Myxomycete diversity in the coastal desert of Peru with emphasis on the lomas formations

Lado, C., Wrigley de Basanta, D., Estrada-Torres, A. & Stephenson, S.L. 2016. Myxomycete diversity in the coastal desert of Peru with emphasis on the lomas formations. Anales Jard. Bot. Madrid 73(1): e032. Results obtained from the first survey for myxomycetes in the hyperarid areas of Peru are reported. Sampling over three consecutive years was carried out at 96 localities between 0 and 1500 m elevation. A total of 78 species from 23 genera in all 6 myxomycete orders were determined from 723 collections of myxomycetes. One new species, Didymium peruvianum, is described, 2 species new to the Neotropics, 4 new records for South America and 41 additional new species for Peru are reported, increasing the country catalogue by over 50%. Results show that arid areas are rich in myxomycetes, and that each area has a unique species assemblage. Endemic plants had a particular relevance as myxomycete substrates. The predominance of the order Physarales in arid areas is reinforced, and the ecological importance of coastal fogs (garúas) is evident from the results. Comments are included on interesting or rare collections, as are SEM micrographs of several species and statistical evaluation


INTRODUCTION
The Myxomycetes or plasmodial slime moulds are a group of microorganisms included in the Amoebozoa, a major taxonomic group of amoeboid protists, considered as a protozoan fungal analogue (Kirk & al., 2011) due to their characteristic of producing spores in static fruiting bodies.Myxomycetes have been found in all terrestrial ecosystems, and have, until recently, been associated with temperate humid environments, on account of the need for water to complete their life cycle.However it has become evident that an exclusive and interesting number of this myxobiota, thrive in extremely arid or semi-arid environments, such as warm and cold deserts where the availability of liquid water from rain is practically nil.The number of species in these arid regions, especially in the Neotropics, is surprisingly high, since in the Atacama Desert 24 species were recorded by Lado & al. (2007), more than 100 taxa were reported by Estrada-Torres & al. (2009) in the deserts of Mexico, and in the Monte desert of Argentina 72 species were isolated (Lado & al., 2011).Research carried out in the Patagonian steppe registered 133 different species (Lado & al., 2014), almost the 15% of the total number of species known worldwide.These arid regions have become even more interesting because of the number of new species that have been discovered and described from them (Lado & al., 1999(Lado & al., , 2007(Lado & al., , 2009(Lado & al., , 2013(Lado & al., , 2014;;Estrada-Torres & al., 2001, 2009;Wrigley de Basanta & al., 2008b, 2009, 2010a, 2011, 2012, 2015).These new species represent four different orders, and the genera Cribraria, Didymium, Licea, Macbrideola, Perichaena and Physarum.
Since the publication of the first records by Rudolphi (1829), only 31 species of myxomycete had been published from Peru, until a recent paper increased the number to 80 (Rojas & al., 2011), but almost all of these records were from the tropical forests of the Amazon basin, to the East of the Peruvian Andes.In the checklist compiled by Lado & Wrigley de Basanta (2008) the country was among the least studied in the whole of the Neotropics with only five publications referring to Myxomycetes.There are no data at all from the arid areas of Peru, and no systematic study, catalogue or inventory had ever been carried out on this enormous dry Peruvian territory.A three-year study, supported by a grant from the Spanish Government, was undertaken to carry out a field-based investigation of the myxobiota of these arid areas.In this paper the first substantial data set on the Myxomycetes of the drylands of Peru is provided.
The ecoregion to be included in the study is the Peruvian Coastal Desert.This desert forms a continuous belt of arid land for more than 2500 km, extending between the Pacific Ocean and the Western slopes of the Andes.It is limited to the North by the border with Ecuador (3°25′S) and to the South by the Chilean border and the Atacama Desert (18°21′S), and makes up about 11% of the country (Rundel & al., 1991).It is a hyper-arid territory, receiving 0-300 mm of rain annually.The coast bordering the desert to the West is bathed by the cold Humboldt Current, and subsequent temperature inversion results in dense coastal fogs or "garuas" that settle on the lower westerly slopes of the Andes, and provide the only moisture for certain plants to grow in an otherwise barren desert (Dillon, 1997).The hyper-aridity of the southern Peruvian Desert appears to be very old, at least regionally, and has largely prevailed for the last 13-15 million years.The uplift of central Andean Cordillera during the Oligocene and Early Miocene was Fig. 1.Map of the general study area and localities sampled that were positive for myxomycete species (numbers refer to Table 1).
a critical factor in the formation of these arid conditions (Rundel & al., 1991).In the North, the Sechura Desert, sometimes has torrential rain associated with the phenomenon of El Niño, the fact that the Humboldt Current moves West away from the coast at these latitudes, and the proximity of the area is to the humid tropical weather from the Equator.
