The basis and bane of vegetation sampling is the random numbers table. This clever invention assures that within the mathematically selected primary sample area, the researcher dose not somehow bring biased elements into the study. So within each quad, we systematically laid out five straight transect lines with string every thirty metered. Own far we walked along that string and how far we turned either left or right to find our nested 100-10, and 1- square meter sample plots, however were subject to the dictates of the random numbers table. And it was a dictator. No sooner had we returned from a precarious descent into a ravine than the next magic number would require that we go back down over the same cliff. This perverse sequence also seemed to guarantee return passes through especially nasty patches of elephant’s nettles. At times like these, only Amy’s strong commitment to good science over whelmed Bills strong desire to sample some place more accommodating.

We completed vegetation sampling in mid June. The heavy rainy season was ending with a vengeance, we had experienced several days in a row of powerful afternoon storms. On our last day, we were determined to start early and finish by noon if at all possible. We were on schedule when a rare flash of lighting signaled the start of an hour long blitz of hail. The marble sized ice balls stung our exposed hands and heads as they shredded much of the surrounding vegetation. By the time we finish our last samples we were standing in ice water in a daze, but were still aware enough to realize that we were both in the early stages of hypothermia. Our solution to the problem to pack our bags and race back up the mountains to camp made sense given our poor mental state and excellent physical condition. Our half hour descent turned into a ten minute return sprint uphill that warmed our bodies leaving us almost delirious with laughter as we pulled up about fifty yards below our cabin. There we held hands and continued walking until a strange, deep ripping sound brought us to a halt. The sound turned into a ror as a giant Hagenia tore its roots from the earth and lurched toward our moss that crashed through our roof, cutting our home almost in half. Had we arrived five minutes earlier, we would have been peeling off our wet clothes exactly where the colossus came to rest by the side of our bed. Instead we spent the remainder of the day cutting up the tree, replacing a few key pieces of the wooden frame work, and patching new panels of corrugate tin in place. Among the few advantages of living in a tin shack is its ease of repair.

OUR SAMPLING REVEALED one disturbing fact about the virunga ecosystem. In more than 1,500 plots, we found not one Hagenia sapling. Mature Hagenia were quiet common from the park boundary at 8,800 feet up to the treeline at 10,600 feet. But there was little sign of any re-generation of this virunga giant: the largest life form and one of only two major species in the entire park. Gorillas didn’t eat any part of the tree, but they did scale its sloping trunk to feed on its abundant lichens and ferns. Birds, squirrels, hyrax and many other creatures certainly depended on the tree for food and shelter. Outside of our sample areas we did see some signs of regenerations, but always as shoots from the fallen trunks of dead trees. We also knew of one patch of young saplings in a disturbed area near the park boundary. Our finding was not definitive, but it did raise serious questions about karisoke’s continuing use of Hagenia as its primary source of firewood-a practice that ended a few years later when sandy Harcourt became director of the station.

AMY’S SEVENTEEN MONTHS of research revealed that the Virunga gorillas made very selective choices from among more than one hundred distinct food types. They preferred more nutritious and higher quality foods whenever possible and actively sought diversity in their diet. Bamboo shoots made up the majority of their diet during one third of the year. Furthermore, many of their foods were relatively wide spread and abundant, with the table exception of bamboo. Some areas with high food values appeared in our sampling yet were rarely visited by group 5.Apparently; they didn’t need to visit them. The gorillas lived in a giant salad bowl that still offered an ample supply of diverse nutritious foods.

This conclusion was contrary to our expectations and to widely held perceptions. Citing their low numbers and loss of habitat, a growing number of writers and conservationists were already consigning the mountain gorillas to the dustbin of evolutionary history, or to a constrained life of captive breeding. Fossey herself spoke openly of their imminent extinction. Ian Redmond had told us that he believed it was our task “to record all that we could about the gorillas lives and behavior before they disappear “ Ian and Dian were as committed to the gorillas as anyone could possibly be, but we believed that our overriding challenge was to save the mountain gorilla, not document its demise. Our research was demonstrating that the gorillas had numbers and perhaps grows. In a later demographic analysis we would show that the population even had the reproductive potential to return to Schaller’s level of four hundred to dive hundred gorillas. We were determined to use this new information to help make appositive case for mountain gorilla conservation in the wild.

Yet many problems persisted in the salad bowl paradise. Poaching continued thought our tenure at Karisoke. The long rainy season of 1979 was punctuated by the constant coughing and sniffling of gorillas that no longer had a lower elevation refuge from periods of extreme rain and cold. And the rising tide of human population lapped steadily higher on their forested shore. Without a strong dike of political support, the gorillas would be swamped.

PUCK AND TUCK WERE BROTHERS. But that was before Puck gave birth. From the time Dian Fossey began her work at Karisoke in 1969, Group 4 was the primary focus for research. Other nearby groups was monitored irregularly to track their movements, membership and interactions. But only in the mid 1970s did any serious research begin on GROUP 5, when Sandy Harcourt began working at Karisoke and used data from both Groups 4 and 5 for his behavioral research. When Sandy left Karisoke to complete his doctorate at Cambridge University, Group 5 was again left to infrequent monitoring until our arrival in 1978.At that time, Dian gave us a family composition list for group 5, but there was some confusion over certain of the twelve listed individuals. Amy quickly determined that there were actually fourteen in the family, two silver back males, Beethoven and icarus, four adult females-Effie, Marchessa, Pantsy and Liza; two black back (eight-to ten-year-old) males –ziz and Puck; three younger males –Tuck, Pablo and Shinda and three younger females- quince, poppy and Muraha.

This structure remained quite stable, except for quince’s death and Liza’s transfer, throughout our time at Karisoke until November 14, 1978, when things changed dramatically. Amy had taken two days to catch up on transcribing her field notes and was sitting at our desk when David watts knocked and walked in.