We applaud Coall & Hertwig (C&H) for highlighting the central role of evolutionary principles in understanding the important and understudied phenomenon of grandparental investment. But we can take their approach further. A proper fitness-based analysis of kin-directed behavior reveals considerable conflict and a richer, more complex picture of the selective forces shaping patterns of investment.
Although equally related to all of their grandchildren, grandparents are not necessarily selected to distribute their resources evenly. The optimal allocation pattern depends on the shape of the curve describing fitness returns on their investment (Fig. 1). If fitness returns are linear (curve A), all distribution patterns yield the same total return; it does not matter whether they spread their investment evenly or favoritize particular grandchildren. If returns are decelerating (B), it always pays to invest in the grandchild with fewest resources, which will tend to promote an even allocation pattern (though see below). If returns are sigmoidal (C), intermediate levels of investment are favored; so grandparents with limited resources might be better off directing them all to one grandchild rather than trying to spread them evenly. Other curves are possible and may favor alternative patterns of investment.
Figure 1. Three hypothetical relationships between the total amount of resources invested in a grandchild and the expected fitness (lifetime reproductive success) of that grandchild. A (solid line): linear returns on investment; B (dashed line): decelerating returns on investment; C (dotted line): sigmoidal returns on investment. The points x and y refer to two different grandchildren who have previously received different levels of investment from their parents (see text for details).
This situation is complicated further by the fact that grandchildren may receive different levels of investment from their own parents (i.e., the grandparents' offspring and sons-/daughters-in-law), putting them at different positions on the returns-on-investment curve. As an example, consider two grandchildren, x and y (shown in Figure 1), who are first cousins (they share grandparents, but not parents). Prior to any grandparental investment, these individuals have received different amounts of investment from their own parents. Grandchild y is in a relatively privileged position, in that its parents have been able to invest more resources in it than the parents of grandchild x. How should the grandparents optimally invest in these grandchildren? The answer depends on the shape of the returns-on-investment curve. If returns are decelerating, grandparents should invest preferentially in grandchild x; if returns are sigmoidal, they should favor grandchild y; if returns are linear, it does not matter which grandchild they invest in.
Whatever pattern of investment is favored from the point of view of the grandparents, the beneficiaries are unlikely to agree with this. The evolutionary basis for intergenerational conflict over parental investment is well established (Trivers Reference Trivers1974), and the same principles will lead to a similar intergenerational conflict over grandparental investment. Put simply, grandchildren are more related to themselves (coefficient of relatedness, r=1) than to their siblings (r=0.5) and cousins (r=0.125), and so will seek a disproportionate share of grandparental resources. Their parents will partly support this bias, since they are more related to their own children (r=0.5) than to their nieces and nephews (r=0.25). Thus, as recognized by C&H, grandparental investment is not a one-way street; but this is the case even when the investment is not reciprocated. More directly, descendants may be looking to exploit their grandparents' willingness to invest.
How will this conflict manifest itself? The relatedness asymmetry will favor behaviors in the grandchildren and their parents that increase the chances of resources being allocated disproportionately to their own direct line of inheritance. This may entail overt manipulation – for example, direct requests for resources – but also more subtle tactics such as maintaining regular contact with the grandparents. Thus, while grandparents could conceivably invest resources in their grandchildren as a way of encouraging support from those grandchildren or their parents (Friedman et al. Reference Friedman, Hechter and Kreager2008; Laferrère & Wolff Reference Laferrère, Wolff, Kolm and Ythier2006), a contrasting view is that grandchildren and their parents actively seek contact with the grandparents and provide support as a means of encouraging greater resource investment.
Of course, grandparents are not passive partners in this process, and will be selected to resist any attempts to divert resources that conflict with their own fitness interests. Typically, grandparents will favor a more equal allocation of resources than their grandchildren, but depending on the returns-on-investment curves and the amount of resources contributed by the parents (Fig. 1), this may not always be the case. Where it pays grandparents to favoritize particular grandchildren, the conflict with those grandchildren will be weaker, while that with the others will be intensified.
Trivers (Reference Trivers1974) recognized that intergenerational conflict over care might also impinge on mate choice, and our own work (Buunk et al. Reference Buunk, Park and Dubbs2008; van den Berg et al. Reference van den Berg, Fawcett, Weissing and Buunk2009) supports this. As explained above, grandparental investment decisions should be sensitive to inequality in the resources their grandchildren receive from other sources. A major determinant of this resource distribution is the care provided by the sons- and daughters-in-law of the grandparents, that is, the parents who marry into the descendant line. If these in-laws differ markedly in their contributions to parental care, grandparents may adjust their own levels of investment accordingly. But they might also be able to manipulate the resource distribution in their favor at an earlier stage, by actively influencing their offspring's choice of spouse in the first place. Arranged marriages and other forms of control over offspring's mate choice are common across cultures and throughout history (Apostolou Reference Apostolou2007) and reveal considerable disagreement over the characteristics of a suitable partner (Buunk et al. Reference Buunk, Park and Dubbs2008; in press). Our evolutionary analysis (van den Berg et al. Reference van den Berg, Fawcett, Weissing and Buunk2009) confirms that such disagreement can arise from the intergenerational conflict over grandparental investment outlined earlier. Typically, individuals will value caring qualities more highly in their offspring's mates (i.e., their sons-/daughters-in-law) than in their own mate.
In a general sense, we fully support C&H's first steps towards a theory of grandparental investment, in which evolutionary principles play a prominent role. But for a deeper understanding of investment patterns, we should explicitly consider the details of the selective forces involved. Grandparental investment is not solely in the hands of the grandparents; it is subject to an intergenerational evolutionary conflict in which the fitness interests of three generations are at stake, with each party selected to manipulate the pattern of investment for its own benefit.