Emphasis of the study reported in this paper was placed on the lower elevations (0-1,500 m) that include the "lomas formations" (Weberbauer, 1945).These are terrestrial islands surrounded by hyper-arid areas devoid of plants.They cover only about 5000 km 2 of this immense desert, but a recent checklist of vegetation in the lomas formations registered 847 plant species from 83 families (Dillon & al., 2011).Some are species found in other deserts of North America, but others are endemics sometimes occurring in only a single locality.These plants are potential substrates for myxomycetes, providing rich and varied microhabitats in which these organisms can complete their development.
The study area covers the coastal areas in the administrative Departments of Tumbes, Piura, Lambayeque, Figs.2-16.Landscapes and some vegetation types of the coastal desert of Peru.2-3.Trees such as Prosopis sp. and cacti in the sandy soils of Tumbes.4-5.Flat sandy plains and active dunes in the Sechura Desert.6. Hills near the coastal areas.7. Dunes stabilised by species of the genus Capparis.8. Neoraimondia sp.near Lima.9-10.Dunes with islands of terrestrial bromeliads of the genus Tillandsia.11.Arid valleys.12. Haageocereus sp.near Moche, La Libertad.13.Fog retained by the hills in the area of lomas de Atiquipa.14. Vegetation influenced by the "garua" in the lomas de Atiquipa.15.Haageocereus sp. in Sama Grande, Tacna.16.Population of Neoraimondia aticensis near the coast of Sama, Tacna.
Cajamarca, La Libertad, Ancash, Lima, Ica, Arequipa, Moquegua and Tacna (Figs. 1,[2][3][4][5][6][7][8][9][10][11][12][13][14][15][16].The abundance of vegetation in Tumbes and the presence of trees (Figs.2-3) are the result of greater precipitation due to the influence of El Niño and its proximity to the tropics as mentioned above.In Piura the Sechura Desert, with extensive areas of flat sandy plains and active dunes , is where the coastal desert spreads out to its widest and can reach up to a width of 100 km.The dunes are stabilised by species of the genus Capparis (Fig. 7), plants that have proved to be most productive substrates for myxomycetes.Parts of the desert along courses of rivers that flow to the Pacific Ocean in Lambayeque, Cajamarca and La Libertad, have been transformed into agricultural land and planted with crops like sugar cane, asparagus or orchards of fruit trees.On the lower sandy slopes, below 200 m, a mixture of cacti (Espostoa, Haageocereus, Melocactus) and terrestrial bromeliads of the genus Tillandsia can be found 12,(15)(16).On the hills and stony mountain slopes, under the influence of the "garuas" (Figs.6,[13][14], a number of endemic species of plants can be found (Leiva González & al., 2014).In the central zone (Ancash, Lima, Ica, North of Arequipa) the coast and land below 100 m is devoid of vegetation, and from 100-300 m a community of cryptogams predominates with species of Nostoc and foliose and fruticose lichens (Fig. 11).Around 300 m herbaceous and woody plants appear and above 300 m, the dry rocky slopes support stands of Tillandsia latifolia and Puya ferruginea.Most of the original vegetation around Lima has disappeared under the significant human pressure on the land.In Ica there are steep coastal ridges with a series of marine terraces.The fierce coastal winds generate enormous dunes that are stabilised in some places only by the growth of species of Tillandsia.The lomas de Atiquipa , one of the best-preserved examples (Gómez-Sosa, 1986) and one of the richest in species, was among the areas studied in Arequipa.In the southern zone (South of Arequipa, Moquegua and Tacna) the vegetation contains sparsely distributed shrubs and herbaceous species and a population of Neoraimondia aticensis occurs near the ocean (Fig. 16).Large colonies of Tillandsia purpurea cover the sand dunes (Dillon, 1997).

MATERIAL AND METHODS
Sampling was carried out in 96 localities (Table 1), between 0 and 1500 m elevation of which 83 yielded identifiable myxomycetes.Work was carried out in three expeditions in September to October 2012, February to March 2013 and in May 2014.An additional site is included from a different elevation (PER-12-23), as the new species was isolated from substrate material found there, in addition to the two coastal collections.Localities were selected at intervals of approximately 25-50 km.At each site, the microhabitats in which myxomycetes are known or suspected to occur were examined carefully, and were georeferenced with a portable GPS unit (Garmin eTrex Vista HCX, Datum WGS84).Samples were collected in the field and substrate samples were also removed for moist chamber culture.Methods used for collecting myxomycetes in the field and substrates for laboratory culture can be found in Stephenson (1989), and Rossman & al. (1998).
All the fieldwork was carried out by A. Estrada-Torres, C. Lado and D. Wrigley de Basanta, with the help of S.L. Stephenson, A. Rollins, G. Rebaza and I. Treviño in 2012, with J. Garcia and A. Cano in 2013, and with J. Rojas-Fox, J. Molina and S. Castillo in 2014.Four investigators collected at any one time with approximately one hour spent in each collecting locality.
