HORMONES, MEMORY AND THE HIPPOCAMPUS

It is well known that corticosterone has an effect on memory function (Sapolsky 1996; McEwen & Sapolsky 1995; McEwen 2000). It has generally been assumed that short-term elevations in corticosterone enhance spatial memory whereas prolonged, chronic elevation has extremely deleterious effects. Most of the research addressing the relationship between corticosterone elevation and memory has been focused on mammals (McEwen & Sapolsky 1995; McEwen 2000) but there are also studies suggesting that corticosterone has an effect on memory in birds (Loscertales et al. 1997; Sandi et al. 1995; Sandi & Rose 1997).

Unlike short-term, chronic corticosterone elevation has been reported to result in memory impairments (Sapolsky 1996; McEwen & Sapolsky 1995; McEwen 2000).  It appears, however, that baseline corticosterone levels may remain elevated in wintering parids over several months (Silverin 1998 – willow tits). Experimental manipulations of food supply designed to decrease food predictability and/or availability also triggered prolonged elevation of baseline corticosterone in mountain chickadees (Pravosudov et al. 2001) and in red knots (Calidris canutus; Reneerkens et al. 2002). It is important to note that such elevations are only moderate and they exceed the baseline levels only two or three times compared to stress-induced levels which might exceed baseline concentration by more than 10 times (Silverin 1998; Wingfield et al. 1997, 1998).

            It appears that prolonged corticosterone elevations occur when birds need to be feeding most efficiently in order to survive and when memory for previously made food caches could also be most critical for survival (Pravosudov & Clayton 2001). Foraging conditions which caused long-term moderate corticosterone elevation also resulted in enhanced spatial memory and more efficient cache retrieval (Pravosudov & Clayton 2001). Thus, prolonged moderate elevation of corticosterone correlated with enhanced spatial memory in mountain chickadees but this was only a correlational link suggesting that corticosterone could have been responsible for discovered changes in spatial memory.

            When implanted with corticosterone implants designed to maintain moderate corticosterone elevation (about 150% over the baseline) for several weeks, mountain chickadees demonstrated better spatial memory performance compared to placebo-implanted individuals (Pravosudov 2003) suggesting a direct link between elevated corticosterone and spatial memory. Furthermore, compared to placebo-implanted birds, corticosterone-implanted chickadees fed more intensively and cached significantly more food (Pravosudov 2003). Thus, long-term moderately elevated corticosterone facilitates more foraging and food caching while enhancing spatial memory required for successful cache retrieval. This entire “suite” of behaviors is likely to increase birds’ probability of survival when environmental conditions are energetically demanding such as during the winter. These results suggest that prolonged but moderately elevated corticosterone levels may be highly adaptive by enhancing memory rather than impairing it as it has been widely suggested (Sapolsky 1996; McEwen & Sapolsky 1995; McEwen 2000). The fact that birds do not always maintain elevated corticosterone levels despite its apparent benefits suggests that even moderate corticosterone elevations carry costs (e.g. weakening immune function, Wingfield et al. 1997, 1998). These costs, however, do not seem to concern memory function, at least in a short term.

            It has been suggested that chronically elevated corticosterone might impair memory through increased neuronal death but most experimental studies used high doses of corticosterone resulting in clinically high elevations (Sapolsky 1996; McEwen & Sapolsky 1995; McEwen 2000). Such elevations might not necessarily occur under naturally stressful conditions and it is important to understand the consequences of naturally occurring corticosterone elevations.

Limited and unpredictable food supply triggered moderate corticosterone elevation and resulted in enhanced spatial memory performance in mountain chickadees (Pravosudov & Clayton 2001; Pravosudov et al. 2001). There were no significant differences in hippocampal volume or the total number of neurons between the birds maintained on ad libitum food (lower baseline corticosterone) and on limited and unpredictable food (elevated corticosterone; Pravosudov et al. 2002). Similarly, there were no significant differences between corticosterone- and placebo-implanted mountain chickadees in either hippocampal volume, total number of hippocampal neurons or cell proliferation rates in the ventricular zone adjacent to the hippocampus and to the mesopallium (Pravosudov & Omanska 2005). Thus relatively long-term (49 days with corticosterone implants and 94 days on limited and unpredictable food supply) moderate corticosterone elevation had no detectable effect on hippocampal structure and cell proliferation rates. It has been reported that prolonged stress and associated prolonged corticosterone elevation result in reduced neurogenesis rates (Gould & Tanapat 1999). Neurogenesis consists of cell proliferation and neuron survival (Prickaerts et al. 2004) and thus it is still possible that even moderately elevated corticosterone levels can negatively impact neuron survival rates.

Experimental results with mountain chickadees suggest that corticosterone elevation in response to challenging environmental conditions is adaptive, at least in wintering food-caching birds. When foraging conditions deteriorate, corticosterone levels rise facilitating more active foraging behavior and more food caching while enhancing spatial memory which is important for successful cache retrieval. When environmental conditions improve, corticosterone levels return to normal suggesting that corticosterone elevation has some costs, which might be outweighed by the benefits during energetically demanding periods.