Turmoil first struck the Virunga Volcanoes region in October 1990 when the Rwandan Patriotic Front went to war along the Rwanda-Uganda border. This conflict culminated in the Rwandan genocide of April to July 1994. Then, between 1994 and 1996, the region struggled with millions of fleeing Rwandans who took refuge in camps inside Congo-Zaire. Next, the first civil war in Zaire struck the region, lasting from October 1996 to May 1997. Under the successor government, Zaire became the Democratic Republic of Congo. After a brief lull, a second civil war started in August 1998, striking the DRC and then drawing in both rebel forces and the armies of surrounding nations. In 2001, this war was still in progress. (For a detailed chronology of these conflicts, refer to Appendix 1).

Rwanda's guerrilla war was enormously harmful to biodiversity conservation and to protected habitats inside the Virunga Volcanoes region, but also had a serious impact on adjacent regions.

The first protected area overrun by the offensive launched by the Rwandan Patriotic Front (RPF) in October 1990 was Akagera National Park in eastern Rwanda. This offensive was rebuffed by a coalition composed of the Rwandan Armed Forces (RAF) and certain units of the Zairean army. Although that offensive did not touch the Virungas, the region was impacted by an immediate and sharp decrease in tourism. Expeditions to view the gorillas dropped off. Tourist revenue nearly collapsed. Conservation activities around the country lost most of their financial support. The agency responsible for protected areas, ORTPN (Rwandan Office for Tourism and National Parks), suddenly was faced with a significant fiscal deficit. At the same time, Rwandan authorities began orchestrating a new campaign of ethnic pogroms and arbitrary arrests, and ORTPN's staff was not exempt from this general climate of distrust and suspicion.

Beginning in January 1991, the date of the first RPF offensive in northwestern Rwanda and the first attack on Ruhengeri, the Virunga Volcanoes became a direct theater of military operations. Military leaders quickly understood the immense strategic value of the Virungas. The region is the sole wooded area between Rwanda and Uganda. As such, it offers armed units dense cover and secure escape routes. Soldiers discovered they could move freely through the region without attracting their enemies' notice. In 1991, eastern VNP and all of MGNP were affected by military operations. Very quickly, the RPF adopted the tactic of circling around the Sabyinyo volcano (a feature that borders all three countries), so that military operations next encroached into the Mikeno sector of Virunga National Park (ViNP) inside Zaire. Between 1991 and 1994, the Virungas experienced intensive movements for infiltration purposes (by RPF troops) or for patrols and searches (by RAF troops). The military presence in the vast forest grew and intensified. Both rebel and regular units increased in number and in activity. The RAF also established permanent positions in strategic locations such as the edge of the forest and the hollows located between the volcanoes of the eastern Virunga region (Sabyinyo, Gahinga, and Muhabura). Several hundred mines were laid in the forest region, mainly along the paths leading into the forest and along the Rwanda-Zaire border. After numerous forest ambushes, military personnel also cut the vegetation along several access trails leading to volcano passes in eastern VNP (Plumptre, Bizumuremyi, 1996). In the vicinity of Kinigi, rebel forces killed a number of ORTPN agents. The offices and lodgings of the park were pillaged and in some cases destroyed; Karisoke research station was heavily damaged. In addition, at least two gorillas were slain during this period, including Mrithi, the male silverback, who died in 1992 (Cooper, Cooper, 1996).

At the same time, mainly because of the drop in revenue from tourism, but also because of the military takeover of management operations, the authority of institutions responsible for protected areas crumbled. ORTPN, for instance, lost a good part of its control over the VNP to the military authorities during this period (Gombe, 1995).

Despite the enormous impact of the 1994 genocide on the Rwandan population, the environment as a whole and the protected areas in particular were not severely affected during the 100 days of massacres. During the final offensive of the RPF, which liberated Rwanda in July 1994, several thousand people took refuge in Zaire by following paths and roads through the Virunga region, bringing with them a good part of their livestock. . This sizeable exodus was the precursor of one of the most acute humanitarian crises of the century, and its environmental impact would fall heavily upon Virunga National Park.

The takeover of Rwanda by the Rwandan Patriotic Front in July 1994 following the genocide resulted in the most massive and most intensive population movement ever observed in recorded history. In July 1994, in just a few days, almost two million people left Rwanda and took refuge in neighboring countries, particularly Zaire. On July 15 alone, in a single day, 500,000 persons crossed the border and arrived in Goma, where they were joined over the next few days by 300,000 others. Goma, the desperate refugees hoped, could provide them with their most basic needs: water, firewood, and food. All these things were available in and around the southern part of the ViNP.

