**2. Hypothesis**

Understanding the causes and dynamics of naturally occurring range limits in plants has become a central issue in both basic (evolutionary ecology) and applied (conservation and agriculture) areas of biology because of climate change and land use concerns [5-8]. Most transplant studies show decreased performance just across geographic range boundaries ([6] for review); therefore, it is generally assumed that many range boundaries are spatial mani‐ festations of niche limits, requiring adaptation for local range expansion or for the persistence of populations at range edges as climate changes. However, the existence of range limits suggests that adaptation to stressful environments just outside the range is often prevented. What prevents this adaptation from occurring?

Many factors and processes can contribute to range limit development. These factors include lack of genetic variation in range margin populations, barriers to dispersal, gene flow from elsewhere within the range, and various kinds of tradeoffs [6]. Any of these factors, alone or in combination with other factors, could prevent adaptation to stressful environments outside the range. However, there is often sufficient genetic variation within and among range margin populations [9] for natural selection to presumably act upon, and often there are no obvious barriers to dispersal at range boundaries. In these cases, possible constraints on the process of adaptation to stressful environments just outside the range would include gene flow and tradeoffs. But because many range margin populations are geographically and genetically isolated (e.g., [10]), it is thought that the study of range limit development should often focus on molecular, physiological, or developmental tradeoffs [11]. What kind of tradeoffs might be important to mustards at low elevation or low latitudinal range limits?

The process of adaptation often proceeds by modifying existing structures and pathways. Within ranges, stress response signal transduction pathways help plants to survive temporary challenges from abiotic and biotic stressors [12]. Just across range boundaries, some of these same stressors increase in frequency; therefore, one would predict that adaptation to stressful environments across range limits would involve the upregulation of stress response pathways such that the pathways and the traits that they regulate were expressed more frequently or stably.

However, evolutionary models predict that a problem may arise when antagonistic response pathways are co-opted simultaneously for evolutionary change [13]. The problem is with negative pleiotropic and epistatic effects. Multiple signaling pathways often form networks of regulatory genes (transcription factors) that may interact for multiple positive and negative integrative effects. An excellent example is the flowering time signaling network in Arabi‐ dopsis, which involves many positive and negative interactions among photoperiod, circadian clock, vernalization, autonomous and Gibberellic acid pathways. In general, quantitative geneticists predict that the evolution of complex traits may involve many genetic correlations, which is why there has been such an interest in the analysis of quantitative genetic variance– covariance G-matrices [14]. Indeed, early theoreticians such as Fisher predicted that the evolution of complex traits would only involve many genes of small effects to avoid such pleiotropic effects [15]. Epistatic interactions between major flowering time network genes FRI and FLC, for example, were one of the contributing factors in the maintenance of genetic variation in Arabidopsis flowering time [16] and therefore may represent an evolutionary constraint. FRI and FLC are major transcription factors (TFs) in the flowering time signaling network that allow large behavioral shifts involving many genes, but these major effects might impede evolution through multiple epistatic and pleiotropic effects.

spatially restricted distributions. This hypothesis is in contrast to previous hypotheses on defense evolution that argue the opposite that variation in defensive chemistry is the conse‐ quence of spatial distributions, life history patterns, inherent growth rates, etc. [3, 4]. As such, we begin this chapter with a description of the central hypothesis, followed by a set of predictions and then a description of the interdisciplinary approach needed to evaluate the

Understanding the causes and dynamics of naturally occurring range limits in plants has become a central issue in both basic (evolutionary ecology) and applied (conservation and agriculture) areas of biology because of climate change and land use concerns [5-8]. Most transplant studies show decreased performance just across geographic range boundaries ([6] for review); therefore, it is generally assumed that many range boundaries are spatial mani‐ festations of niche limits, requiring adaptation for local range expansion or for the persistence of populations at range edges as climate changes. However, the existence of range limits suggests that adaptation to stressful environments just outside the range is often prevented.

Many factors and processes can contribute to range limit development. These factors include lack of genetic variation in range margin populations, barriers to dispersal, gene flow from elsewhere within the range, and various kinds of tradeoffs [6]. Any of these factors, alone or in combination with other factors, could prevent adaptation to stressful environments outside the range. However, there is often sufficient genetic variation within and among range margin populations [9] for natural selection to presumably act upon, and often there are no obvious barriers to dispersal at range boundaries. In these cases, possible constraints on the process of adaptation to stressful environments just outside the range would include gene flow and tradeoffs. But because many range margin populations are geographically and genetically isolated (e.g., [10]), it is thought that the study of range limit development should often focus on molecular, physiological, or developmental tradeoffs [11]. What kind of tradeoffs might be

The process of adaptation often proceeds by modifying existing structures and pathways. Within ranges, stress response signal transduction pathways help plants to survive temporary challenges from abiotic and biotic stressors [12]. Just across range boundaries, some of these same stressors increase in frequency; therefore, one would predict that adaptation to stressful environments across range limits would involve the upregulation of stress response pathways such that the pathways and the traits that they regulate were expressed more frequently or

However, evolutionary models predict that a problem may arise when antagonistic response pathways are co-opted simultaneously for evolutionary change [13]. The problem is with negative pleiotropic and epistatic effects. Multiple signaling pathways often form networks of regulatory genes (transcription factors) that may interact for multiple positive and negative integrative effects. An excellent example is the flowering time signaling network in Arabi‐

important to mustards at low elevation or low latitudinal range limits?

hypothesis.

stably.

**2. Hypothesis**

What prevents this adaptation from occurring?

690 Abiotic and Biotic Stress in Plants - Recent Advances and Future Perspectives

In another example, it is generally assumed that plants face both abiotic and biotic stressors across range boundaries, especially at low elevation "trailing edges" of species ranges [17], yet it is well known that stress response pathways, such as abscisic acid (ABA) signaling for coping with abiotic stressors (e.g., drought) and jasmonic acid (JA) signaling for coping with biotic stressors (e.g., herbivores) may negatively interfere with one another ([18 - 20] for reviews). Thus, the simultaneous co-option of these antagonistic pathways for low-elevation range expansion where organisms face both increased abiotic and biotic stressors may be problematic because of the crosstalk. Although limited phylogenetic evidence suggests that ancient antagonistic crosstalk between signaling pathways may not constrain evolution [21], more thorough and experimental work is needed to address this issue.

**Figure 1.** Crosstalk between JA/ET and ABA signaling. MYBs, MYCs, and NAC are transcription factors. The key can‐ didate TFs in *Arabidopsis thaliana* either regulate aliphatic GS toxins (MYB28 [At5g61420] and MYB29 [At5g07690]) (44-48) or regulate both defense and drought tolerance pathways (AtMYC2 At1g32640, AtMYB2 At2g47190, and a NAC TF AtAF2 At5g08790) (18, 19, 20 for reviews).

Specifically, our hypothesis states that components of defense (e.g., JA – jasmonic acid) and stress tolerance (e.g., ABA – abscisic acid) signaling pathways are not genetically independent of one another, which may constrain the simultaneous evolution of defense and stress tolerance. Candidate genes in Arabidopsis for the evolutionary constraint include specific TFs that are involved in gene regulation within and among signaling pathways (Figure 1). For example, if there exists genetic variation in TFs MYC2 or AF2 that natural or artificial selection could act upon to increase drought stress tolerance traits for more stable expression, we predict that [1] a glucosinolate defense response would also change because these TFs help to regulate both pathways, and [2] other pleiotropic or epistatic effects would reduce fitness.
