Downloaded from http://rsbl.royalsocietypublishing.org/ on November 22, 2014
Evolutionary biology
Why are so many trees hollow? rsbl.royalsocietypublishing.org
Graeme D. Ruxton School of Biology, University of St Andrews, St Andrews KY16 9TH, UK
Opinion piece Cite this article: Ruxton GD. 2014 Why are so many trees hollow? Biol. Lett. 10: 20140555. http://dx.doi.org/10.1098/rsbl.2014.0555
Received: 15 July 2014 Accepted: 19 October 2014
Subject Areas: biomechanics, ecology, evolution, plant science Keywords: herbivory, plant defences, termites, wood
Author for correspondence: Graeme D. Ruxton e-mail: [email protected]
In many living trees, much of the interior of the trunk can be rotten or even hollowed out. Previously, this has been suggested to be adaptive, with microbial or animal consumption of interior wood producing a rain of nutrients to the soil beneath the tree that allows recycling of those nutrients into new growth via the trees roots. Here I propose an alternative (non-exclusive) explanation: such loss of wood comes at very little cost to the tree and so investment in costly chemical defence of this wood is not economic. I discuss how this theory can be tested empirically.
Many trees have a hollow or otherwise rotten core to their main trunk. For example, surveys of savannah woodland in Australia have found hollow cores (a phenomenon called piping) in 66–89% of trees of different species; on average, hollow cores extended to 50% of the total diameter [1– 3]. A study in the Amazonian rainforest found 37% of trees from a broad range of species to be piped [4]. The most relevant hypothesis to explain this phenomenon is due to the renowned ecologist Daniel H. Janzen, who in 1976 published a paper entitled ‘Why tropical trees have rotten cores’ [5]. He hypothesized that ‘the rotten core is often an adaptive trait . . . A rotten core is a site of animal nests, animal defecation, and microbial metabolism that should result in a steady fertilization of the soil under the base of the tree. . . . Hollow cores are expected more frequently in nutrient poor sites . . . In short, the hollow core becomes a clever use of an otherwise useless piece of wood.’ Mutualistic interactions between microbes and plants are well known in other contexts, most obviously in the rhizosphere. Although it seems logically plausible, empirical support for the predictions of this hypothesis have been lacking. There may be low return of nitrogen to a living plant from termite activity within its trunk, with much of the nitrogen harvested from the consumption of the wood of a living tree being lost through predation and nuptial flights and as a result of heavy rains [6]. Further, any nutrients that are released into the soil are more likely to be captured by the shallower roots of the herbaceous understorey that compete with larger deeper-rooted canopy trees [7]. One key study found that in nutrient-poor Australian woodlands, growth and survival were lower in piped trees than size-matched unpiped trees growing in the same plot [8]. A follow-up study found that for one species of Eucalyptus, growth rate was independent of the extent of piping in trees of a particular total trunk diameter, whereas in the other study species growth decreased with extent of piping [9]. Here I offer an alternative (non-exclusive) hypothesis. In contrast to Janzen’s hypothesis, I suggest that the tree gains no advantage from a rotten core, but rather ( provided the rottenness is not too extensive) pays a very small price (in terms of reduced structural integrity) for allowing the central wood to rot. Furthermore, this cost can be lower than the metabolic costs required to chemically protect the central part of the trunk. So like Janzen’s hypothesis, my hypothesis suggests that the prevalence of rotten cores can be understood in terms of selection pressures. But my hypothesis is very different in suggesting that there is no selective benefit from a rotten core (in terms of freeing up nutrients for recycling by the tree’s roots), but that it is often not economical to chemically defend the central part of the trunk; and rotten cores emerge as an epi-phenonenon of this judicious lack of investment.
& 2014 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rsbl.royalsocietypublishing.org/ on November 22, 2014
Acknowledgement. This manuscript benefited from perceptive and constructive reviewing from Hanns-Christof Spatz and two anonymous referees.