We applaud Coall & Hertwig (C&H) for highlighting the central role of evolutionary principles in understanding the important and understudied phenomenon of grandparental investment. But we can take their approach further. A proper fitness-based analysis of kin-directed behavior reveals considerable conflict and a richer, more complex picture of the selective forces shaping patterns of investment.
Although equally related to all of their grandchildren, grandparents are not necessarily selected to distribute their resources evenly. The optimal allocation pattern depends on the shape of the curve describing fitness returns on their investment (Fig. 1). If fitness returns are linear (curve A), all distribution patterns yield the same total return; it does not matter whether they spread their investment evenly or favoritize particular grandchildren. If returns are decelerating (B), it always pays to invest in the grandchild with fewest resources, which will tend to promote an even allocation pattern (though see below). If returns are sigmoidal (C), intermediate levels of investment are favored; so grandparents with limited resources might be better off directing them all to one grandchild rather than trying to spread them evenly. Other curves are possible and may favor alternative patterns of investment.
Figure 1. Three hypothetical relationships between the total amount of resources invested in a grandchild and the expected fitness (lifetime reproductive success) of that grandchild. A (solid line): linear returns on investment; B (dashed line): decelerating returns on investment; C (dotted line): sigmoidal returns on investment. The points x and y refer to two different grandchildren who have previously received different levels of investment from their parents (see text for details).
This situation is complicated further by the fact that grandchildren may receive different levels of investment from their own parents (i.e., the grandparents' offspring and sons-/daughters-in-law), putting them at different positions on the returns-on-investment curve. As an example, consider two grandchildren, x and y (shown in Figure 1), who are first cousins (they share grandparents, but not parents). Prior to any grandparental investment, these individuals have received different amounts of investment from their own parents. Grandchild y is in a relatively privileged position, in that its parents have been able to invest more resources in it than the parents of grandchild x. How should the grandparents optimally invest in these grandchildren? The answer depends on the shape of the returns-on-investment curve. If returns are decelerating, grandparents should invest preferentially in grandchild x; if returns are sigmoidal, they should favor grandchild y; if returns are linear, it does not matter which grandchild they invest in.
Whatever pattern of investment is favored from the point of view of the grandparents, the beneficiaries are unlikely to agree with this. The evolutionary basis for intergenerational conflict over parental investment is well established (Trivers Reference Trivers1974), and the same principles will lead to a similar intergenerational conflict over grandparental investment. Put simply, grandchildren are more related to themselves (coefficient of relatedness, r=1) than to their siblings (r=0.5) and cousins (r=0.125), and so will seek a disproportionate share of grandparental resources. Their parents will partly support this bias, since they are more related to their own children (r=0.5) than to their nieces and nephews (r=0.25). Thus, as recognized by C&H, grandparental investment is not a one-way street; but this is the case even when the investment is not reciprocated. More directly, descendants may be looking to exploit their grandparents' willingness to invest.
How will this conflict manifest itself? The relatedness asymmetry will favor behaviors in the grandchildren and their parents that increase the chances of resources being allocated disproportionately to their own direct line of inheritance. This may entail overt manipulation – for example, direct requests for resources – but also more subtle tactics such as maintaining regular contact with the grandparents. Thus, while grandparents could conceivably invest resources in their grandchildren as a way of encouraging support from those grandchildren or their parents (Friedman et al. Reference Friedman, Hechter and Kreager2008; Laferrère & Wolff Reference Laferrère, Wolff, Kolm and Ythier2006), a contrasting view is that grandchildren and their parents actively seek contact with the grandparents and provide support as a means of encouraging greater resource investment.
Of course, grandparents are not passive partners in this process, and will be selected to resist any attempts to divert resources that conflict with their own fitness interests. Typically, grandparents will favor a more equal allocation of resources than their grandchildren, but depending on the returns-on-investment curves and the amount of resources contributed by the parents (Fig. 1), this may not always be the case. Where it pays grandparents to favoritize particular grandchildren, the conflict with those grandchildren will be weaker, while that with the others will be intensified.
Trivers (Reference Trivers1974) recognized that intergenerational conflict over care might also impinge on mate choice, and our own work (Buunk et al. Reference Buunk, Park and Dubbs2008; van den Berg et al. Reference van den Berg, Fawcett, Weissing and Buunk2009) supports this. As explained above, grandparental investment decisions should be sensitive to inequality in the resources their grandchildren receive from other sources. A major determinant of this resource distribution is the care provided by the sons- and daughters-in-law of the grandparents, that is, the parents who marry into the descendant line. If these in-laws differ markedly in their contributions to parental care, grandparents may adjust their own levels of investment accordingly. But they might also be able to manipulate the resource distribution in their favor at an earlier stage, by actively influencing their offspring's choice of spouse in the first place. Arranged marriages and other forms of control over offspring's mate choice are common across cultures and throughout history (Apostolou Reference Apostolou2007) and reveal considerable disagreement over the characteristics of a suitable partner (Buunk et al. Reference Buunk, Park and Dubbs2008; in press). Our evolutionary analysis (van den Berg et al. Reference van den Berg, Fawcett, Weissing and Buunk2009) confirms that such disagreement can arise from the intergenerational conflict over grandparental investment outlined earlier. Typically, individuals will value caring qualities more highly in their offspring's mates (i.e., their sons-/daughters-in-law) than in their own mate.
In a general sense, we fully support C&H's first steps towards a theory of grandparental investment, in which evolutionary principles play a prominent role. But for a deeper understanding of investment patterns, we should explicitly consider the details of the selective forces involved. Grandparental investment is not solely in the hands of the grandparents; it is subject to an intergenerational evolutionary conflict in which the fitness interests of three generations are at stake, with each party selected to manipulate the pattern of investment for its own benefit.