Humanity has a miserable track record in conserving large carnivores: from Paleolithic hunters skinning the enormous cave lion 15,000 years ago to the contemporary loss of the marsupial Tasmanian tiger. Today, several iconic members of the order Carnivora are on the brink of extinction (Amur leopards, Asiatic cheetahs), and over 75% of the world's 31 large carnivore species have experienced alarming population declines, often directly from human persecution. Yet, several species of large predators have dramatically rebounded (European gray wolf, American black bear) in the most unlikely of places: heavily human-dominated landscapes. For example, the black bear population in northwestern New Jersey (NJ), the state with the highest human densities in the United States, has exponentially increased over sixfold in just 15 years. During this period of unprecedented suburban sprawl in NJ there have been over 26,500 reported human-bear interactions including seven attacks on humans and one human fatality. Given accelerating anthropogenic landscape transformation, there simply are not enough large tracts of wildlands remaining to alone support expanding bear populations. Thus, American black bear conservation in the Anthropocene may ultimately depend upon society's tolerance for this large carnivore in areas where people live, work, and recreate. In an effort to curb bear population growth and reduce conflicts, the first regulated NJ black bear harvest in over three decades was held in 2003 resulting in an acrimonious public debate. How can objective population ecology help us make informed decisions about management actions that elicit such strong emotional responses among different stakeholder groups?

## LEARNING OUTCOMES

Students will evaluate how sex, age-class, and behavior (problem vs. normal) affect the probability that black bears in northwestern NJ die from harvest, lethal control, and other causes of mortality like vehicle strikes. Given these results, students will then propose possible explanations for the observed correlation between bear harvest rates and subsequent declines in nuisance bear behaviors reported. Informed by this remarkable dataset comprised of over 3,500 individual bears collected over 33 years, students will ultimately have a meaningful discussion about whether a carefully regulated bear harvest should be included in an integrated management strategy to conserve American black bears.

## INTRODUCTION

Large, terrestrial predators (e.g., wolves, big cats, and bears) are some of humanity's most beloved animals. Ironically, they are also some of our planet's most imperiled species (Slide #2) as 24 of the 31 largest carnivore species are experiencing population declines, and they currently occupy an average of only 47% of their historical ranges [1]. This is especially disconcerting to ecologists as the functional removal of apex predators can trigger trophic cascades in food webs resulting in unanticipated and far-reaching ecological consequences [2] (Slide #3). Surprisingly though, populations of several species of large carnivores have stabilized and are increasing in landscapes that have been heavily transformed by humans such as the European gray wolf (Canis lupus) [3] and the American black bear (Ursus americanus) [4] (Slide #4). Black bears were almost extirpated in New Jersey (NJ), United States, during the 1950s reduced to less than an estimated 100 individuals. Yet the population recovered to 450–500 bears by the mid-1990s, and then exponentially increased in abundance and spatial range reaching 3,200–3,400 bears by 2010, with densities in some areas exceeding 1.3 bears/km2 [5] (Slide #5).

### Estimating Cause-Specific Mortality Probabilities

To address questions 1 and 2, we employed a demographic modeling tool used to estimate survival and cause-specific mortality rates of mobile species, aptly named, capture-mark-recapture (CMR) methods [10] (Slide #11). The technique CMR initially allows us to estimate the probability that an animal, a black bear in our case, previously marked with a unique metal ear tag will be recaptured on another trapping occasion or physically recovered if it dies. Based on these “detection probabilities,” we then estimate the probability that a bear will transition from an “Alive State” to one of our three defined dead states (Harvested, Lethally Managed, or Dead All Other Causes) in any given year. This probability of transitioning to a dead state is one technique population ecologists employ to estimate “cause-specific mortality rates” of organisms as a function of sex, age, and physical and/or behavioral traits. For example, we now have an empirically derived estimate of the probability that an adult male will die from harvest in a given year during the study period (Slide #12).

