TRENDS IN OPHIOGNOMONIA HOST EVOLUTION Several trends indicating host conservation, divergence, specialization, and switching were noted in each clade of the phylogeny of Ophiognomonia. When comparing broad trends, a major host-splitting event was suggested in more basally positioned nodes (nodes 1:2, 2:2, 2:6, and 3:5) when compared with the relative position of terminal nodes (e.g. node 3:12) for all three clades of Ophiognomonia. Patterns of host specialization differed from family → species for nodes 1:2 and 3:5, when compared with nodes 2:2 and 2:6, which show specialization from order → species (Figs 2–4). A general trend (with exceptions) of host genus and species specialization was observed at more terminally positioned nodes (relative to basal nodes) in all three clades of extant species of Ophiognomonia (Figs 2–4). For example, host genus and species specialization was observed in the more terminally positioned nodes 1:6, 1:7, 1:12 → 1: 15 in clade 1, 2:7 → 2: 10 in clade 2, and 3:7, 3:8, 3:9 → 3: 12 in clade 3 (Figs 2–4). Although the nodes in this phylogeny were not dated, this may suggest major host splits (order, family) early in the evolution of Ophiognomonia, and a trend of host genus and species specialization more recently. The remaining sections will discuss defined patterns of host conservation, divergence, specialization, and switching. HOST PLANT CONSERVATISM The time-for-speciation effect states that species richness will be greater in an area that has been evolutionarily conserved for a dramatically longer time when compared with richness in a recently occupied niche (Stephens & Wiens, 2003). We hypothesize that this may explain the rich diversity of species occurring on hosts in the Juglandaceae (node 2:3), Rosaceae (node 2:7) in clade 2, and Fagaceae in clade 1 (node 1:12; Fig. 2). For example, the subclade (node 1:12) including O. asiatica, O. kobayashii, O. otanii, O. setacea, and O. sogonovii shows host conservatism at the family rank for plants in Fagaceae (Fagales; Fig. 2). The subclade of species (nodes 2:7 → 2:10) including O. gei, O. nipponicae, O. padicola, O. rosae, and O. rubi-idaei suggests a pattern of host conservatism at the order and family rank, and specialization within host genus and species (Fig. 3). This subclade (node 2:7) only associates with hosts in Rosaceae; therefore, we hypothesize that these species will be collected on other genera in Rosaceae and will continue to speciate in this family. Species in clade 3 have the narrowest host range when compared with the other two clades of Ophiognomonia (Fig. 4; Table 1). New species are known to diverge gradually and occupy niches similar to their most recent ancestors, and therefore associate with phylogenetically related hosts (Prinzing et al., 2001; Martínez-Meyer et al., 2004; Wiens & Graham, 2005). For clade 3, much of the phylogenetic structure in host–fungus range can be considered niche conservatism for the host family Betulaceae (nodes 3:3, 3:4, and 3:6; Fig. 4). When simply listing host–fungus relationships, 12 of 15 species in clade 3 associate with Alnus spp. and Betula spp. (Betulaceae), only Ophiognomonia balsamiferae occurs on Populus balsamiferae in Salicaceae, and O. clavigignentijuglandacearum and Ophiognomonia pterocaryae associate with Juglans spp. and Pterocarya rhoifolia, respectively (Juglandaceae) (Fig. 4; Table 1). Within a phylogenetic context, a clear pattern of host order conservatism was observed from nodes 3:3 → 3:15, and host family conservatism was observed from nodes 3:6 → 3:12 (Fig. 4). Within Betulaceae, the species O. bugabensis, O. ibarakiensis, O. maximowiczianae, O. multirostrata, O. nana, and O. tucumanensis (node 3:6) occur only on either Alnus spp. or Betula spp. (Betulaceae and Fagales), whereas the species O. alni-viridis and O. intermedia have an expanded host range occurring on both Alnus spp. and Betula spp. Sogonov et al. (2008) found that the genus Gnomonia (Gnomoniaceae) has tight host association patterns with the family Betulaceae. In fact, species from this genus are not known to occur on plants outside of the Betulaceae. We hypothesize that most of the species in clade 3 (node 3:3, excluding O. clavigignenti-juglandacearum / O. pterocaryae) have associated and possibly co-evolved with Alnus and Betula. HOST