2
Biol. Lett. 10: 20140555
will be more at a premium, and differential investment in chemical protection across the circumference of trees, with the outer portion being much more strongly protected than the inner. There is evidence that chemical protection of heartwood can vary according to local conditions, being affected by thinning during silviculture [16], but that same study found no evidence of consistent variation in defensive investment with heartwood age. It might also be valuable to explore extent of hollowing in large lateral branches as well as the central trunk of suitable species. Horizontal branches will often experience greater tensile and compressive stresses than an upright trunk of similar thickness, and might be predicted to be particularly strongly protected against decay. We might also predict that trees should avoid any hollowing of the trunk in the vicinity of large lateral branches. A lateral branch causes stresses to be imposed on the trunk, and if there is hollowing of the trunk near the branch then stresses will be concentrated on the outer wall and potential for local buckling increases relative to a situation where a solid trunk allows spreading of those stresses across the cross-section of the trunk (see [17] for further discussion). This illustrates that there will be a complex suite of selection pressures impinging on the design of tree trunks (and all other plant structures [18]). Furthermore, as Janzen’s theory suggests a benefit to the plant from heartwood consumption, his theory would predict that plants should be selected to make this wood more accessible and/or attractive to decomposers. By contrast, if any such traits are otherwise costly then they should not be selected if the mechanism proposed here operates. It may also be illuminating to examine the diversity of bacterial and fungal communities in decaying central wood and contrast this with those in rotting wood associated with damage, to explore if there are functional differences in the different assemblages. Naturally hollow trees are very unusual, but do occur in the Neotropical pioneer species of the genus Cecropia, where internodes are either hollow or filled with pith—depending on the species. These species are fast-growing, relatively short-lived gap colonizers in tropical forests. They are characterized by having a relatively small number of slender branches at the top of the trunk carrying sparse leaf cover. Thus, the central trunk does not have to bear the same loads as those of more robust and more branched trees. Accordingly, the trunk is essentially columnar (like bamboo stems) rather than the more traditional tapered shape created by thickening at the bottom of the trunk over time. Cecropia does show, however, that naturally hollow trees are possible, but in most cases the complex of selection pressures on tree trunk design seem not to have led to this solution, although the sometimes extensive occurrence of microbially driven hollowing mentioned at the start of this article perhaps suggests that the costs of hollowing may often not be high. But we require more empirical exploration of hollowing, and I very much hope this article encourages and focusses such investigation.
rsbl.royalsocietypublishing.org
Chemical defences against both microbial and animal damage to tissues are widespread across plants and can often involve very substantial metabolic expenditure. Hence, it may not be economical for a plant to protect all its tissue equally, and I will argue that the interior of tree trunks will often not be well defended. As a tree grows, so the trunk must thicken to carry the increased weight; such thickening involves the formation of new sapwood around the circumference of the trunk. Sapwood acts not only structurally but also to transport water from roots to leaves and as a repository for nutrients. However, sapwood is metabolically active and the maintenance of such tissue has been estimated as costing 5–13% of annual net energy gain through photosynthesis [10]. For this reason, many trees convert their innermost wood from sapwood to heartwood. Heartwood is metabolically inactive, plays no role in nutrient storage or water transport and offers no structural advantage over sapwood [11]. The attraction of heartwood is that it is dead and incurs no recurrent metabolic costs. Although some chemical defences may be laid down at the time of heartwood formation, the lack of metabolism in this wood means that such defences cannot be replenished when they decay, or be ramped up in response to specific attack. Thus, there may be metabolic savings a tree can obtain by avoiding investing heavily in the chemical defence of wood in the interior of the trunk. Given that this interior is often heartwood, the primary cost to this strategy is likely to be structural weakness as a result of decay or hollowing of this wood; below I will argue that this cost can be very low. As a simplification, the trunk of a tree can be considered as a cantilevered cylinder. As such, the greatest tensile and compressive stresses occur towards the surface and the interior contributes relatively little to structural strength [12]. A survey of previous research of the effects of trunk hollowing on the structural failure of trees found strong agreement across studies, involving a broad range of different species and broad range of tree sizes, that there was a critical amount of hollowing above which structural failure was considerably more likely [13]. This critical point occurs when the radius of the inner hollow region is approximately 70% of the total radius of the trunk. Hollowness less than this critical amount involves very little cost in reduced structural stability. This 70% critical value is broadly used in the management of trees, and in particular is a very widely used criterion for the removal of trees considered to be structurally at risk [14], although its theoretical basis is an area of still-active discussion [15]. In conclusion, I agree with Janzen that the widespread occurrence of hollow trees demands explanation. I also agree that an evolutionary approach can help generate hypotheses that might explain this phenomenon. Janzen offered one such explanation in terms of recycling of nutrients; here I offer what I believe is a more generally applicable alternative or complementary hypothesis: that the central wood of trees is allowed to decay because the costs of chemically defending it are not justified by the small reduction in structural stability that is likely to occur. This hypothesis is empirically testable. I predict a greater investment in chemical protection of wood against destruction in exposed and windy environments where structural stability
Downloaded from http://rsbl.royalsocietypublishing.org/ on November 22, 2014
References
2.
