Extraordinary movements connecting distant ecosystems by African wild dogs ( L ycaon pictus) Authors: Scott Creel, Stephani Matsushima, Henry Mwape, Bridget Mayani‐Nkhoma, Kachama Banda, Ben Goodheart, Johnathan Reyes de Merkle, Anna Kusler, Daan Smit, Johane Njobvu, Moses Muwowo, Reuben Kabungo, Michelo Mungolo, Mwamba Sichande, Catherine Sun, Carolyn A. Sanguinetti, Charlotte Krag, Lucky Chama, Mwansa Chisanga, Twakundine Simpamba, Howard Maimbo, Ella Bradford, Clive Chifunte, Craig Reid, Bruce Ellender, David Christianson, Egil Dröge, Matthew S. Becker Published: 2026 EWD‐1355, a 3‐year‐old African wild dog (Lycaon pictus), dispersed from the Luamfwa pack in Zambia's South Luangwa National Park in October of 2021, together with two of her littermate sisters. African wild dogs live in cooperatively breeding packs in which a single dominant individual of each sex typically breeds (Creel et al., 1997; Girman et al., 1997; Malcolm & Marten, 1982). Subordinates like EWD‐1355 either wait to become dominant and gain inclusive fitness by helping relatives, or disperse in search of more immediate breeding opportunities (Behr et al. 2020; Cozzi et al., 2020; Creel & Creel, 2002; Jordan et al., 2023). Dispersal is most common among 2‐ and 3‐year‐olds, particularly if a large same‐sexed cohort reaches this age together and all packmates of the opposite sex are close relatives (Behr et al. 2020; Creel & Creel, 2002). Larger dispersing groups hunt more effectively (although anthropogenic prey depletion is now changing this pattern: Creel et al., 2026) and are more likely to take over another pack by ousting the residents of their sex (Behr et al. 2020; Creel & Creel, 2002). Thus, EWD‐1355 and her sisters' emigration was not surprising, but their subsequent movement over more than 2300 km into Mozambique and back was. Over 11 months, the sisters repeatedly climbed and descended the Rift Valley wall and crossed major rivers and highways. They moved through well‐protected National Parks, less‐protected Game Management Areas (GMAs), still‐less‐protected Forest Reserves, and unprotected areas with appreciable human settlement. In unprotected areas they made rapid, linear movements (Creel et al., 2020), and rested in rocky areas or small woodland patches where encounters with people were less likely (Figure 1: EWD‐1355's track is shown by orange points). They ultimately reached the home range of a large pack in Lower Zambezi National Park on the border with Zimbabwe, but did not join them or entice any males to emigrate. Failing to find a reproductive opportunity, they moved on, and in June of 2022 EWD‐1355 died in a wire snare set for bushmeat hunting (Video S1). Her sisters' fate is not known, but several individuals are often snared in a single event because wild dogs are highly reluctant to abandon snared packmates (Becker et al., 2024). EWD‐1355's movements and fate, and those of subsequent long‐distance dispersers described below, are revealing important insights about connectivity at the national and continental scale, and the threat posed to it by bushmeat poaching with wire snares. Movements by African wild dogs dispersing from Zambia's South Luangwa National Park (SLNP) along a protected corridor have connected with North Luangwa (NLNP), Luambe (LuNP), Lukusuzi (LkNP), and Lower Zambezi (LZNP) National Parks. Recently, three males left Kafue National Park (KNP) and moved >2600 km through unprotected areas in the center of the country to connect with four females from SLNP who had moved more than 1500 km, linking the stronghold populations of the Greater Kafue and Luangwa Valley Ecosystems. An unknown wild dog recently immigrated into Liuwa Plains National Park (LPNP) at a time when all individuals (a single pack) were well known. Thus all of the wild dog populations in Zambia remain connected to other populations. Black outline with crosshatch: National Parks. Gray: Game Management Areas, with lower protection than National Parks. The 2020 Human Impact Index (Sanderson et al., 2022) is mapped from low (blues) to high (yellows): Dispersing wild dogs maneuver effectively around areas with high human activity. Wild dog illustration by Bridger Creel, used with permission. While the scale of EWD‐1355's movement was impressive, wild dogs are highly adapted to move through unfavorable landscapes (Creel et al., 2019). We have observed recent movements linking all of the major wild dog populations in Zambia, into Mozambique, and to within a few kilometers of Zimbabwe, Malawi, the Democratic Republic of Congo, and Angola (Figure 1). Within the African large carnivore guild, competitively subordinate species like the wild dog have evolved to move through the landscape in ways that are affected by their dominant competitors, the lion (Panthera leo) and spotted hyena (Crocuta crocuta) (Abrahms et al., 2025; Creel & Creel, 1996; Goodheart et al., 2022; Gorman et al., 1998; Mills & Gorman, 1997; Swanson et al., 2014; Vanak et al., 2013). The competition‐movement‐connectivity hypothesis (Creel et al., 2019) suggests that the morphological, cognitive, and behavioral adaptations that wild dogs use to deal with lions and hyenas are sufficiently general that they allow rapid, efficient passage through areas that are made unfavorable by human activity (Abrahms et al., 2016; Cozzi