Survival
This is the first study to document consistent neighbor facilitation among trees experiencing repeated stressors over an extended time period. Trees of all sizes in close proximity with neighbors were considerably more likely to survive fire throughout the study. Our findings support the literature documenting positive neighbor effects among plants in stressful and highly disturbed environments (Bertness and Callaway 1994, Brooker et al. 2008, Guo et al. 2021).
We believe three factors contributed to this neighbor facilitation in terms of survival. The first involves the reduction of herbaceous fuel under tree canopies. The decline in fire frequency and intensity, and the associated decline in tree mortality that occurs when savannas transition to forests, is believed to be due to the reduced biomass of grasses, which are the primary fuel for surface fires in these environments (Frost and Robertson 1987, Holdo 2005, Archibald et al. 2009, Hoffmann et al. 2012). Rebertus and Burns (1997) found that tree survival in Quercus savannas was inversely related to grass cover. At CCESR, grass abundance is strongly negatively associated with tree canopy cover in the oak savanna/woodland habitats (Peterson et al. 2007). Thus, more neighboring trees means less herbaceous fuel and hence less fire risk.
The second factor involves the reduction of woody fuel. While both oak species contributed to neighbor facilitation involving survival, the survival benefit of having Q. macrocarpa as neighbors exceeded the benefit of having neighboring Q. ellipsoidalis for both species (Figs. 2 and 3). We think the difference in fire-susceptibility between the two oak species may account for this difference. Q. macrocarpa is considered among the most fire-resistant oaks, particularly as compared to Q. ellipsoidallis (Lorimer 1985, Frelich et al. 2015), illustrated by the much higher survival rate (260% higher) of Q. macrocarpa during the study (Davis 2021). This means that Q. ellipsoidalis is more likely to serve as fuel, threatening the survival of their neighbors. Conversely, the more fire-resistant Q. macrocarpa is likely to impede the spread of fire, reducing the threat of fire-induced mortality among its neighbors. Belote et al. (2015) found the fire-resistant Larix occidentalisincreased survival rates among nearby trees in a similar way.
The third factor involves mycorrhizae symbioses. A common explanation for neighbor facilitation among plants involves mycorrhizal sharing (Francis and Read 1984, Hunter and Aarrsen 1988, Simard et al. 1997, Teste and Simard 2008, Das et al. 2008, van der Heijden and Horton 2009, Bennett et al. 2017, Germain and Lutz 2021). Q. macrocarpa andQ. ellipsoidalis are both ectomycorrhizal trees (Avis et al. 2003) and red and white oaks are known to share mycorrhizae (Walker 2003, Avis et al. 2008). Thus, we believe that mycorrhizal sharing among both conspecifics and heterospecifics may also be contributing to the survival-related neighbor facilitation documented in this study. Other possible explanations for the increased likelihood of trees with neighbors surviving fire involve the reduced wind speeds and higher moisture levels associated with denser sites (Hoffmann et al. 2012, Belote et al. 2015).
It is possible an analysis could identify a positive neighbor effect among trees when one does not actually exist. For example, a false positive could result if surviving trees tend to be clumped in desirable microhabitats (Guo et al. 2021). However, while we did not examine all factors that could contribute to a favorable microhabitat, our analyses did include elevation (distance to water table) and soil nitrogen levels, both of which may be expected to influence tree survival.
Although neighbor facilitation with respect to survival persisted throughout the study, remaining strong in most cases, the strength of this relationship slightly declined over time for most size classes among both species (Figs. 2 and 3). The documented decline in neighbor facilitation over time might be due to the substantial reduction in the number of trees during the study. Only about one-third of the trees that were alive at the start of the study were still alive in 2020 (Davis 2021). This means that the number of neighbors for most trees would have declined substantially during this time. As the difference in the magnitude of high and low neighbor measurements declines, the ability of the models to discern positive neighbor relationships would be expected to decline as well.
A second possible explanation for the decline in the strength of neighbor facilitation involves the arrival of oak wilt (Bretziella fagacearum ) in east-central Minnesota several decades ago (Gibbs and French 1980), and which is thought to have spread rapidly at CCESR since 2010 (Sapes et al. 2021, Pellegrini et al. 2021). The fungus is spread primarily through root grafting among neighboring trees and by sap beetles (Nitidulidae) (Kuntz and Riker 1956, Jagemann et al. 2018). Thus, while underground root and mycorrhizal connections may reduce the likelihood of fire-induced mortality for the two oak species, this same network could facilitate the transmittance of oak wilt.
It is possible, then, that oak wilt might have reduced the neighbor benefits over time. However, in a separate study at CCESR, Pellegrini et al. (2021) found that while oak wilt dramatically increased tree mortality in unburned areas, where trees are densely packed and which are dominated by Q. ellipsoidalis , it had little effect in frequently burned sites. Also, since species in the red oak group lack tyloses, they are not as able to effectively compartmentalize scarring injuries as are white oaks, thereby making them considerably more susceptible than white oaks to subsequent fungal infection (Gibbs and French 1980, Juzwik et al. 2011, Dey and Schweitzer 2018). However, while Q. macrocarpa is much less vulnerable to oak wilt thanQ. ellipsoidalis , it exhibited a similar decline in neighbor facilitation. Thus, it seems unlikely oak wilt contributed much, if anything, to the decline in the strength of neighbor facilitation during the study.