3.
5. 6.
7.
Australia. Aust. Ecol. 30, 625– 647. (doi:10.1111/j. 1442-9993.2005.01491.x) 8. Werner PA, Prior LD. 2007 Tree-piping termites and growth and survival of host trees in savanna woodland of north Australia. J. Trop. Ecol. 23, 611 –622. (doi:10.1017/S0266467407004476) 9. Werner PA, Prior LD, Forner J. 2008 Growth and survival of termite-piped Eucalyptus tetrodonta and E. miniata in northern Australian: implications for harvest of trees for didgeridoos. For. Ecol. Manag. 256, 328–334. (doi:10. 1016/j.foreco.2008.04.027) 10. Ryan MG, Gower ST, Hubbard RM, Warig RN, Gholz HL, Cropper Jr WP, Running SW. 1995 Woody tissue maintenance respiration of four conifers in contrasting climates. Oecologia 101, 133–140. (doi:10.1007/BF00317276) 11. Taylor AM, Gartner BL, Morrell JJ. 2002 Heartwood formation and natural durability—a review. Wood Fibre Sci. 34, 587–611. 12. Wainwright SA. 1988 Axis and circumference: the cylindrical shape of plants and animals. Harvard, MA: Harvard University Press.
13. Mattheck C, Bethge K, Tesari I. 2006 Shear effects on failure of hollow trees. Trees 20, 329 –333. (doi:10.1007/s00468-005-0044-0) 14. Fink S. 2009 Hazard tree identification by visual assessment (VTA): scientifically solid and practically approved. Arboricult. J. 32, 139–155. (doi:10.1080/ 03071375.2009.9747570) 15. Spatz H-C, Niklas KJ. 2013 Modes of failure in tubular plant organs. Am. J. Bot. 100, 332 –336. (doi:10.3732/ajb.1200454) 16. Taylor AM, Gartner BL, Morrell JJ. 2003 Co-incident variations in growth rate and heartwood extractive concentration in Douglas-fir. For. Ecol. Manag. 186, 257 –260. (doi:10.1016/S03781127(03)00278-0) 17. Mosbrugger V. 1990 The tree habit in land plants. A functional comparison of trunk constructions with a brief introduction into the biomechanics of trees. Berlin, Germany: Springer. 18. Niklas KJ, Spatz H-C. 2012 Plant physics. Chicago: Chicago University Press.
Biol. Lett. 10: 20140555
4.
Forner J. 2006 The globalisation of the didjeridu and the implications for small scale community based producers in remote northern Australia. Int. J. Environ. Cult. Econ. Soc. Sust. 2, 137–148. Braithwaite RW. 1985 The Kakadu fauna survey: an ecological survey of Kakudu National Park. Canberra, Australia: Australian Parks and Wildlife Service. Fox RE, Clark NB. 1972 The incidence of termites in eucalypts of the Darwin area. Aust. For. Res. 5, 29 –36. Apolinario FE, Martius C. 2004 Ecological role of termites (Insecta, Isoptera) in tree trunks in central Amazonian rain forests. For. Ecol. Manag. 194, 23 –28. (doi:10.1016/j.foreco.2004.01.052) Janzen DH. 1976 Why tropical trees have rotten cores. Biotropica 8, 110. (doi:10.2307/2989630) Birkill A. 1985 Termite ecology in a tropical savanna grazed by cattle. Honours thesis, Flinders University, South Australia. Werner PA. 2005 Impact of feral water and fire on growth and survival of mature savanna trees: an experimental field study in Kakuda National Park,
rsbl.royalsocietypublishing.org
1.