Moist chamber cultures were prepared using sterile disposable plastic Petri dishes (9 cm diameter), in the manner described by Wrigley de Basanta & al. (2009).The pH of each culture was determined with a portable pH meter after 24 hours, and then excess water in each dish was poured off.Cultures were maintained at room temperature (21-25 °C) in diffuse daylight and examined at regular intervals with a dissecting microscope for a period of up to three months.As the myxomycetes matured, the portion of the substrate upon which they occurred was removed from the moist chamber culture, allowed to dry slowly in a closed empty Petri dish and then glued in a small cardboard box.All sporophores of a given species that developed in the same culture were considered to represent a single record.
The completeness of the sampling effort for the studied region was evaluated using the ACE and the CHAO1 abundance índices (Colwell & Coddington 1994;Colwell & al., 2004) and the accumulation curve adjusted according to the Clench function where where S n is the number of species accumulated for a unit of collecting effort (n) (Jimenez-Valverde & Hortal, 2003).Each collecting site was considered as the unit of collecting effort, using the total number of species found with the program EstimateS v 9.1.0(Colwell, 2013), adjusted so that 3 was the upper abundance limit for rare or infrequent species, following the criteria of Stephenson & al. (1993).The latter authors considered species with a relative abundance of ≤0.05 to be rare.With the same program the richness of species with 1.5 times more sampling effort was calculated.The adjustment according to the Clench function was carried out with the program Statistica v 12, using the Simplex and Quasi-Newton method of adjustment ( Jimenez-Valverde & Hortal, 2003).ACE and CHAO1 abundance indices were calculated for all results, only field results and only moist chamber culture results.
To measure the complementarity of the assemblage of myxomycetes in the studied area and other regions in South America, the formula proposed by Colwell & Coddington (1994) was used: C jk =U jk /S jk where U jk =S j +S k -(2V jk ) and S jk =S j +S k -V jk where U jk is the number of species that are different for the two communities, S jk is the total richness of species for both communities, S j is the number of species from the first community, S k is the number of species from the second and V jk is the number of species common to both.
To determine whether the proportion of collections and species from the different orders of myxomycete was dependent on the sampling method used, a test of independence was done using a contingency table and a X 2 test (Zar, 1976).
All specimens are deposited in the MA-Fungi herbarium (sub Lado) with duplicates in TLXM (sub aet), UARK (sub sls) or in the private collection of Wrigley de Basanta (dwb).

RESULTS
As a result of the survey carried out in three consecutive years in these hyperarid areas, 723 collections of myxomycetes were identified, either specimens that had developed in the field under natural conditions or those that were recovered from moist chamber cultures.In total, 78 taxa representing 23 genera of myxomycetes were recorded.

Annotated list of species
Myxomycete collections from this survey are arranged alphabetically in the list that follows by genus and then species.Nomenclature follows Lado (2005Lado ( -2015) ) unless otherwise stated.Information is provided on the source of each record (either a field collection or a collection obtained from a moist chamber [mc] culture), the pH of the culture in which the specimen appeared, the substrate upon which it was collected and the locality from which it was collected (see Table 1).All identified collections are included.Species of particular interest have additional comments.The abbreviation "cf." in the name of a taxon indicates that the specimen representing the source of the record could not be identified with certainty.This sometimes indicates scanty or aberrant material.Comments on the distribution of the species in the Neotropics are based on Lado & Wrigley de Basanta ( 2008), but the information on the area of study has been updated with the added references.New records for the Neotropics, South America or Peru are marked with ( †), (°) or an asterisk (*) respectively.
PER-14-06: Prosopis sp.bark (mc, pH 4.81), dwb 3695.This is a small collection of very typical sporocarps.Originally described on bark from Norway, this is the first record of the species for South America.In the Neotropics it has been previously reported from Mexico.In Peru previously reported by Rojas & al. (2011) from Madre de Dios.Otherwise reported from many countries in the Neotropics.Collections sls 28365, 28367, 28360 and 28368 were characterized by a more globose sporotheca and a more reduced capillitium than those typical of this species, but otherwise they fit the published descriptions of Comatricha laxa.
In Peru previously reported by Rojas & al. (2011) from Madre de Dios.
A widely distributed species in South America, but in Peru previously reported only by Rojas & al. (2011) from Madre de Dios.The collections Lado 23524 and Lado 23525 are the var.scyphoides not previously reported from the country.
A widely distributed species in the Neotropics but in Peru previously reported only by Rojas & al. (2011)  This species is widely distributed throughout the Neotropics.
Anales del Jardín Botánico de Madrid 73(1): e032 2016 In the Neotropics the species is only known from Mexico.This collection has groups of 3-8 sporocarps.They have dark brown spores 17-20 µm diam, that are paler on one side and ornamented with the characteristic prominent, sharp spines of 2-3 µm in length.