PLANT SPECIALIZATION The life cycle of Ophiognomonia consists of dispersing ascospores and conidia into the environment via rain and wind, which must frequently present the opportunity for species to associate with new hosts at the community level in mixed forests (Giraud et al., 2008). Giraud et al. (2008) consider a host shift event as one factor contributing to the evolution of a new species, but not the sole mechanism of speciation; rather, allopatry or reproductive isolation must occur to solidify the event. For example, we hypothesize that the host switch (node 1:9; Fig. 2) in the sister species O. ostryae-virginianae and O. japonica (Ostrya virginiana, Betulaceae → Prunus sp., Rosaceae) represents a speciation event influenced by host association but solidified by spatial allopatry (O. ostryae-virginianae, North America; O. japonica, Japan). Host specificity has been hypothesized to contribute to a reproductive barrier essential for sympatric speciation events (Giraud et al., 2006; Peever, 2007). Two interesting patterns representing hostbased specialization among parapatric/sympatric species were observed in clade 1. The sister species Ophiognomonia ischnostyla and Ophiognomonia pseudoischnostyla (node 1:7) occur on Carpinus / Corylus and Alnus / Betula (Betulaceae) in Europe and Western Asia, from a broadly overlapping geographic range [latitude, (45.98) 46.32–57.08 (57.14); longitude, (6.58) 6.92–31.52 (35.32)]. Similarly, the species O. asiatica and O. sogonovii (node 1:15) associate with Quercus aliena / Quercus crenata and Quercus mongolica / Quercus serrata, respectively, and occur in closely overlapping spatial ranges in East Asia [latitude, (25.14) 35.98–36.23 (36.31); longitude, (102.75) 138.21–140.11 (140.20)]. Ophiognomonia ischnostyla and O. pseudoischnostyla show clear patterns of host genus and species specialization (Carpinus / Corylus and Alnus / Betula, respectively, D = 1.00, node 1:7) and O. asiatica / O. sogonovii show specialization at the species rank (Q. aliena / Q. crenata and Q. mongolica / Q. serrata, respectively, D = 0.85, node 1:15). The latter four species of Ophiognomonia were confirmed as genetically distinct lineages based on the genealogical congruence of multiple molecular markers and the genealogical sorting index (Walker et al., 2012a). We hypothesize that these four species of Ophiognomonia evolved partially because of parapatric/sympatric divergence by host usage, which triggered strong reproductive isolation (Giraud et al., 2006, 2008). The subclade of species (node 2:11) including Ophiognomonia cordicarpa, O. longispora, O. melanostyla, and O. sassafras is associated with hosts in Juglandaceae, Malvaceae, and Lauraceae (Fig. 3; Table 1). After the clear and statistically significant divergence event at node 2:6 (D = 0.79–1.00), these species suggest patterns typical of host specialization from order → species (nodes 2:11 → 2:13; Fig. 3). Within this subclade, O. longispora and O. melanostyla (nodes 2:12 → 2:13) occur on Tilia spp. (Malvaceae), O. cordicarpa occurs on Pterocarya rhoifolia (Juglandaceae), and O. sassafras occurs on Sassafras albidum (Lauraceae). These patterns suggest a recent host jump to Lauraceae and Malvaceae, and evolutionary conservation throughout geological history in the more common host families of Ophiognomonia, including Betulaceae, Fagaceae, Juglandaceae, and Rosaceae (clades 1–3; Figs 2–4). In addition, O. sassafras is the only species of Gnomoniaceae known to occur on hosts in the Laurales. HOST SWITCHING Two separate host jumps from the Fagales to the Rosales (order rank) were observed in the phylogeny of Ophiognomonia (nodes 1:9 and 2:5). The subclade of species including Ophiognomonia micromegala, O. pseudoclavulata, and Ophiognomonia vasiljevae (nodes 2:3 → 2:5) are associated with Carya and Juglans (Juglandaceae), except for O. lenticulispora, which occurs on Prunus in the Rosaceae (node 2:5; Fig. 3). A pattern of host conservatism in this subclade was observed at the order (Fagales) and family (Juglandaceae) ranks (nodes 2:3 → 2:4). Ophiognomonia lenticulispora represents a host switch from Carya tomentosa → Prunus sp., which occurred at the order rank (Fagales → Rosales) in this subclade (node 2:5; Fig. 3). The host genera Carya and Prunus often co-occur in forests