Study area

Three sampling plots were established as follows, one for each pine species:

1. School Forest, Iturbide Nuevo León; pine forest with Pinus pseudostrobus Lindl., other species identified within the plot are Quercus canbyi Trel., Amelanchier denticulata (Kunth) K. Koch, Arbutus xalapensis Kunth, and Prunus serotina Ehrh. 24º42’51” N 099º51’67” W; 1 630 masl.

2. Pablillo, Galeana, Nuevo León; pine forest with Pinus arizonica var. stormiae Mart. 24º36’51” N 100º00’16” W; 2 000 masl.

3. Pablillo, Galeana, Nuevo León; pine forest with Pinus cembroides Zucc. 24º37’11” N 100º00’38 W; 2 000 masl.

Leaf sampling

Ten fascicles of the following leaf types were collected at random, on a fortnightly basis during one year, in the selected plots: green leaves, senescent leaves, dead leaves still attached to the tree, “hanging” dead leaves (i.e. those that have not reached the topsoil and are suspended over the vegetation), topsoil leaves that have not yet begun to decay (L), leaves of the topsoil middle layer with some signs of decay (F1), topsoil leaves of the lower layer with signs of advanced decay (F2).

The samples were utilized for the isolation of fungi through the methods described below.

Isolation of leaf fungi in a culture medium

All the collected leaves were measured in the laboratory, and three leaves were selected at random from each sample and processed for the isolation of the fungi through a modification of the protocol proposed by ^{Kowalski and Zych (2002)}. The leaf surfaces were sterilized with ethanol at 96 % during one minute. Next, they were treated with sodium hypochlorite at 3 % during 10 minutes. Finally, they were submerged in ethanol at 70 % during 30 seconds. The leaves thus treated were cut transversally using a sterilized knife (disinfected by the flame of a burner). The number of cuts varied according to the size of the leaves. The sections were placed ―using tweezers (sterilized by the flame of a burner)― in a Petri dish with 2.5 % Malt Extract Agar and incubated at 22 °C, in a Shellab L120 incubator during the time required for the colonies to become visible to the naked eye.

All the colonies thus obtained were counted and identified by means of routine mycological techniques, through direct observation of preparations mounted on lactic acid, in glycerine-alcohol-water (GAW), or in distilled water, under a compound microscope. The works by ^{Sutton (1980)}, ^{Minter (1981)}, ^{Dennis (1981)}, ^{Ellis (1971}, ¹⁹⁷⁶), ^{Arx (1981)} and ^{Nag Raj (1993)}, and the website of the Index Fungorum (2018) (http://www.indexfungorum.org/) were consulted in order to corroborate the names of the species and their authors.

Subcultures of all the morphospecies present in the dishes were made in order to obtain vouchers of the studied fungi, which were deposited in the strain collection of the Herbario Micológico de la Facultad de Ciencias Forestales (CFNL) (School of Forest Sciences Mycological Herbarium) (CFNL), as well as of those whose species could not be determined because no morphological structures allowing their identification were available at the time.

Leaf fungi in moist chambers

A fascicle of leaves was selected at random from each sample collected fortnightly at the sample site for each leaf type, and it was placed in a moist chamber, which consisted in a Petri dish with sterilized Kraft paper humidified with sterile distilled water. These dishes were kept moist during a month, after which they were placed in an Acros^TM ART14JKX freezer at 5 °C during one week, in order to prevent an infestation by mites. The samples were then examined under a Karl Zeiss^TM Axiostar Plus microscope, as by the routine mycological techniques (^{Hawksworth, 1974}). The species were identified based on specialized literature (^{Sutton, 1980}; ^{Minter, 1981}; ^{Dennis, 1981}; ^{Ellis, 1971}, ¹⁹⁷⁶; ^{Arx, 1981}; ^{Nag Raj, 1993}).

The information was organized by collection and by leaf type, on an Excel 2013 spreadsheet, in order to estimate the richness, abundance, and frequency of the fungi associated to the leaves. Richness corresponded to the number of fungal species observed (total, per leaf type, per pine species); absolute abundance (Aba), to the number of individuals per species, in relation to the total number of individuals present in the study area (ni). Absolute frequency was calculated based on the number of points with the species, divided by the total number of sampled points; the fungal pattern in each of the studied leaves was compared by using a cluster analysis, with the PAST 3 statistical software (^{Hammer et al., 2001}).

Fungal diversity obtained with the method of culture in a Petri dish

26 leaf samplings were carried out for each of the species considered. 1 638 samples were processed for the three pine species.

The analysis included a total of 23 332 fungal colonies (Table 1).

Moreover, 1 953 subcultures were obtained, from which 57 fungal taxa were determined; 56 of these are adscribed to the Phyla Ascomycota, and only one (Gymnopus androsaceus), to the Phyla Basidiomycota. According to ^{Voříšková and Baldrian (2013)}, Ascomycota was the most abundant in green and senescent leaves, with 88.5 % and 99.5 % of amplifications, respectively. These findings agree with the present study, although the above authors used a molecular method to determine the fungal species.