Widely distributed throughout the Neotropics.In Peru previously reported only from Madre de Dios (Rojas & al., 2011).

Didymium anellus Morgan
The Peruvian collection fits the description of the species given by Poulain & al. (2011) except that in the plasmodiocarps no columella is evident.In the sporocarpic forms however a pulvinate calcareous base is present.The spores are 12.6-13.4μm diam and the capillitium is typical for the species, hyaline and forming a complete net.In South America it has only been recorded previously from Argentina (Moreno & al., 2012).
PER-12-32: Tillandsia sp.inflorescence (mc, pH 6.9), aet 12232, (mc, pH 7.3), aet 12235a, (mc, pH 5.6), aet 12257, (mc, pH 6.9), aet 12283.PER-12-33: Tillandsia purpurea inflorescence (mc, pH 7.2), aet 12278, (mc, pH 7.0), aet 12279.PER-12-34: Tillandsia purpurea inflorescence (mc, pH 7.3), aet 12226, (mc, pH 7.0), aet 12228, (mc, pH 7.3), aet 12251.PER-13-10: Puya sp.inflorescence (mc, pH 5.3), aet 13095.Collections aet 12226, 12228, 12232, 12235a, 12279 and 13095 have flattened sporothecae with a non-calcareous base, no columella, a hyaline peridium and spores that measure 7-7-11.3μm, with groups of warts.These characters agree with those noted by Poulain & al. (2011) for Didymium applanatum except for the stalks, which are variable from calcareous to partially or completely without lime and dark in colour, filled with refuse matter and so looking macroscopically like Didymium clavus.Collections aet 12251, 12258 and 12278 have smaller hemispherical sporothecae, but the bases are not umbilicate and they have no columella.Their spore size overlaps with the previous samples, the capilitium is similar and the stalks are bi-coloured with a calcareous apex and darker base, like some of those above.In collection aet 12283 both forms were mixed so it was considered that they probably represent extremes of the same species.Didymium applanatum has recently been considered to be a synonym of D. squamulosum var.claviforme Sturgis (Oltra, 2003), a very variable species.However given that the characters of the Peruvian collections are fairly constant with flattened sporothecae, non umbilicate bases and without a columella we prefer to keep the species separate for the moment until further evidence like molecular data combine or separate these forms.
In the Neotropics it was only previously recorded in Mexico and Ecuador.
Originally described from Brazil (Gottsberger, 1968), it is widely distributed throughout the Neotropics.
A common species in the Neotropics, but in Peru previously reported only from Madre de Dios (Rojas & al., 2011).This species has only been reported previously from Brazil.These specimens show the typical hollow clubshaped columella, (Figs.17-18), filled with lime crystals.The stalk is darker and filled with refuse matter below, and the external surface covered with lime crystals, that are less abundant towards the base.The spores by SEM are densely warted and some warts are fused in a short subreticulate pattern (Fig. 19).
Widely distributed throughout the Neotropics.In Peru previously reported only from Madre de Dios (Rojas & al., 2011).Specimen aet 12239, isolated from cactus flowers, has large flat branched or ring-shaped plasmodiocarps with an egg-shell like outer layer that is yellowish.Microscopically it has the typical characters of D. difforme, spores 11.8-13.0μm, with a paler area.It is possibly an ecotype of the species but has been left as cf.

Didymium peruvianum
Habitat.On dead twigs, aerial and ground plant litter.Distribution.Known only from the South and Central Peru (Ayacucho and Arequipa regions), from sea level to almost 4000 m.