where O. lenticulispora, O. micromegala, and O. pseudoclavulata were collected; therefore, we hypothesize that the ancestor of O. lenticulispora originated on Carya or another host in the Juglandaceae, and has evolved to specialize on Prunus. We also speculate that O. lenticulispora has the ability to switch within and between hosts in Rosaceae and Juglandaceae. Host-shifting events that resemble this example have been documented in Puccinia /Crucifer (Roy, 2001) and Microbotryum / Caryophyllaceae (Refrégier et al., 2008) species complexes. The other host jump from Fagales → Rosales occurred at node 1:9 (Fig. 2). The divergence event at node 1:9 is representative of a host shift from Betulaceae (Ostrya) to Rosaceae (Prunus) for O. japonica (D = 1.00; Fig. 2). The complete subclade (node 1:4) of species is associated with several genera in Betulaceae. Only O. japonica and O. michiganensis occur on Prunus in Rosaceae. We hypothesize that the shift to Prunus in O. japonica is a host jump (node 1:9), whereas the shift in O. michiganensis is an expansion of the host range (node 1:10; Fig. 2). Several authors hypothesized that pathogen association with a narrow host range of species provides an advantage over a generalist species with respect to the endless co-evolution of host–fungus competition (Whitlock, 1996; Kawecki, 1998). The subclade of species indicated at nodes 3:4 → 3:15 associate primarily with Alnus and Betula (Betulaceae), and generally show patterns of host conservatism at the order/family ranks (Fig. 4). Two host-switching events were observed in the subclade consisting of nodes 3:4 → 3:15 from hosts in Betulaceae to hosts in Juglandaceae. The species O. clavigignentijuglandacearum showed a host switch from Alnus to Juglans, and O. pterocaryae switched from Alnus to Pterocarya (nodes 3:14, 3:15; Fig. 4). A similar pattern was observed in the genus Cryptosporella (Gnomoniaceae): all species occur on hosts in Betulaceae, except for Cryptosporella hypodermia (Ulmaceae) and Cryptosporella wehmeyeriana (Malvaceae; Mejía et al., 2011b). Based on this pattern, Mejía et al. (2011b) hypothesize that species of Cryptosporella share a close association with hosts in Betulaceae. In this study we hypothesize that species of Ophiognomonia have co-evolved in close association with several host families in the Fagales, yet show several host shifts to other plant orders. HOST GENERALISTS VERSUS SPECIALISTS Host generalist and specialist strategies have most likely contributed to host associations in species of Ophiognomonia. The species O. michiganensis is associated with Alnus, Betula, and Carpinus in Betulaceae (Fagales), and Prunus in Rosaceae (Rosales) (node 1:10; Fig. 2; Table 1). Another species, O. setacea, occurs on plants from several host orders including Fagales, Proteales, and Sapindales (node 1:14; Fig. 2; Table 1). Both O. michiganensis and O. setacea are the only species of Ophiognomonia known to occur on several host orders, and are considered generalists with respect to host association. This is not uncommon among the Gnomoniaceae; in fact, Sogonov et al. (2007) documented the species Apiognomonia errabunda occurring on a diverse range of host orders, including Fagales, Malpighiales, Malvales, Rosales, and Sapindales. The remaining 43 species of Ophiognomonia associate with a single host genus or several genera from the same family, indicating tight host–fungus evolution for the majority of species in this genus over time. SAMPLE SIZE EFFECT Inevitably, only a subset of the total geographic and host range of Ophiognomonia was sampled in this study, but all data available for scientific studies were included in the study. For example, 11 species of Ophiognomonia are represented by a single herbarium specimen. Struwe et al. (2011) suggest the inclusion of all detailed specimen records, even if they represent a singleton species, to ensure the comprehensive analysis of clades in SEEVA. It is probable that host records for new collections of Ophiognomonia will closely coordinate with host associations from past species records (Martínez-Meyer et al., 2004; Wiens & Graham, 2005). We do not discount the possibility of new host records for any species of Ophiognomonia; however, we hypothesize that this is less likely for some than for others. For