et al., 2020; Creel et al., 2019; Davies‐Mostert et al., 2012) by making fast, linear movements in unfavorable circumstances (Creel et al., 2020; Goodheart et al., 2022). In support of this hypothesis, hidden Markov modeling has shown that GPS‐collared wild dogs outside of protected areas adopt a distinct movement pattern that is faster (1.7‐fold increase in mean speed) and more linear (2.4‐fold decrease in mean turning angle) than typical movements within protected areas (Creel et al., 2020, their fig. 5 and table 1), similar to their movements after encountering lions (Goodheart et al., 2022). Nonetheless, the 2300 km dispersal by EWD‐1355 and her sisters was the second‐longest on record, exceeded only by an astonishing 3892 km movement (520 km straight line) from Loliondo in the Serengeti ecosystem to Tsavo West National Park (Davies‐Mostert et al., 2012; Masenga et al., 2016; Sandoval‐Serés et al., 2022). Since EWD‐1355's remarkable journey, several groups of wild dogs in Zambia have dispersed distances that approached or exceeded 2000 km, with a maximum path length of 4126 km and a maximum straight‐line distance of 419 km. In July of 2025, four 3‐year‐old females dispersed from the Mayukuyuku Pack in Kafue National Park, and three 3‐year‐old males dispersed from the same pack 2 weeks later. Prior to these departures, the pack held 18 adults and yearlings, and both sets of dispersers were part of a cohort of 19 pups that survived to adulthood. With large cohorts of same‐sexed 3‐year‐olds living in a large pack where they had low odds of breeding (Cozzi et al., 2020; Creel & Creel, 2002; McNutt, 1996; Woodroffe et al., 2020), these departures were expected. For several weeks, the females (down to three, soon after dispersing) roamed through the Lunga Lushwishi GMA, a buffer zone adjacent to their natal home range, while the three males moved eastward into unprotected areas (Figure 1, Video S2). The movements of dispersing groups are somewhat predictable when they are moving along protected corridors, but become quite unpredictable in unprotected areas (Figure 1, Video S2). The region of central Zambia to which the males moved (and eventually the females: Figure 1, Video S2) has a relatively high human footprint (Venter, Sanderson, Magrach, Allan, Beher, Jones, Possingham, Laurance, & Wood, 2016a; Venter, Sanderson, Magrach, Allan, Beher, Jones, Possingham, Laurance, Wood, & Fekete, 2016b) with a diffuse distribution that yields no single, well‐identified path between Kafue and protected areas to the east (Creel et al., 2019). Both groups reached the outskirts of major cities including Kabwe (288,598 residents), Kitwe (665,961), and Ndola (627,503) (Video S2). Each of these encounters caused the dogs to double back in the direction they had come: after moving north to Ndola, the males retraced their previous track almost exactly for 200 km to the south (Video S2). Remarkably, the males and females re‐encountered one another on September 18th and 19th in an unprotected area 72 km from Kitwe and 78 km from Ndola, after the females had moved 1404 km and the males had moved 1203 km from their natal homerange. Wild dogs have a great capacity to locate one another by scent, honed by the need to regroup without making noise that might attract dominant competitors if they are scattered when hunting. The two groups separated after one day, and the females arrived in a small forested area near Kabwe on October 10th, having moved 1899 km. Within this small forest patch, the collared female moved repeatedly back and forth to a fixed location, an unusual pattern for wild dogs except when they have newborn pups who cannot move with the group, or are returning to an injured packmate. Wild dogs are not adapted to avoid wire snares and are highly reluctant to abandon injured packmates. We have observed that repeated movement to a fixed location is common when a packmate is snared because the snared animal cannot move unless it breaks the snare wire (Video S1). Based on repeated movements to the same location over a week, in an area with many active snares at the time, we infer that one of the uncollared females was snared. On October 23rd, the collared female also died, probably in a snare (her collar had been cut off with a knife). With the death of the collared female, contact with the remaining sister was lost. If she survived, her location is unknown. Although competition with dominant competitors has adapted wild dogs to move efficiently through unfavorable landscapes, they have no adaptations to avoid wire snares. Snares are easily concealed on paths through vegetation, are silent, have little scent, and do not resemble anything dangerous in the natural environment. Consequently, wild dogs and other large carnivores do not avoid them, unless perhaps they have survived a prior snaring. Individual wild dogs have been snared as many as four times, so it is debatable whether even survivors learn to avoid snares (Kasozi et al., 2023). Snaring for bushmeat is now having strong, widespread effects on wild dogs and other large carnivores, by direct mortality, depleting prey populations, and altering carnivore community dynamics (Becker et al., 2024; Benhaiem et al., 2022; Creel et al., 2018; Creel et al., 2025; Feldmeier et al., 2025; Goodheart et al., 2024; Lindsey et al., 