### Estimating If There is a Correlation Between Harvest and Lethal Control Rates in Year t, and Problem Behaviors Reported in Year t + 1

To achieve objective 3, we first plotted all bear behaviors reported (i.e., Threat, Nuisance, and Normal), as well as our cause-specific mortality rates (described above) for each year we had reporting data (2001 through 2013; Slide #13). We also plotted the proportion of each bear behavior relative to other behaviors during these same years (Slide #14). We then used the statistical tool linear regression to determine if there was a significant (ɑ = 0.05) relationship between the rates of bears dying from all causes combined (total mortality) in year t and the number of nuisance behaviors reported in year t + 1. Additionally, we used the same methods to evaluate if just harvest and lethal control rates in year t explained variation in the number of nuisance behaviors reported during year t + 1 (Slide #15). Note, these rates were coupled to assess the additional influence of harvest on the background, inter-annual lethal control rates, and ultimately, increase our sample size (see Teaching Notes for more detail).

## A FINAL NOTE ON “SIGNIFICANCE” AND UNCERTAINTY IN SOCIAL-ECOLOGICAL SYSTEMS

In the ensuing discussion it is important to remind students that a defining attribute of the scientific process is an explicit attempt to quantify uncertainty. For example, we present a “significant” correlation between increasing harvest and lethal control rates and subsequent declines in nuisance behaviors reported by expressing a low calculated probability (i.e., p-value = 0.028) which informs the audience that these data are highly unlikely if the null hypothesis is true, i.e., there is no relationship between our predictor and response variables. That said, there is much we do not understand as ecological drivers (e.g., increasing bear densities, climate variability) interact in complex ways with the social dimensions inherent to the system (e.g., the probability that a person who experiences a negative bear interaction reports this interaction). As an example, untangling how much of the observed decline in nuisance bear behaviors is due to humans altering their behavior in response to educational outreach versus bears adapting to the presence of a new predator in the system, i.e., the human hunter, is likely inextricably confounded with this dataset. However, it could be addressed with the collection of additional data within a well-executed experimental design.

## CONCLUSION

The conservation of large, ecologically-influential carnivores necessitates prioritizing innovative interventions outside of protected areas as coexistence between humans and carnivores likely represents the only realistic way forward in ensuring carnivore persistence in heavily transformed, novel social-ecological systems [11]. Optimizing human-carnivore coexistence in these shared landscapes demands mutualistic coadaptation, i.e., both humans and carnivores learning from experience and altering their behaviors to minimize negative impacts on each other [12] (Slide #18).

## CASE STUDY QUESTIONS

1. Examine our tables of cause-specific mortality estimates to answer the following questions:

• How does the age-class (cub, yearling, and adult) and sex of the black bear (female vs. male) interact to influence its probability of being harvested? Remember to examine whether 95% confidence intervals overlap to determine if differences are significant. (Slide #16)

• What is the dominant source of mortality for young bears (cubs and yearlings) in this human-dominated landscape? Why do you think this category is so high? (Slide #16)

• How does behavior affect the probability that black bears in northwestern NJ are harvested and lethally controlled? (Slide #17)

2. Examine our temporal line/bar graph on Slide #13: Do you see a relationship between nuisance behaviors reported and the implementation of harvest in years 2003, 2005, and 2010–2013?

3. Examine our temporal line graph on Slide #14: In what year does the proportion of normal behaviors reported begin increasing relative to problem behaviors (that begin decreasing at this inflection point)? What event occurred during this time, and what are the implications for bear management?

4. Examine our linear regressions in Slide #15: Is the change in nuisance behaviors reported in year t + 1 more strongly correlated with total mortality in year t or harvest + management mortality rates in year t?

5. Provide three hypotheses explaining the observed correlation between implementing the new bear harvest and subsequent declines in nuisance bear behaviors reported. They do not have to be “mutually-exclusive,” i.e., two or more hypotheses could be occurring simultaneously.