The principal characters that distinguish this minute species are the brown spores with broad pale bands  and groups of more prominent pigmented warts, and the eggshell-like outer peridium .Other Didymium species covered with an eggshell-like layer of packed lime crystals include D. annulisporum H.W. Keller & Schokn., D. trachysporum G. Lister, D. listeri Massee, D. rugulosporum Kowalski, D. quitense (Pat.)Torrend and D. difforme (Pers.)Gray.The first three have smaller spores from 8-10 µm diam.vs 12.5-14 µm in the new species 36).Didymium rugulosporum has larger spores (18-20 µm) and a dense rigid capillitium forming a wide meshed reticulum whereas D. peruvianum has scanty capillitium, scarcely branched and with few cross connections (Fig. 27) Didymium quitense has very different spores with large warts forming a network of irregular muri by SEM (Lado & al., 2011).Didymium difforme also has different spores, pale on one side and densely covered with very fine warts (Nannenga-Bremekamp, 1975;Neubert & al., 1995: 340).Didymium nullifilum (Kowalski) M.L. Farr has densely compacted lime scales on the peridium as well as smaller spores (8-10 µm diam.vs 12.5-14 µm) and the spores have scattered prominent spines (Kowalski, 1972).Short stalked species that could be confused with D. peruvianum include D. umbilicatum, D. mexicanum and D. nigrisporum but they all have calcareous stalks not very short and brownish stalks or merely a restricted base to the sporotheca (Wrigley de Basanta & al., 2008b;Moreno & al., 1997;Nannenga-Bremekamp & al., 1984).The first two are also umbilicate above, but both have a single membranous peridium and scattered lime crystals, and are not covered with a continuous layer of eggshell like lime.They also have different spores.Didymium nigrisporum Nann.-Bremek., K.G.Mukerji & Pasricha also has very dark, strongly warted spores and not broad pale bands between the more prominent warts.The colour of the sporocarps in D. inconspicuum Nann.-Bremek.& D.W. Mitch.and D. saturnus H.W. Keller differentiate these from the new species since the former is ochraceous to fawn (Nannenga-Bremekamp, 1989) and the latter straw yellow to silvery gray or brown (Keller, 1970).In addition both have abundant capillitium and D. saturnus has spores with a prominent ring.Some other small, sessile Didymium species are D. comatum (Lister) Nann.-Bremek., D. circumscissile K.D. Whitney & L.S. Olive and D. atrichum Henney & alexop.In D. comatum the capillitium is profuse, expanding elastically when the lid comes off, not scanty like in D. peruvianum and the spores are violaceous, not brown, and paler on one side, minutely warted, and the warts are arranged in rows, sometimes forming a small-meshed reticulum (Nannenga-Bremekamp, 1966).The shape of the sporocarps in D. circumscissile is different from the new species as they are globose, obovate or turbinate, occasionally reniform, not discoid.The distribution of lime is also different as the sporotheca is lightly sprinkled with crystals, especially on the operculum (Whitney and Olive, 1983), whereas in D. peruvianum it is a layer like an eggshell.In D. atrichum capillitium is totally lacking and it has smaller spores that have spinulose ornamentation arranged in a reticulate lacey pattern (Henney & al., 1980), not with broad pale bands between more prominent warts.
Didymium peruvianum was collected in the field, where it fruited under natural conditions, and was also isolated from plant litter put into moist chamber culture.The collections are from two different regions in Peru and at very different elevations.The combination of characters do not match any described species of Didymium and so we describe it here as a new species.
This species, described originally from Quito (Ecuador) as Chondrioderma quitense Pat.(Patouillard & Lagerheim, 1895) has also been reported from Brazil (Farr, 1968), Argentina (Lado & al., 2011(Lado & al., , 2014) ) and Chile (Lado & al., 2013).This is the first record for Peru.†Didymium cf.rugulosporum Kowalski PER-12-47: Tillandsia purpurea inflorescence (mc, pH 6.8), aet 12264.This is a new record for the Neotropics.The Peruvian material is macroscopically similar to D. difforme but it has spores with prominent warts sometimes connected in lines and a wide mesh reticulum.This is a character of D. rugulosporum, a species described by Kowalski (1969) from the United States and known so far only from the type locality.The spores of the Peruvian material however measure 11.8-15.0μm diam, while Kowalski describes the US material as having spores of 18-22 μm diam, and so the collection has been left as cf.Widely distributed throughout the Neotropics.In Peru previously reported from Loreto (Wrigley de Basanta & al., 2008a).The spores of the collection Lado 22547 are illustrated by  and are typical in form (globose) and size (10-11 µm diam).The collection Lado 23526 has spores that fit the dimensions of the species, 10-11 µm diam, but are densely warted and slightly angular when observed by SEM (Fig. 21).This ornamentation, has also been reported by Nannenga-Bremekamp (1991) for specimens from temperate zones, and could be an ecotype of this variable species.The delimitation of some of these collections with D. nigrisporum was difficult but the dimension of the spores 10-11(-12) µm diam vs 12-15(-16) µm diam in D. nigrisporum  was the main character used to separate them.A species widely distributed throughout the Neotropics and usually associated with succulent plants.In Peru previously reported from Cuzco (Rojas & al., 2011)  purpurea litter (mc, pH 6.13), dwb 3538; Tillandsia purpurea litter (mc, pH 5.20), dwb 3563; Tillandsia purpurea litter (mc, pH 6.41), dwb 3570.This species is usually associated with xeric environments often developing on the remains of cacti and other desert plants.These collections are typical.The sporothecae are covered with pale yellow to white lime crystals, sporocarps have short calcareous stalks, and spores 8-10 µm densely and uniformly warted.In the Neotropics previously reported from Mexico (Lado & al., 2007;Estrada-Torres & al., 2009;Esqueda & al., 2011;Wrigley de Basanta & al., 2015), Argentina (Lado & al., 2011) and Chile (Lado & al., 2013).These are the first records for Peru.
In South America it was only known from Ecuador (McHugh, 2005) and Argentina (Lado & al., 2014).