2013; Reyes de Merkle et al., 2024). When wild dogs leave well‐protected areas such as National Parks, they are exposed to an increased risk of being snared (Becker et al., 2024; Kendon et al., 2022; Watson et al., 2013). After leaving the females on September 19th, the Mayukuyuku males doubled back to the south, avoiding the Great North Road (which links the capital Lusaka, with a population of 2.2 million, and the major industrial cities of the Copperbelt near the border with the Democratic Republic of Congo) and moving precisely along the same route they had taken to the north (Video S2). Upon reaching Kabwe a second time they found their way around the city, to the east. They soon encountered two large reservoirs that are uncrossable for 20–30 km, but skirted both successfully and continued eastward until they reached the Chisomo and Petauke GMAs on November 9th. These GMAs are well‐connected to South Luangwa National Park (SLNP) and are part of the Luangwa Valley Ecosystem, which holds one of the largest and highest density wild dog populations on the continent (Creel et al., 2024; Reyes de Merkle et al., 2024). Dispersing wild dogs have repeatedly moved into this area from SLNP (Figure 1), and on November 14th, the males were within 14 km of a group of four females that had dispersed from the Kasweta Pack in SLNP (Video S2). The paths of the two groups passed within 1.6 km of one another in a period of 6 days, after the Kasweta females had moved 1538 km and the Mayukuyuku males 2356 km, from natal homeranges separated by 546 km in a straight line. Getting there is half the battle—immigration remains difficult and dangerous, and the groups did not remain together. The Kasweta females remained in the GMA (possibly with other males), but over the next few days the Mayukuyuku males moved 107 km to the southwest into an unprotected area. As of April 1st, 2026 the males had moved an astonishing 4126 km since dispersing and settling in a new range in the Luano GMA in the Luangwa Valley Ecosystem, at the top of the Great Rift Escarpment (Video S2). The stories of the Mayukuyuku males, Mayukuyuku females, Luamfwa females, and Kasweta females highlight an important distinction between movement and connectivity. Despite combined movements of 9863 km (for comparison, highway I‐90 crosses the entire United States in 4860 km) these four movements have yielded little genetic or demographic connectivity. Because wild dogs usually fight (often fatally) when they encounter unknown individuals, immigration itself is difficult, dangerous, and often unsuccessful (Creel & Creel, 2002). It is increasingly common to consider dispersal as a process of movement affected by environmental variables such as vegetation structure, roads, or human settlement (Hofmann et al., 2021). This perspective is valuable, but it remains important to remember that dispersal is fundamentally a social process. Dispersing wild dogs are likely to move through areas with favorable conditions, but they are seeking other wild dogs, and the social and demographic conditions that promote emigration and allow immigration are drivers of connection between populations, just as they are drivers of connection between packs (Behr et al., 2020; Cozzi et al., 2020; Creel & Creel, 2002). Nonetheless, the long‐distance movements between seven Zambian National Parks that we have observed (Figure 1) confirm results from genetic data (Creel et al., 2019; Tensen et al., 2022) that wild dogs remain capable of connecting distant ecosystems with extraordinary movements. Despite their invariably small population sizes and the high risk of mortality in human‐dominated landscapes (due to snaring, roadkill, and conflict due to livestock killing), wild dogs' enduring capacity to move long distances through unfavorable landscapes provides an extraordinary opportunity to conserve a functional meta‐population of an endangered large carnivore at the continental scale. Efforts to protect corridors that connect stronghold populations (Creel et al., 2019; Hofmann et al., 2021) and allow movements like those described here should be a high priority for wild dog conservation. Clearly, they can still be successful. With long distance movements also seen in Tanzania, Botswana, and Zimbabwe (Cozzi et al., 2020; Masenga et al., 2016; Sandoval‐Serés et al., 2022), it is plausible to hypothesize that most of the wild dog populations in Eastern and Southern Africa remain connected (Creel et al., 2019; Hofmann et al., 2021). An impressive 40% of Zambia has some form of protection for wildlife, and some important corridors are protected (notably along the Luangwa Valley), but others are not (Figure 1, Video S2). Significant investments are now being made within many protected areas via collaboration between the Zambian Government and NGO partners. If these efforts yield growing wildlife populations, it is likely that dispersal out of protected areas will increase, for wild dogs and other species (Creel et al., 2024; Reyes de Merkle et al., 2024). This creates the potential for even greater connectivity, but also for human–wildlife conflict, and policies that anticipate these changes will be of value. CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest. Supporting information Video S1. Video S2. Video S1_Metadata. Video S2_Metadata.