increasing frequencies of wolf-human interactions. Over time, the frequency

of wolf-human interactions (per wolf) may decline with wild-born wolves that are less tolerant of

humans. Because wolves can pass information between generations, the attraction to humans

may take some time to extinguish. In the Republic of Georgia, for example, captive-born wolves

were intensively trained to kill wild prey and to avoid humans before their reintroduction. This

release procedure was considered successful after the third generation of wild-born wolves still

showed the same behavior as their hand raised parents (J. Badrize pers. comm.).

The Program has responded well to wolf-human interactions, although documentation and

data recording have been poor. For example, in the databases USFWS provided us no response

dates or times were recorded for 7 events. It is critical that the Interagency Field Team keep

comprehensive notes on wolf-human interactions. The Program should continue its practice of

responding to all wolf-human interactions with immediate on site inspections and investigations.

The Interagency Field Team appears to have made responsible decisions regarding the recapture

of wolves involved in human interactions.

Mexican wolf review . . . Paquet et al. 2001 Page 60 of 86

12. OVERALL CONCLUSIONS AND RECOMMENDATIONS

a. PREFACE

On 25 April we convened a meeting in Globe, Arizona to present our draft report to the

Mexican Wolf Interagency Management Advisory Group (IMAG). We purposefully presented a

draft to provide the IMAG a chance to make substantive contributions to our review. Many

comments we received during the meeting clarified issues, thus materially improving our review.

During the week of 30 April the draft report was, without our knowledge released to the media.

During the following weeks several newspaper stories presented the findings of our draft review

as final determinations. Moreover, on 12 May the Arizona Game and Fish Commission received a

briefing about the reintroduction from representatives from the Arizona Game and Fish

Department who also presented our draft findings as final determinations. Draft reports are by

definition works in progress. Any discrepancy between the conclusions and recommendations

presented in the draft report and those presented here are a result of that simple fact.

Our conclusions and recommendations are based on our analysis of the data. We believe

the long term objective is to protect the wolf population and meet human needs by reducing the

potential for one to seriously encroach upon the other. Current circumstances demand that

wolves be conserved in a human dominated landscape. This requires a systematic and rigorous

approach to wolf recovery that integrates the social and economic aspirations of humans with the

ecological necessities of wolves.

b. CONCLUSIONS

The ultimate factor determining population viability for wolves is human attitude. Thus,

an active and fully enabled Recovery Program comprising private interests, non governmental

conservation organizations, local, state, federal, and tribal agencies is essential to ensure success

of any restoration. The biology, politics, and sociology of wolf reintroduction in the Blue Range

Wolf Recovery Area are too complex for recovery to be successful without a fully engaged and

participatory Program. Fortunately, the Service has a successful history of reintroducing and

effectively managing recovered wolf populations in other parts of the country (Refsnider 2000).

Based on this success and the first 3 years of the Mexican wolf reintroduction, we think that

expecting a similar outcome in the Blue Range Wolf Recovery Area is reasonable.

Overall we are satisfied with the progress of the reintroduction project since its inception

in 1998. During May 2001, the Service reported that at least 28 wolves were free-ranging. Most

of these animals are in social groups and the Service reports up to 5 litters have been produced in

the wild this spring. Monitoring of reintroduced wolves has revealed that captive-born Mexican

wolves can adjust to life in the wild by primarily preying on elk. This fact combined with the

likely presence of several litters in the wild bodes well for the future. We believe the likelihood is

high that continued application of the Service’s current practices will result in the restoration of a

self-sustaining population of Mexican wolves in the Blue Range Wolf Recovery Area. We

believe, however, the Program should continue with some adjustments and modifications.

Not surprisingly, our review revealed room for improvement. Restoration of any wildlife

population is fraught with uncertainty and work elsewhere shows that many projects are

Mexican wolf review . . . Paquet et al. 2001 Page 61 of 86

unsuccessful because of a failure to accommodate new information (Breitenmoser et al. in press).

Several factors currently hinder recovery of a self-sustaining and viable wolf population. Those

that predominate are:

1. The small areal extent of the primary recovery zone, which greatly hinders the vigor of the

reintroduction phase of the reestablishment project

2. The Service’s insistence that wolves only inhabit the small Blue Range Recovery area,

which is at odds with the naturally extensive movements that characterize gray wolves and

current thinking regarding the viability of large carnivore populations (Noss et al. 1996).