Widely distributed throughout the Neotropics, but not reported previously from Peru.
Widely distributed throughout the Neotropics.These collections were found in the North of the country, in dunes fixed by Capparis scabrida.The spores by SEM (Fig. 39) show the chracteristic ornamentation of warts connected by minute crests in a reticulate pattern.
Widely distributed throughout the Neotropics, but not reported previously from Peru.
Widely distributed throughout the Neotropics.In Peru previously reported from Madre de Dios (Rojas & al., 2011).
These collections on the same substrate, but from two separate years and different places, show a partially chequered peridium.The small, scattered, sessile sporocarps are less than 0.5 mm diam.The yellowish peridium breaks up into irregular sized platelets.The capillitium is scanty, the threads of 2-3 µm diam., irregular, with constrictions and ornamented with warts.The spores however are larger than those described by Lister (1931), from 12-14(-15) µm diam.vs 9-10 µm diam., yellow, and thick walled but with a thinner area to the wall.The sporocarp colour is also lighter than the description of P. tesselata, yellowish brown not purplish black.The Peruvian material is apparently the closest to this species, but only further collections can confirm if it is the same.In the Neotropics it was only reported from El Salvador (Rojas & al., 2013), and is otherwise a rare species.If confirmed these collections would be the first records of the species in South America.
Widely distributed throughout the Neotropics.In Peru previously reported from Loreto (Wrigley de Basanta & al., 2008a).The moist chamber collections of this species on Capparis scabrida were very extensive, sometimes covering almost all the substrate in a 9 cm culture dish.Widely distributed throughout the Neotropics.In Peru previously reported from Madre de Dios (Rojas & al., 2011).

*Physarum atacamense D. Wrigley, Lado & Estrada
Only previously known from the Atacama Desert in Chile (Wrigley de Basanta & al., 2012).These records expand the distribution of the species to the coastal arid lands of Peru, where it was almost the most frequently recorded species, 135 collections.The collections were extensive both in the field and in moist chamber cultures.From these numerous collections, the description of the species can be broadened to include some sessile or very short stalked sporocarps, the stalk in some cases is flat, entirely yellow, not darker at the base, longitudinally striate and common to several sporocarps.*Physarum bitectum G. Lister PER-12-19: aerial twigs (mc, pH 5.52), sls 29170.
The single collection was typical in all respects.
*Physarum cinereum (Batsch) Pers.Second record for South America, previously reported only from Chile (Lado & al., 2013).Again this is a rare species with a restricted distribution that is appears abundantly in the Peruvian coastal desert.Macroscopically, the Peruvian specimens are almost indistinguishable from P. bitectum, but detailed observation of the spores reveals disperse prominent spines in P. clavisporum  and uniformly distributed warts in P. bitectum.Our collections agree, in general, with the original description of the species (Moreno & al., 2009), since the spores have the typical dark purple brown colour with a paler area, bearing dispersed and prominent spines, and the spore surface covered by tiny densely distributed warts visible only by .But the spines of the spores by SEM, in our collections, are formed of spines or bacula  instead of the pila with widened to slightly coralloid apex of the type material.In addition, in the Peruvian material a gradation in the quantity and density of the spines, from few and dispersed  to abundant and regularly distributed (Figs.49-50) can be observed, differentiating this material from that described by Moreno & al. (2009), and for that reason the collections have been left as cf.They may indicate an ecotype or variety of the species, but further investigation is needed to confirm or refute this.
Widely distributed throughout the Neotropics.In Peru previously reported from Loreto (Wrigley de Basanta & al., 2008a).Collection sls 31148 has been left as cf.since the collection was poorly developed, but it appeared to fit this species in most respects.
Widely distributed throughout the Neotropics.In Peru only reported from Loreto (Stephenson & Mitchell, 1994).
Reported previously from Loreto, Peru by Wrigley de Basanta & al. (2008) on lianas.Otherwise only reported in the Neotropics from Costa Rica by Rojas & al. (2010).This is a rare species that appeared frequently in the moist chamber cultures of Tillandsia remains.The sporocarp, the stalk and details of the capillitial net are illustrated for the first time by .The stalk is clearly formed of concentric layers making a tube , not of fibers.The spores by LM are defined as warted-reticulate, by SEM (Figs. 62-69) they show a well defined net and a complex ornamentation of warts connected in rows in a pattern of a "simple reticulate type with perforated muri" according to the terminology of Rammeloo (1975), and, in some cases, the pillars supporting the muri surpass the top of the muri (Fig. 67), while in other the muri appear more solid .
Distributed throughout the Neotropics, but not previously reported from Peru.