Three refugee camps formed during July 1994, springing up in the places where the refugees stopped: Kibumba, Mugunga and Katale. As more refugees continued flowing out of Rwanda, aid organizations developed two more campsites in late 1994 and early 1995, Lac Vert and Kahindo. At the end of 1994, the refugee population was estimated at approximately 720,000 persons (Delvingt, 1994). All these refugees, with their urgent human needs, were now poised on the edge of Virunga National Park.

It quickly became clear that moving the refugees to other sites would be impossible. The first stumbling block was the sheer number of displaced people. Logistically, it would have been extremely difficult to move nearly three-quarters of a million people. The second barrier to moving the refugees was that many of them were heavily armed ex-soldiers who resisted the idea of relocation. And third, there was no local civil authority over the camps that could have coordinated, directed and assisted humanitarian agencies. As a result, the refugee camps in the Goma region remained on the edge of the national park for more than two years. They only dispersed when the region was directly struck by a new outbreak of war.

The same problems that made the camps impossible to relocate also made them hard to manage. Additional problems were generated by the volcanic nature of the soil in this region, soil that also made it very difficult to provide such basic services as fresh water and solid and liquid waste disposal. Even digging pit latrines was a formidable task (UNHCR, 1998b). (Note that UNHCR developed its environmental guidelines after this crisis.)

The Rwandan refugee crisis in Zaire posed such an unprecedented threat to Virunga National Park that, in December 1994, the World Heritage Committee placed this protected area on its list of eight endangered natural world heritage sites (UNESCO, 1995).

Deforestation was among the most visible and best documented impacts during the refugee crisis. Aid agencies provided the refugees with shelter and raw food—but to cook their food, people had to fend for themselves. Wood gathering and cutting for firewood quickly became a huge threat to the environment. Trees were cut down for firewood, for construction materials, even for commercial purposes--large-scale charcoal manufacture, for instance, became a thriving business (Tombola, Sanders, 1994). At the beginning of the crisis, 40,000 people on average entered the park every day in search of wood (Tombola, Sanders, 1994). But the rate of deforestation increased even more. On some days, as many as 80,000 people penetrated the park, cutting approximately 1,000 tons of wood daily (Languy, 1995; Blondel, 1997). The refugee camps were present along the ViNP border for 27 months, and during that entire time woodcutting continued to intensify. That was particularly true in the Nyamulagira sector (the active volcano sector).

The impact of deforestation is illustrated by the following figures and data (Henquin, Blondel, 1997). Two years after the arrival of the refugees, 105 square kilometers of forestland had been impacted by deforestation. Thirty-five square kilometers of that were totally denuded. This total deforestation was equivalent to 63 square kilometers of clear cutting ("equivalent cleared area"). The table below provides details of deforestation by camp:

From a qualitative point of view, at least two-thirds of the deforestation occurred in lava-field forests—areas that are relatively poor in terms of biodiversity, compared for instance to the primary afro-montane forest in other parts of southern Virunga NP. Furthermore, at least 50 percent of the zones that were cleared or severely affected by the refugees were parts of young forests composed of pioneering species in the initial stages of recolonizing lava flows.

The most irreversible damage occurred in the Mikeno sector, within the zone of influence of the Kibumba camp, where extensive areas were deforested. Podocarpus milanjianus, in the mountain forest, was especially hard hit.

The following table shows the trend of deforestation around the different camps during the first and second years when refugees were present.

(hectares = equivalent cleared area, including the park buffer zone).

In the areas surrounding the Katale and Kahindo camps (containing 290,000 refugees in all), the level of deforestation was already relatively low during the first year, probably because when the camps were established, aid agencies also implemented park-protection programs.

In contrast, the wooded areas adjacent to the Kibumba camp suffered intensive deforestation during the first year. During Kibumba camp's second year, considerable means were mobilized to protect the area's important ecosystem. By 1996, the damage was practically halted.

In the areas surrounding the Mugunga and Lac Vert camps (holding approximately 200,000 refugees), the clearing of forestland was especially extensive. There were two reasons for this major deforestation. First, camp residents gathered large amounts of firewood for their own use. Second, camp residents also created a vast commercial enterprise selling wood and charcoal in the city of Goma (Languy, 1995). Because no security forces protected the forest, this commercial activity actually grew, and the rate of deforestation in the second year was higher than in the first.