3
Evolutionary biology
Why are so many trees hollow? rsbl.royalsocietypublishing.org
Graeme D. Ruxton School of Biology, University of St Andrews, St Andrews KY16 9TH, UK
Opinion piece Cite this article: Ruxton GD. 2014 Why are so many trees hollow? Biol. Lett. 10: 20140555. http://dx.doi.org/10.1098/rsbl.2014.0555
Received: 15 July 2014 Accepted: 19 October 2014
Subject Areas: biomechanics, ecology, evolution, plant science Keywords: herbivory, plant defences, termites, wood
Author for correspondence: Graeme D. Ruxton e-mail: [email protected]
In many living trees, much of the interior of the trunk can be rotten or even hollowed out. Previously, this has been suggested to be adaptive, with microbial or animal consumption of interior wood producing a rain of nutrients to the soil beneath the tree that allows recycling of those nutrients into new growth via the trees roots. Here I propose an alternative (non-exclusive) explanation: such loss of wood comes at very little cost to the tree and so investment in costly chemical defence of this wood is not economic. I discuss how this theory can be tested empirically.
Many trees have a hollow or otherwise rotten core to their main trunk. For example, surveys of savannah woodland in Australia have found hollow cores (a phenomenon called piping) in 66–89% of trees of different species; on average, hollow cores extended to 50% of the total diameter [1– 3]. A study in the Amazonian rainforest found 37% of trees from a broad range of species to be piped [4]. The most relevant hypothesis to explain this phenomenon is due to the renowned ecologist Daniel H. Janzen, who in 1976 published a paper entitled ‘Why tropical trees have rotten cores’ [5]. He hypothesized that ‘the rotten core is often an adaptive trait . . . A rotten core is a site of animal nests, animal defecation, and microbial metabolism that should result in a steady fertilization of the soil under the base of the tree. . . . Hollow cores are expected more frequently in nutrient poor sites . . . In short, the hollow core becomes a clever use of an otherwise useless piece of wood.’ Mutualistic interactions between microbes and plants are well known in other contexts, most obviously in the rhizosphere. Although it seems logically plausible, empirical support for the predictions of this hypothesis have been lacking. There may be low return of nitrogen to a living plant from termite activity within its trunk, with much of the nitrogen harvested from the consumption of the wood of a living tree being lost through predation and nuptial flights and as a result of heavy rains [6]. Further, any nutrients that are released into the soil are more likely to be captured by the shallower roots of the herbaceous understorey that compete with larger deeper-rooted canopy trees [7]. One key study found that in nutrient-poor Australian woodlands, growth and survival were lower in piped trees than size-matched unpiped trees growing in the same plot [8]. A follow-up study found that for one species of Eucalyptus, growth rate was independent of the extent of piping in trees of a particular total trunk diameter, whereas in the other study species growth decreased with extent of piping [9]. Here I offer an alternative (non-exclusive) hypothesis. In contrast to Janzen’s hypothesis, I suggest that the tree gains no advantage from a rotten core, but rather ( provided the rottenness is not too extensive) pays a very small price (in terms of reduced structural integrity) for allowing the central wood to rot. Furthermore, this cost can be lower than the metabolic costs required to chemically protect the central part of the trunk. So like Janzen’s hypothesis, my hypothesis suggests that the prevalence of rotten cores can be understood in terms of selection pressures. But my hypothesis is very different in suggesting that there is no selective benefit from a rotten core (in terms of freeing up nutrients for recycling by the tree’s roots), but that it is often not economical to chemically defend the central part of the trunk; and rotten cores emerge as an epi-phenonenon of this judicious lack of investment.
& 2014 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rsbl.royalsocietypublishing.org/ on November 22, 2014
Acknowledgement. This manuscript benefited from perceptive and constructive reviewing from Hanns-Christof Spatz and two anonymous referees.