example, in clade 2 the subclade of species including O. gei, O. nipponicae, O. padicola, O. rosae, and O. rubi-idaei is restricted to the family Rosaceae. We hypothesize that this is a tight host–fungus association, and that these species are unlikely to occur on plants outside of the Rosaceae. In clade 3 (node 3:3 → 3:15), nearly all species are associated with Alnus and Betula (Betulaceae), except for O. clavigignenti-juglandacearum and O. pterocaryae, which occur on hosts in the Juglandaceae. We hypothesize that this indicates the potential for other species in this subclade (nodes 3:3 → 3:15) to shift to hosts in the Juglandaceae, but still maintain a strong association within the Betulaceae., Published as part of Walker, Donald M., Castlebury, Lisa A., Rossman, Amy Y. & Struwe, Lena, 2014, Host conservatism or host specialization? Patterns of fungal diversification are influenced by host plant specificity in Ophiognomonia (Gnomoniaceae: Diaporthales), pp. 1-16 in Biological Journal of the Linnean Society 111 (1) on pages 9-14, DOI: 10.1111/bij.12189, http://zenodo.org/record/7848455, {"references":["Stephens PR, Wiens JJ. 2003. Explaining species richness from continents to communities: the time-for-speciation effect in emydid turtles. American Naturalist 161: 112 - 128.","Prinzing A, Durka W, Klotz S, Brandl R. 2001. The niche of higher plants: evidence for phylogenetic conservatism. Proceedings of the Royal Society of London. Series B: Biological Sciences 268: 2383 - 2389.","Martinez-Meyer E, Townsend-Peterson A, Hargrove WW. 2004. Ecological niches as stable distributional constraints on mammal species, with implications for Pleistocene extinctions and climate change projections for biodiversity. Global Ecology and Biogeograpy 13: 305 - 314.","Wiens JJ, Graham CH. 2005. Niche conservatism: integrating evolution, ecology, and conservation biology. Annual Review of Ecology, Evolution, and Systematics 36: 519 - 539.","Sogonov MV, Castlebury LA, Rossman AY, Mejia LC, White JF. 2008. Leaf-inhabiting genera of the Gnomoniaceae, Diaporthales. Studies in Mycology 62: 1 - 77.","Giraud T, Refregier G, Le Gac M, de Vienne DM, Hood ME. 2008. Speciation in fungi. Fungal Genetics and Biology 45: 791 - 802.","Giraud T, Villareal LM, Austerlitz F, Le Gac M, Lavigne C. 2006. Importance of the life cycle in sympatric host race formation and speciation of pathogens. Phytopathology 96: 280 - 287.","Peever T. 2007. Role of host specificity in the speciation of Ascochyta pathogens of cool season food legumes. European Journal of Plant Pathology 119: 119 - 126.","Walker DM, Castlebury LA, Rossman AY, Mejia LC, White JF. 2012 a. Phylogeny and taxonomy of Ophiognomonia (Gnomoniaceae, Diaporthales), including twentyfive new species in this highly diverse genus. Fungal Diversity 57: 85 - 147.","Roy BA. 2001. Patterns of association between crucifers and their flower-mimic pathogens: host-jumps are more common than coevolution or cospeciation. Evolution 55: 41 - 53.","Refregier G, Le Gac M, Jabbour F, Widmer A, Hood M, Yockteng R, Shykoff J, Giraud T. 2008. Cophylogeny of the anther smut fungi and their Caryophyllaceous hosts: prevalence of host shifts and importance of delimiting parasite species. BMC Evolutionary Biology 8: 100. doi: 10.1186 / 1471 - 2148 - 1188 - 1100.","Whitlock MC. 1996. The Red Queen beats the Jack-of-alltrades: the limitations of phenotypic plasticity and niche breadth. American Naturalist 148: S 65 - S 77.","Kawecki TJ. 1998. Red Queen meets Santa Rosalia: arms races and the evolution of host specialization in organisms with parasitic lifestyles. American Naturalist 152: 635 - 651.","Mejia LC, Rossman AY, Castlebury LA, White JF. 2011 b. New species, phylogeny, host-associations, and geographic distribution of the genus Cryptosporella (Gnomoniaceae, Diaporthales). Mycologia 103: 379 - 399.","Sogonov MV, Castlebury LA, Rossman AY, White JF. 2007. The type species of Apiognomonia, A. veneta, with its Discula anamorph is distinct from A. errabunda. Mycological Research 111: 693 - 709.","Struwe L, Smouse PE, Heiberg E, Haag S, Lathrop RG. 2011. Spatial evolutionary and ecological vicariance analysis (SEEVA), a novel approach to biogeography and speciation research, with an example from Brazilian Gentianaceae. Journal of Biogeography 38: 1841 - 1854."]}