When the red deer project begun in 1971 the study area was routinely culled. The last cull within the North Block occurred in 1973, after which the population increased to an approximately constant size by the early 1980s. Since then the total number of deer in the study area has fluctuated (see figure 1), although the sex ratio has continued to become more female skewed. The fluctuations in size are relatively small, and are comparatively tiny compared to those seen in the Soay sheep (http://soaysheep.biology.ed.ac.uk/population-ecology).
Figure 1: Changes in the number of adult females and males in the population. 
The first research on understanding the population dynamics focused on factors influencing survival rates. Early studies identified how age, sex, density and maternal characteristics influenced survival rates (Clutton-Brock et al. 1987a, b), with later work examining the role of spatial variation, weather and genetics (Coulson et al. 1997, 1998, Catchpole et al. JABES). These survival analyses were later coupled to analyses of fecundity rates (Coulson et al. 2000) to allow the construction of demographic models of the population dynamics of red deer (Milner-Gulland et al. 2000, 2004; Clutton-Brock et al. 2002). These models revealed that the change in population sex ratio observed was primarily generated by an increase in density impacting male mortality and dispersal rates to a greater extent than female rates.
The primary reason for constructing population models of the red deer of Rum was to allow a set of applied questions to be asked: what would be the likely consequences on the Scottish deer population of a wolf reintroduction (Nilsen et al. 2007), and what is the optimal deer management strategy to maximize income for a Scottish estate (Clutton-Brock et al. 2002; Milner-Gulland et al. 2002)? Both these pieces of work have fed into government policy in Scotland.
The relatively stable dynamics of this population have inspired work on ‘transient population dynamics’ (Coulson et al. 2004). In order to understand this work it is necessary to consider a little more why fluctuations in population size are comparatively small in red deer. The fluctuations are small because density-dependence stabilizes population numbers: births almost perfectly balance deaths and both fertility and mortality rates respond in a similar way to any change in population numbers. This is because an increase in density decreases survival and fertility rates gradually, and equivalently. Although density-dependence in reproduction and survival are similar, not all rates responded identically to the increase in density observed in the 1970s. The first part of the life history to respond to an increase in density was the age at first reproduction, followed in turn by the fertility rate of prime aged adults, juvenile survival rates and finally adult survival (Coulson et al. 2004). This pattern in the way different rates responded to an increase in density is primarily a result of effects early in life taking a generation to play out. For example, we only observed density-dependence in adult survival once individuals born into a high-density population had themselves survived to reach adulthood. These cohort effects (Albon et al. 1987) generated a long period of transience in the population dynamics, lasting a couple of decades (Coulson et al. 2004).
The current focus on the population dynamics of red deer is on identifying climatic drivers associated with aspects of demography and deer life history, and on linking together the on-going evolutionary work with the detailed and comprehensive demographic data our dedicated field assistants have collected over the past four decades.