A two-year study (Henquin, Blondel, 1997) was able to calculate, by extrapolation, the total deforestation within the Park caused by refugees during the two years and three months of the camps' existence:

Illegal bamboo cutting was mainly organized and perpetrated by refugees in the Kibumba camp (Bremer, 1996). The bamboo, Arundinaria alpina, could only be found in the higher-altitude Mikeno sector (part of the gorillas' habitat). Refugees used it for many purposes-to make baskets and mats, for instance, and to construct shelters. At one point, an international NGO even started a project that encouraged the refugees to make artisanal products—a project that abruptly ended when it was realized that the bamboo was coming, in fact, from the park! A total of 192 hectares of bamboo was exploited at a rate of 50 percent in the Mikeno (DRC) sector (meaning that the area affected was 50 percent cleared--i.e., one stem out of two was cut) (Henquin, Blondel, 1996).

During the two years that the refugees remained in the region, poaching intensified in the two southern sectors of ViNP (Biswas, Tortajada-Quiroz, 1996). The poachers preferred two antelope species, bushbucks and duikers, but also targeted forest buffaloes and elephants (Wathaut, 1996).

It is important to distinguish between two types of poaching. Traditional poaching in the region used rudimentary tools, such as wire or rope snares or spears, to capture game. This old technique was perpetrated in the region by local people and by some refugees. The new method of poaching, in contrast, was a result of the presence in the camps of numerous ex-soldiers who had smuggled their firearms out of Rwanda. Automatic weapons, that is, were easily available in the camps, and one of the purposes they were put to was poaching. The meat taken in this way was usually sold, not to other, impoverished refugees but to local populations who could afford to pay for it—particularly in such cities as Goma and Rutshuru. Many Zairean soldiers helped to intensify the toll of poaching on the ecosystem by creating large-scale enterprises that organized all phases of commerce, from the slaying of the animal to the marketing of its carcass (Mushenzi Lusenge, 1995).

Various armed groups who continued to move through Volcanoes National Park also poached wildlife, and the unstable situation caused by such troops' presence also promoted poaching. The toll of bushbucks, duikers, and buffaloes increased, particularly between 1991 and 1996 (Plumptre et al. 1997).

In July and August 1995, poachers killed four mountain gorillas--three silverback males and one adult female (Cooper, Cooper, 1996)--the first cases of gorilla killings in the Virunga region in 10 years (Weber, 1989). Although there was probably no direct connection between the refugees and the poachers who committed these acts, it is nonetheless very likely that this type of poaching was the result of the general state of disorder and lack of security stemming from the crisis (GTZ, 2000).

Among the hundreds of thousands of refugees who fled Rwanda in July 1994, several thousand reached Zaire by crossing the Virunga Volcanoes. Some even lived in the forest for several weeks before exiting on the Congolese side. Most made the journey with their livestock--cows, goats and sheep. This type of activity understandably disturbed the forest ecosystem; it also increased the risk of transmitting disease to wildlife.

During the first year the camps operated, and in particular during the first months, many refugees needed emergency medical treatment, generating significant amounts of medical waste. A number of organizations working in the medical sector got rid of this waste matter simply by dumping it in the park. This contaminated matter included used syringes, human waste, and materials containing human blood (Biswas et al. 1994). By the second year, such dumping practices were largely halted.

Security in the Congolese region significantly deteriorated during the refugee-camp period because of the presence of tens of thousands of former Rwandan army soldiers still bearing modern weapons—a situation that greatly reduced the effectiveness of ICCN ranger patrols. At times, armed groups of ex-soldiers even confronted and threatened ICCN field agents. The ICCN lost all control of two zones in southern ViNP, one in the vicinity of the Mugunga-Lac Vert camps in the Nyamulagira sector, and the other at the heart of the Mikeno sector (a zone located between the Mikeno, Karisimbi and Visoke volcanoes).

The intensive utilization of the forest by human populations and by domestic animals for more than two years posed a serious risk to wildlife in southern ViNP.

As members of the anthropoid family, gorillas are among our closest relatives, genetically speaking, and are susceptible to a large number of pathogens of human origin. Maladies that may be transmitted from people to gorillas include a number of respiratory diseases (for example, measles, tuberculosis, pneumonia, and influenza), as well as diseases transmitted by fecal-oral contact (including shigellosis, hepatitis, herpes, scabies, intestinal worms, and polio) (Homsy, 1999). Because full access to the entire Virunga forest has not yet been guaranteed at this writing, it remains impossible to accurately assess the impact of the vast and complex movements of human population through the forest on the health status of the mountain gorilla. But it is nonetheless highly probable that the risk of gorillas falling ill with human pathogens increased dramatically.