2
Biol. Lett. 10: 20140555
will be more at a premium, and differential investment in chemical protection across the circumference of trees, with the outer portion being much more strongly protected than the inner. There is evidence that chemical protection of heartwood can vary according to local conditions, being affected by thinning during silviculture [16], but that same study found no evidence of consistent variation in defensive investment with heartwood age. It might also be valuable to explore extent of hollowing in large lateral branches as well as the central trunk of suitable species. Horizontal branches will often experience greater tensile and compressive stresses than an upright trunk of similar thickness, and might be predicted to be particularly strongly protected against decay. We might also predict that trees should avoid any hollowing of the trunk in the vicinity of large lateral branches. A lateral branch causes stresses to be imposed on the trunk, and if there is hollowing of the trunk near the branch then stresses will be concentrated on the outer wall and potential for local buckling increases relative to a situation where a solid trunk allows spreading of those stresses across the cross-section of the trunk (see [17] for further discussion). This illustrates that there will be a complex suite of selection pressures impinging on the design of tree trunks (and all other plant structures [18]). Furthermore, as Janzen’s theory suggests a benefit to the plant from heartwood consumption, his theory would predict that plants should be selected to make this wood more accessible and/or attractive to decomposers. By contrast, if any such traits are otherwise costly then they should not be selected if the mechanism proposed here operates. It may also be illuminating to examine the diversity of bacterial and fungal communities in decaying central wood and contrast this with those in rotting wood associated with damage, to explore if there are functional differences in the different assemblages. Naturally hollow trees are very unusual, but do occur in the Neotropical pioneer species of the genus Cecropia, where internodes are either hollow or filled with pith—depending on the species. These species are fast-growing, relatively short-lived gap colonizers in tropical forests. They are characterized by having a relatively small number of slender branches at the top of the trunk carrying sparse leaf cover. Thus, the central trunk does not have to bear the same loads as those of more robust and more branched trees. Accordingly, the trunk is essentially columnar (like bamboo stems) rather than the more traditional tapered shape created by thickening at the bottom of the trunk over time. Cecropia does show, however, that naturally hollow trees are possible, but in most cases the complex of selection pressures on tree trunk design seem not to have led to this solution, although the sometimes extensive occurrence of microbially driven hollowing mentioned at the start of this article perhaps suggests that the costs of hollowing may often not be high. But we require more empirical exploration of hollowing, and I very much hope this article encourages and focusses such investigation.
rsbl.royalsocietypublishing.org
Chemical defences against both microbial and animal damage to tissues are widespread across plants and can often involve very substantial metabolic expenditure. Hence, it may not be economical for a plant to protect all its tissue equally, and I will argue that the interior of tree trunks will often not be well defended. As a tree grows, so the trunk must thicken to carry the increased weight; such thickening involves the formation of new sapwood around the circumference of the trunk. Sapwood acts not only structurally but also to transport water from roots to leaves and as a repository for nutrients. However, sapwood is metabolically active and the maintenance of such tissue has been estimated as costing 5–13% of annual net energy gain through photosynthesis [10]. For this reason, many trees convert their innermost wood from sapwood to heartwood. Heartwood is metabolically inactive, plays no role in nutrient storage or water transport and offers no structural advantage over sapwood [11]. The attraction of heartwood is that it is dead and incurs no recurrent metabolic costs. Although some chemical defences may be laid down at the time of heartwood formation, the lack of metabolism in this wood means that such defences cannot be replenished when they decay, or be ramped up in response to specific attack. Thus, there may be metabolic savings a tree can obtain by avoiding investing heavily in the chemical defence of wood in the interior of the trunk. Given that this interior is often heartwood, the primary cost to this strategy is likely to be structural weakness as a result of decay or hollowing of this wood; below I will argue that this cost can be very low. As a simplification, the trunk of a tree can be considered as a cantilevered cylinder. As such, the greatest tensile and compressive stresses occur towards the surface and the interior contributes relatively little to structural strength [12]. A survey of previous research of the effects of trunk hollowing on the structural failure of trees found strong agreement across studies, involving a broad range of different species and broad range of tree sizes, that there was a critical amount of hollowing above which structural failure was considerably more likely [13]. This critical point occurs when the radius of the inner hollow region is approximately 70% of the total radius of the trunk. Hollowness less than this critical amount involves very little cost in reduced structural stability. This 70% critical value is broadly used in the management of trees, and in particular is a very widely used criterion for the removal of trees considered to be structurally at risk [14], although its theoretical basis is an area of still-active discussion [15]. In conclusion, I agree with Janzen that the widespread occurrence of hollow trees demands explanation. I also agree that an evolutionary approach can help generate hypotheses that might explain this phenomenon. Janzen offered one such explanation in terms of recycling of nutrients; here I offer what I believe is a more generally applicable alternative or complementary hypothesis: that the central wood of trees is allowed to decay because the costs of chemically defending it are not justified by the small reduction in structural stability that is likely to occur. This hypothesis is empirically testable. I predict a greater investment in chemical protection of wood against destruction in exposed and windy environments where structural stability
Downloaded from http://rsbl.royalsocietypublishing.org/ on November 22, 2014
References
2.