3. The Service’s embrace of a target population of 100 wolves (EIS, page 2) when such a

population is not viable over the long term (Shaffer 1987, IUCN 1994, Noss et al. 1996,

Breitenmoser et al. in press).

c. RECOMMENDATIONS

The architects of the Mexican wolf reintroduction program properly accounted for the

inevitable uncertainty and difficulty of the project by establishing adaptive management as the

overarching operational paradigm. Consequently, our recommendations are largely the inevitable

result of the reintroduction project’s maturation. In this regard, we predict that the next review

will also identify changes that can be made for improving the program..

If the Service adopts the recommendations presented below then the effectiveness of the

reintroduction project and prospects for success will improve. Proper adoption of our

recommendations will require a long-term and diligent effort by the Service. For many of the

recommendations to be effective, biologists involved in the daily matters of the reintroduction

effort must embrace them as standard operating procedures.

The current reintroduction project will greatly influence the future of the Mexican wolf

recovery program since additional reintroduction projects will be required to remove Canis lupus

baileyi from the list of endangered and threatened wildlife. Accordingly, we used our review to

develop a few recommendations that consider Mexican wolf recovery overall. We also decided to

consider programmatic issues that are germane to reintroduction, and issues the Service did not

provide data for such as injuries resulting from capture. All of the recommendations below relate

directly to the successful restoration of Mexican wolves the BRWRC. We did not elaborate on

several biological issues, identified in our recommendations as important, because the

reintroduction process is in too early a stage to have accumulated sufficient data.

Biological and Technical Aspects

WE RECOMMEND THAT THE SERVICE:

Continue to develop appropriate opportunities to release (and re-release) wolves for at least

2 years to ensure the restoration of a self-sustaining population.

Begin developing population estimation techniques that are not based exclusively on

telemetric monitoring. As the wolf population grows it will become increasingly difficult to

Mexican wolf review . . . Paquet et al. 2001 Page 62 of 86

maintain telemetric contact with all known or suspected packs. Consequently, the Service needs

to develop non-telemetrically-based methodology (e.g., track station surveys, genetic sampling of

hair or fecal material) for assessing the distribution and size of the wolf population.

Develop data collection forms and data collection and management procedures similar to

those used by the red wolf restoration program in North Carolina.

Require biologist to promptly and carefully enter field data into a computer program for

storage and analysis. The Service should require biologists to record data on a per wolf and per

day basis. Data checking should be improved to eliminate data entry errors. In this regard,

picklists and auto filling fields can simplify data entry and improve accuracy. Lastly, the Service

should require that data files be proofed at least once before they conduct analyses. We remind

field biologist working on the project that generally 1 hour of productive time in the field requires

2 hours in the office for data management and initial analyses.

Make all data available for research and peer review.

Carefully consider using a modified #3 soft-catch trap for capturing Mexican wolves rather

than the McBride #7. We are concerned that the #7 might cause unacceptably frequent and

serious foot injuries. The Service might find that a modified #3 soft-catch trap is more

appropriate for capturing wolves that have a high probability of being re-released or that are fairly

small (e.g., smallish adults or pups). Modified soft-catch traps have been used to capture

hundreds of red wolves that are similar in size to Mexican wolves and larger gray wolves

(Quebec) with no serious foot injuries (M. Phillips unpublished data, P. Paquet unpublished data).

However, careful consideration of all aspects of capturing wolves with leghold traps will lead to a

proper decision about the use of a modified trap for capturing Mexican wolves.

Encourage research that will help to inform future Program evaluations and adjustments.

The research we suggest is beyond the scope of the current Mexican wolf program because of

resource limitations (personnel and fiscal) and the need to focus on the central mission of

reintroducing wolves. However, research partnerships with universities and other organizations

should be developed. Increasing the capacity of the Mexican wolf recovery Program, should be a

principle charge of the Recovery Team. The following areas are of contemporary conservation

and academic interest and should be research priorities:

1. Population modeling (PVA and metapopulation model) and sensitivity analysis of short-and

long-term demography and distribution

1. Assessment of new threats to population including new guild structure, disease, and

human activity.

2. Habitat viability analyses of the release area and projected population range (environment,

resources, carrying capacity, spatial characteristics, etc.)

3. Development of guidelines for decision-making in conflict situations

4. Reassessment of policies for intervention in the release phase

5. Assessment of monitoring programs

Mexican wolf review . . . Paquet et al. 2001 Page 63 of 86

6. Evaluation and design of long-term management program, including

a. Evaluation design of long-term monitoring program

1. demography and population range

2. genetic surveillance

3. health surveillance

4. long-term adaptation of individuals and population to ecosystem

5. effects on ecosystem (predation, displacement)

7. The