Not previously reported from South America.Only one collection of grouped and dispersed but well developed sporocarps, up to 2 mm tall, and cylindrical.The capillitial threads are undulate, without membranous expansions, forming a broken surface net.The spores are pale brown by TL, with a paler area corresponding to the area of germination, and minutely warted, these characters agree with the description of Nannenga-Bremekamp (1975).However, the spores in the Peruvian specimens are larger, 7-9(-10) µm diam vs 5-7 µm diam as described by Nannenga-Bremekamp (1975), and by SEM they are densely warted with some of the warts fused or connected by a tenuous basal "muri" .
Distributed throughout the Neotropics.In Peru only previously known from a record by Farr (1976)  In South America previously reported only from Brazil (Cavalcanti, 2010).The three collectios were obtained on the same typical substrate of the species, Agave americana, and the characters agree with the descriptions available in the literature.
Widely distributed throughout the Neotropics.In Peru only reported from Loreto (Wrigley de Basanta & al., 2008a).

DISCUSSION
As stated above, the results of the survey of these hyperarid habitats yielded 723 collections of myxomycetes representing 78 species in 23 genera in all six orders recognised in this group.These include one species that is new to science, another 2 species that have not been reported before from the Neotropics, a further 4 new records for South America and 41 additional species reported for the first time in Peru.This increases the number of myxomycetes recorded from the country so far by over 50%, to 128 species.
The known distribution of many species has been extended by these results, which also fill a prominent gap in the knowledge of these microorganisms in a virtually unexplored biogeographical area.These data complement work carried out in other South American deserts such as the Atacama, Chile (Lado & al., 2007(Lado & al., , 2013) ) and Monte, Argentina (Lado & al., 2011) and help to develop an idea of the true myxobiota of the Neotropics.It is becoming increasingly evident that both warm and cold deserts are reservoirs of surprising myxomycete biodiversity, as results from the Patagonian steppe (Lado & al., 2014) and work in other arid areas of Europe and Asia (Schnittler, 2001;Novozhilov & al., 2006;Schnittler & al., 2013) also indicate.
The predominant order among the results was the Physarales with 44 species, 56% of the recovered taxa.This was followed by the Trichiales, with 14 species, and the Stemonitales, with 10 species.The orders with the lowest number of species were the Liceales (5 species), Echinosteliales (4 species) and Ceratiomyxales with only one species.The predominance of the Physarales in arid zones in the Americas has previously been documented by surveys carried out in the Valle de Tehuacán-Cuicatlán, Mexico (Estrada-Torres & al., 2009) and Monte Desert, Argentina (Lado & al., 2011), and has also been registered in the cold Central Asian deserts in Kazakhstan, Mongolia and Caspian lowland by Schnittler (2001) and Novozhilov & al. (2006).The prevalence of this order is even more obvious when the total number of collections is taken into consideration.Almost 79% ( 566  considered to be particulary abundant (>6% relative abundance) and the most representative of this Peruvian coastal desert since they make up 53.5% of all collections.At the other extreme 41% of the species studied were represented by only a single collection.
Of the 78 species identified from this survey, 33 were only collected in the field and 20 were obtained only from moist chamber culture while 25 were recovered from both.In the field collections, 70% of the specimens found were represented by 5 species, whereas in moist chamber cultures this percentage was made up of 11 species.The six additional species, abundant in moist chamber culture were rarely collected in the field or not at all.Both techniques returned the same three most dominant species.The two methods of survey are evidently complementary to each other and give a more accurate account of the myxobiota actually present in an area, especially in arid places like the study area where environmental extremes can affect the detection of many species in the field.This is reinforced by the tests of independence done on the proportion of collections (Table 2) and species (Table 3) recovered from the field or from moist chamber culture.For the proportion of collections from different orders X 2 =68.63 (df=2; p<0.0001), indicating that the number of collections of different groups of myxomycetes depends on whether they were collected in the field or from moist chamber culture.For the proportion of species from different orders, however, X 2 =1.32 (df=2; p=0.5177), which indicates that this is independent of the method used to collect the species (field or culture).
The fact that the Physarales were also the predominant order considering only the field collections, since the five most common species were from this order, could be due to various factors.One of these could be that this group may have strategies to complete their development, even under such extreme conditions, that the other orders of myxomycetes do not possess.It is also possible that the reproductive structures of the Physarales are more resistant to destruction by biotic and abiotic elements in arid zones and so remain for longer periods in situ.The first of these hypotheses is supported in this study by the fact that even in moist chamber culture, where available moisture is a constant, the Physarales were the dominant group, and in addition the moisture and favourable conditions permitted the development of a larger number of specimens from other orders, the Echinosteliales, Liceales and Stemonitales.In the field, under natural conditions, species from the latter orders would be opportunistic, developing only when sporadic moisture was available in these dry environments of the Peruvian coast.