The thousands of cows, goats and sheep that passed through the southern ViNP similarly posed a significant health threat to the region's wild ungulates—buffaloes and antelopes. Foot-and-mouth disease and bovine tuberculosis are among the illnesses that can be transmitted from domestic cattle to wild ungulates.

In the most normal of times, people commit crimes. When the general level of security deteriorated throughout the region, illegal activities were exacerbated. Local people at times took advantage of the new opportunities presented by the general disorder. Two examples that have already been mentioned are increased poaching and charcoal production (Werikhe et al. 1997).

Both the presence of enormous numbers of refugees and the region's unrest seriously lowered the volume of tourism during the entire crisis. But tourist visits to observe gorillas never really stopped. Modest numbers of tourists came to watch the gorillas at Jomba, a ranger outpost near the Ugandan border inside Zaire, later DRC, and an excellent point of entry for tourists into ViNP.

Most deforestation that occurred during the refugee crisis took place within the park (Henquin and Blondel, 1997). But tree plantations outside the park were also harmed. In a region that was already short of firewood before the crisis, the presence of refugee camps not only worsened the immediate situation but also, in the long term, threatened the local populations' ability to meet their energy needs after the crisis.

The two civil wars that struck DRC had a series of direct impacts on the Virunga Volcanoes. 2 [Parallel to the events observed in and around the Virunga Volcanoes, it is also important to mention the attack by the Interahamwe rebels in the Bwindi Impenetrable National Park in Uganda in March 1999. The rebels killed an Ugandan warden and eight foreign tourists, systematically pillaged and looted the park buildings and those of tourism agencies and dealt a very serious blow to tourist activities throughout Uganda (Lanjouw, 1999).]

ICCN personnel were greatly affected by these wars. Many agents were killed or injured by armed groups from all sides (Mushenzi Lusenge, 1996). The facilities of southern ViNP (Rumangabo station and patrol posts in the Mikeno sector) were seriously damaged (Werikhe et al. 1997)—perhaps by local people searching for building materials, perhaps by the armed gangs who circulated freely through the region and inflicted systematic destruction.

Several gorillas were killed, either by armed groups or by poachers (Cooper, Cooper, 1996). Considering the gravity of the various crises gripping the region, it is almost miraculous that there were not more gorilla losses during this period. Gorillas may have been spared because the local populations do not usually eat gorilla meat, and also because gorillas were often viewed as a national asset thanks to their known economic value for tourism. As a result, both the Rwandan authorities and various rebel groups vowed not to harm the gorillas.

Armed groups who entered the region committed such illegal activities as woodcutting and poaching. All the various military forces were involved: rebels, government troops, even allied forces arriving from other countries.

Military units--as well as several thousand refugees--also engaged in farming activities inside the forest, both in the Mikeno sector of ViNP and in VNP, growing such exotic plants as potatoes, tobacco, wheat and hemp (Rutagarama, 1999).

Deforestation for strategic purposes was also carried out. Such strategic deforestation had its greatest impact along the Mwaro ecological corridor connecting the Mikeno and Nyamulagira sectors of ViNP in DRC. A number of animal populations, including elephants, formerly used the corridor to migrate between the dormant volcanoes and the rest of Virunga National Park inside DRC. That changed when military forces controlling the region tore up vegetation alongside the Goma-Rutshuru road in order to reduce the threat of ambush. This slashed corridor, initially 20 meters wide, was quickly widened (up to 50-70 meters in places) by local people, sometimes with the complicity of military officers, severing the very important ecological link that elephant and other animal populations used in their seasonal movements.

The blatant lack of security throughout the region had enormous consequences. Armed rebel groups freely used certain parts of the ViNP forest in DRC—especially the areas around Mikeno volcano and the patrol posts at Gatovu, Kibumba, Bukima and Bikenge. VNP was not secured from such incursions until July 1999, the date of its reopening.

Another consequence of this state of near-anarchy was that the institutions responsible for protected areas were enormously enfeebled. That was particularly true for ICCN in DRC and Rwanda's ORTPN. Two factors--the drastic fall in tourism income and the military takeover of park surveillance--especially acted together to weaken both their official mandates and their operating capacities.

The presence of armed groups also offers terrible potential for the transmission of human diseases to wildlife--especially to the great apes. One enormous risk was posed by the many latrines discovered in the VNP following the period of insurrection that prevailed in northwestern Rwanda from 1997 to 1998 (Rutagarama, 1999)