Experimental tests of minimum viable population size

Fitness and rates of extinction were compared among populations of the housefly, Musca domestica L., kept either at constant effective sizes of 50, 500 or 1500 or passed through extreme founder events reducing effective size to 5. Populations were maintained for 24 generations, which for small to medium-sized mammals would be less than the 200 years suggested by Soulé et al. (1986) as necessary for maintaining viable populations of endangered species. The results demonstrate that effective population sizes have to be greater than the 50 individuals suggested by Franklin (1980) to retain fitness and escape extinction, even in the short term. In contrast to populations of constant size that exhibited monotonic decreases in fitness through time, populations established with few founders rebounded from initial inbreeding depression. However, they were less adaptable to environmental stress than constant size populations, suggesting that populations founded with few numbers may do well within a single environment but may do far less well if they are reintroduced to natural environments or exposed to rapid environmental changes. All correspondence to E. H. Bryant. Tel:713-743-2651; Fax: 713743-2636; E-mail:[email protected]. accumulate within small populations to cause loss of fitness and eventual extinction (Lynch & Gabriel, 1990; Charlesworth, Morgan & Charlesworth, 1993; Lande, 1994; Lynch, Conery & Burger, 1995a,b). While shortterm population decline is expected to be minimal for populations with effective sizes greater than 100 or so (Lande, 1994), many endangered species are kept under optimal conditions that minimize selection. As a result, accumulated effects of mutations can be rapidly accelerated, leading to losses in competitive ability of 1–2% per generation (Shabalina, Yampolsky & Kondrashov, 1997; Bryant & Reed, 1999). Any detrimental effects might not be evident within a benign captive environment, but these captive populations may not prosper in more natural regimes (Kondrashov & Houle, 1994; Shabalina et al., 1997; Bryant & Reed 1999). It is also difficult to predict population extinction from current population fitness. It is widely accepted that inbreeding depression in small populations, for example, decreases fitness and increases the risk of extinction (Soulé, 1980; Ralls et al., 1988; Miller & Hedrick, 1993; Thornhill, 1993; Frankham, 1995; Falconer & Mackay, 1996). Indeed, the rate of extinction within small populations appears to be greater than large populations, for both experimental and natural populations (Pimm, Jones & Diamond, 1988; Stacy & Taper, 1992; Latter, Mulley et al., 1995; Frankham, 1996; Newman & Pilson, 1997). Nevertheless, if extinction bears a non-linear relationship to level of inbreeding, as suggested by Frankham (1995), it may be particularly difficult to predict, based on current levels of inbreeding and/or genetic variation, which populations are under risk of extinction. Moreover, it is often difficult to separate genetic from environmental causes of extinction (Pimm et al., 1988; Schoener & Spiller, 1992), so there are few documented cases where inbreeding depression has directly led to extinction in wild populations (e.g. Saccheri et al., 1998). This has led some authors to question any important link between genetic variation, inbreeding depression and population extinction (Harcourt, 1991; Young, 1991; Shields, 1993; Caughley, 1994). Most of the work on establishing minimum viable population size has been theoretical, and there is a need for additional empirical evidence to parallel these theoretical predictions and establish clearer relationships between population size, fitness and extinction. While empirical tests of long-term effects are beyond the scope of most experiments, the relationship of population size to short-term population health is directly amenable to experimental verification. The purpose of this study was to develop an empirical assessment of fitness and extinction in relation to population size in experimental populations of the housefly, Musca domestica L. Specifically, we compared life histories among populations derived from a large outbred population of this species and kept at constant effective population sizes of 50, 500 or 1500, respectively, for a total of 24 generations. In addition, some managed populations of endangered species have been initiated with few founders (Foose, 1990; Hedrick & Miller, 1992). To investigate the effect of small founder size separately from maintenance size, we also included experimental populations that were founded with few individuals and allowed to recover to large size. Finally, we compared the fitness of experimental populations in benign as well as stressful environments, to extrapolate our results to more natural conditions. For mammals weighing ≤ 330 kg, 24 generations translates into ≤ 200 years (Millar & Zammuto, 1986), and thus our experiment directly addresses the minimum of 200 years suggested by Soulé et al. (1986) necessary for the management of endangered species.