^{Haňáčková et al. (2015)} utilized a culture method and a molecular method, and they concluded that both produced similar dominant species, and that in the early stage of decay, ascomycetes are predominant, and basidiomycetes occur only occasionally; in this, they also agree with the results documented here.

40 fungal taxa were registered for Pinus arizonica, 28 for P. cembroides and 32 for P. pseudostrobus. The complete list is shown in detail in Table 2. The most abundant species were Lophodermium austral Dearn., Pestalotia stevensonii Peck and Rhizosphaera kalkhoffii Bubák, while Cladosporium cladosporioides (Fresen.) G.A. de Vries, Alternaria alternata (Fr.) Keissl. and Ceuthospora sp. were the most frequent. According to ^{Yuan and Chen (2014)}, Lophodermium taxa are frequent pioneers on leaves and play an important role in litter decay.

The taxon with the greatest richness of fungi was P. arizonica, followed by P. pseudostrobus and P. cembroides. However, the highest number of colonies was registered in P. pseudostrobus, followed by P. arizonica. The number of colonies in P. cembroides, compared to the other pines, was 10 to 11 times smaller. This may be related to the length of the leaves. The average lengths of the sampled leaves were 173.9 mm for P. pseudostrobus, 156.1 mm for P. arizonica, and 27 mm for P. cembroides. Table 3 shows the minimum, mean and maximum number of fungal species per pine species and leaf type.

In all cases, leaf litter and gray leaves exhibited the highest richness. Green leaves ―especially those of P. cembroides― exhibited the lowest richness. The number of fungal species always diminished with the progress of the decomposition of the leaves. ^{Tokumasu et al. (1994)} cite similar results and state that hanging leaves and leaf litter exhibited the largest number of isolations.

As for the number of colonies, it was highest among the dead hanging leaves and the leaf litter. Fragmented and green leaves were the ones with the lowest number of colonies (Figure 1).

Número de colonias = Number of colonies; Tipo de hojas = Leaf type; Verde Green; Senescente = Senescent; Adherida = Attached; Colgantes = Hanging dead leaves; Suelo = Litter; Grises = Grey; Fragmentadas = Fragmented.

Figure 1 Number of colonies per leaf type in the studied pine species. 

In terms of the fungal richness patterns on the seven leaf types, the cluster analysis (Figure 2) distinguishes three groups: one made up of green and senescent leaves; another, by dead leaves (attached, hanging and litter), and one consisting of fragmented leaves only.

Verde Green; Senescente = Senescent; Adherida = Attached; Colgantes = Hanging dead leaves; Suelo = Litter; Grises = Grey; Fragmentadas = Fragmented.

Figure 2 Cluster analysis of the richness of fungal species in the various leaf types of the three studied pine species 

Fungal diversity of litterfall with moist chambers

27 samplings were carried out for Pinus arizonica and P. cembroides and 26 for P. pseudostrobus. A total of 400 samples were analyzed; 28 fungal taxa were identified on the attached leaves, the hanging leaves and the leaf litter. Table 4 shows the complete list of species.

Nine species were recorded for Pinus cembroides, of which Rhizosphaera kalkhoffii, Pestalotia stevensonii and Preussia sp. were most commonly found in attached leaves. Desmazierella acicola and Discosia strobilina were the most commonly found fungal species on leaf litter (Table 5). ^{Tokumasu et al. (1994)} observed similar patterns for Pinus sylvestris in Europe and noted the presence of D. acicola in leaf litter.

14 fungal species were identified in P. pseudostrobus: Lophodermium australe, Niesslia exilis and Gymnopus androsaceus were the most abundant in attached leaves and Lophodermium australe and Discosia strobilina were the most abundant in the leaf litter.

18 taxa were determined in P. arizonica. This pine species exhibited the highest richness; it also had the lowest absolute frequencies.

Well defined patterns were observed for the three pine species (Figure 3). Only Discosia strobilina and Preussia sp. were common fungi in the studied pine species.

Distancia = Distancia

Figure 3 Cluster analysis of the fungal richness on the leaves of the three pine species. 

When comparing the results obtained with the two techniques utilized, the moist chamber (direct method) turned out to be more effective for the determination of the fungal species. ^{Parungao et al. (2002)} reached to similar conclusions in their study on the fungal diversity in a tropical forest. Although the isolation in a culture medium (indirect method) made it possible to obtain twice as many taxa using the direct method, many of the cultures failed to form reproductive structures; for this reason, the list of unidentified species was long (50.87 %). This problem is discussed by ^{Bills and Polishook (1994)} and ^{Polishook et al. (1996)}, who cite 37 to 45 % of the taxa as unidentifiable. Furthermore, better patterns were defined for each pine species through the moist chamber technique. Many of the taxa identified with this means were not isolated in a culture medium. An example of this was Desmazierella acicola, very frequently found in the dead leaves of Pinus cembroides, which, according to ^{Ponge (1991)} is also a frequent fungus in the leaf litter of Pinus sylvestris. 13 species were identified for both taxa; of these, the most abundant was Lophodermium australe.