3.
5. 6.
7.
Australia. Aust. Ecol. 30, 625– 647. (doi:10.1111/j. 1442-9993.2005.01491.x) 8. Werner PA, Prior LD. 2007 Tree-piping termites and growth and survival of host trees in savanna woodland of north Australia. J. Trop. Ecol. 23, 611 –622. (doi:10.1017/S0266467407004476) 9. Werner PA, Prior LD, Forner J. 2008 Growth and survival of termite-piped Eucalyptus tetrodonta and E. miniata in northern Australian: implications for harvest of trees for didgeridoos. For. Ecol. Manag. 256, 328–334. (doi:10. 1016/j.foreco.2008.04.027) 10. Ryan MG, Gower ST, Hubbard RM, Warig RN, Gholz HL, Cropper Jr WP, Running SW. 1995 Woody tissue maintenance respiration of four conifers in contrasting climates. Oecologia 101, 133–140. (doi:10.1007/BF00317276) 11. Taylor AM, Gartner BL, Morrell JJ. 2002 Heartwood formation and natural durability—a review. Wood Fibre Sci. 34, 587–611. 12. Wainwright SA. 1988 Axis and circumference: the cylindrical shape of plants and animals. Harvard, MA: Harvard University Press.
13. Mattheck C, Bethge K, Tesari I. 2006 Shear effects on failure of hollow trees. Trees 20, 329 –333. (doi:10.1007/s00468-005-0044-0) 14. Fink S. 2009 Hazard tree identification by visual assessment (VTA): scientifically solid and practically approved. Arboricult. J. 32, 139–155. (doi:10.1080/ 03071375.2009.9747570) 15. Spatz H-C, Niklas KJ. 2013 Modes of failure in tubular plant organs. Am. J. Bot. 100, 332 –336. (doi:10.3732/ajb.1200454) 16. Taylor AM, Gartner BL, Morrell JJ. 2003 Co-incident variations in growth rate and heartwood extractive concentration in Douglas-fir. For. Ecol. Manag. 186, 257 –260. (doi:10.1016/S03781127(03)00278-0) 17. Mosbrugger V. 1990 The tree habit in land plants. A functional comparison of trunk constructions with a brief introduction into the biomechanics of trees. Berlin, Germany: Springer. 18. Niklas KJ, Spatz H-C. 2012 Plant physics. Chicago: Chicago University Press.
Biol. Lett. 10: 20140555
4.
Forner J. 2006 The globalisation of the didjeridu and the implications for small scale community based producers in remote northern Australia. Int. J. Environ. Cult. Econ. Soc. Sust. 2, 137–148. Braithwaite RW. 1985 The Kakadu fauna survey: an ecological survey of Kakudu National Park. Canberra, Australia: Australian Parks and Wildlife Service. Fox RE, Clark NB. 1972 The incidence of termites in eucalypts of the Darwin area. Aust. For. Res. 5, 29 –36. Apolinario FE, Martius C. 2004 Ecological role of termites (Insecta, Isoptera) in tree trunks in central Amazonian rain forests. For. Ecol. Manag. 194, 23 –28. (doi:10.1016/j.foreco.2004.01.052) Janzen DH. 1976 Why tropical trees have rotten cores. Biotropica 8, 110. (doi:10.2307/2989630) Birkill A. 1985 Termite ecology in a tropical savanna grazed by cattle. Honours thesis, Flinders University, South Australia. Werner PA. 2005 Impact of feral water and fire on growth and survival of mature savanna trees: an experimental field study in Kakuda National Park,
rsbl.royalsocietypublishing.org
1.
3
![[Why there are so many overdiagnoses in clinics].](/assets/img/hold.png)