## AUTHOR CONTRIBUTIONS

 Conceptualization JDR, LMA, MJRH Methodology LMA, JDR Software JDR, LMA Validation MJRH, PCC Formal Analysis JDR, LMA Investigation PCC, JDR Resources PCC, MJRH, LMA Data Curation PCC, JDR Writing—Original Draft JDR, LMA Writing—Reviewing and Editing MJRH, PCC Visualization JDR, LMA, MJRH Supervision LMA, MJRH, PCC Project Administration LMA, MJRH, PCC Funding Acquisition LMA, MJRH, PCC, JDR
 Conceptualization JDR, LMA, MJRH Methodology LMA, JDR Software JDR, LMA Validation MJRH, PCC Formal Analysis JDR, LMA Investigation PCC, JDR Resources PCC, MJRH, LMA Data Curation PCC, JDR Writing—Original Draft JDR, LMA Writing—Reviewing and Editing MJRH, PCC Visualization JDR, LMA, MJRH Supervision LMA, MJRH, PCC Project Administration LMA, MJRH, PCC Funding Acquisition LMA, MJRH, PCC, JDR

## FUNDING

Financial and logistical support was provided by the Federal Aid in Wildlife Restoration Act, New Jersey Wildlife Research W-68-R, the New Jersey Division of Fish and Wildlife, Bear Trust International, Hudson Farm Foundation, the Utah State University Presidential Doctoral Research Fellows Program, Office of Research and Graduate Studies, Ecology Center and Wildland Resources Department.

## COMPETING INTERESTS

The authors have declared that no competing interests exist.

## SUPPORTING INFORMATION

Teaching Notes:

Article title: Why does the regulated harvest of black bears affect the rate of human-bear conflicts in New Jersey?

Contact the authors for enquiries about teaching this case, including answers to any questions posed by the Case, the Teaching Notes, or Case Study Questions.

Slides:

Article title: Why does the regulated harvest of black bears affect the rate of human-bear conflicts in New Jersey?

## REFERENCES

REFERENCES
1.
Ripple WJ, Estes JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, et al.
Status and ecological effects of the world's largest carnivores
.
Science
.
2014
;
343
(
6167
):
151
62
.
2.
Estes JA, Terborgh J, Brashares JS, Power ME, Berger J, Bond WJ, et al.
.
Science
.
2011
;
333
(
6040
):
301
6
.
3.
Chapron G, Kaczensky P, Linnell JDC, von Arx M, Huber D, Andrén H, et al.
Recovery of large carnivores in Europe's modern human-dominated landscapes
.
Science
.
2014
;
346
(
6216
):
1517
9
.
4.
Gompper ME, Belant JL, Kays R.
Carnivore coexistence: America's recovery
.
Science
.
2015
;
347
(
6220
):
382
.
5.
Carr PC, Burguess K. New Jersey Black Bear Annual Status Report 2011.
New Jersey Division of Fish and Wildlife, Wildlife Services Section
,
Hampton, New Jersey, USA 08827
;
2011
.
6.
Beckmann JP, Berger J.
Rapid ecological and behavioural changes in carnivores: the responses of black bears (Ursus americanus) to altered food
.
Journal of Zoology
.
2003
;
261
(
2
):
207
12
.
7.
Wolgast LJ, Burke D, Kertz E, Messeroll J, Vreeland J. Comprehensive black bear management policy—New Jersey Department of Environmental Protection Division of Fish and Wildlife. In:
Protection NJDoE
, editor.
New Jersey
:
New Jersey Fish and Game Council
;
2010
:
45
.
8.
Johnson BB, Sciascia J.
Views on black bear management in New Jersey
.
Human Dimensions of Wildlife
.
2013
;
18
(
4
):
249
62
.
9.
Raithel JD, Reynolds‐Hogland MJ, Koons DN, Carr PC, Aubry LM.
Recreational harvest and incident‐response management reduce human–carnivore conflicts in an anthropogenic landscape
.
J Appl Ecol
.
2016
. DOI:
10.
Assessing the relative importance of different sources of mortality from recoveries of marked animals
.
Ecology
.
2004
;
85
(
4
):
930
8
.
11.
Di Minin E, Slotow R, Hunter LT, Pouzols FM, Toivonen T, Verburg PH, et al.
Global priorities for national carnivore conservation under land use change
.
Scientific Reports
.
2016
;
6
. DOI:
12.
Carter NH, Linnell JD.
Co-adaptation is key to coexisting with large carnivores
.
Trends in Ecology & Evolution
.
2016
;
31
(
8
):
575
8
.