The most productive substrates were leaves and other dead parts of species of the genus Capparis with 158 collections from 14 myxomycete species.The species of this plant genus, as noted above, stabilise the dunes in the coastal desert and were particularly common in the more northerly areas (Fig. 7).The species of Tillandsia, found sometimes as isolated islands of vegetation sprouting directly out of the sand dunes (Figs.9-10), produced 86 collections of myxomycete (12%) but these represented 20 species, and the remains of species of Haageocereus were the third most productive substrate genus with 57 collections.However if the remains of all the species of cacti are taken together they accounted for 23% (167) of the collections but representing only 15 species of myxomycete, and so accounting for considerably less diversity than the Tillandsias.The locality that produced the greatest number of myxomycetes with 65 collections was Yauyos (PER-12-34) in the administrative department of Lima at 13°S latitude and an elevation of 400 m.The greatest variety (number of species) was collected from Pariacoto in Ancash, 9°S latitude and at 1478 m (locality PER-13-32) where 18 different species were recorded.More than half the collections were produced by only 14 localities (375 collections).The localities from South latitude 11° to South latitude 17° had a greater number of collections than South latitudes 3° to 9°, when corrected for the number of collecting localities sampled, which is surprising, given the proximity of the latter to the tropical areas and the subsequent increased variety in vegetation.However the elevations where the greatest number of collections were made were between 400 and 600 m, when corrected for the number of localities sampled at each elevation.This suggests a reason for the number and variety of myxomycetes at the above localities, since at these latitudes and elevations the coastal fogs or garuas constantly bathe the slopes providing more moisture for the development of these microorganisms .This moisture is responsible for the great variety of plants now known from the lomas formations in these areas, known by some authors as "fog oases" (Dillon, 1997).These pockets of vegetation enable the myxomycetes to exploit the multiple microhabitats available to complete their life cycles in such an otherwise hostile environment of barren desert, and are most important areas of conservation.
The substrate pH in moist chamber cultures has been shown to influence the number and variety of myxomycetes isolated from a particular substrate (Wrigley de Basanta, 2004).In these cultures the range of substrate pH of the 225 moist chamber collections was from 4.39 to 8.13, but more than half the collections were obtained from cultures with circumneutral pH (Fig. 70).This has been found in other surveys (Wrigley de Basanta & al., 2008a;Estrada Torres & al., 2009;Lado & al., 2011).
According to the estimators CHAO1 and ACE, the expected species richness if the sampling effort had been exhaustive would be 119 and 114 respectively (Fig. 71).The estimated number of species based on the Clench function, with an r 2 =0.996, was 124, indicating a good fit of the  data to the curve (Fig. 72) and giving a value slightly higher than that estimated using CHAO 1.This means that in this survey almost 70% (63% -69%) of all possible species in the sampling area were recorded, a very representative result considering the extreme conditions of the coastal desert.Extrapolation of the species richness using the program EstimateS showed that if 49 more collecting localities had been visited this would have returned only 17 species more.
The Clench function suggests that 102 more collecting sites would be necessary to achieve these 17 more species, with the enormous effort in both the field and laboratory work implied.In fact double the time of 3 weeks a year for three years invested in sampling, and the consequent 3 years of laboratory work, would be required.This is because of the large number of rare species among the myxobiota of this region.
Analysis of the results of field collections alone using the above estimators showed that an approximate 63% of the species theoretically possible were recorded.Both CHAO1 and ACE showed that 66% and 73% of possible species were recovered using moist chamber cultures, a slightly more efficient result probably on account of the more stable favourable conditions that existed in the culture.This productivity is similar to that of other surveys (Lado & al., 2011(Lado & al., , 2013(Lado & al., , 2014) ) and surprisingly high considering the arid environment and the multiple reasons why all the myxomycetes actually present do not fruit normally and produce identifiable specimens.The exact phenological optimum varies among species and will not always coincide with the moment of sampling in the field.In culture many sclerotia and probably microcysts never developed further, and some abnormal fruitings were found that were impossible to identify.
The results are also surprising, given the vastness of the area studied, the isolation of many of the groups of plants that act as substrates for the development of these organisms, in pockets or islands, and the fact that even in spite of these factors, from 86% of the localities visited identifiable myxomycetes were recovered.
The biotic distinctness, or complementarity, of the assemblage of myxomycetes in the study area compared to that of other arid areas in South America is shown in Table 4.
Not surprisingly the lowest values, or most similar communities are the closest arid areas in North Chile and Argentina.All three belong to the South American Transition Zone, according to Morrone (2004).However, in both cases there is an approximate 70% difference in the species present, explainable by the peculiarity of the myxobiota of the Peruvian coast.Rarer species such as Didymium nigrisporum, Didymium anomalum, Didymium cf.rugulosporum, Licea bulbosa and Stemonaria gracilis are unique to the Peruvian biota, as is the newly described species, Didymium peruvianum.Some species are even shared with those found in North American arid areas such as Diderma acanthosporum and Perichaena luteola (Estrada-Torres & al., 2009).However the shared species between the three South American deserts include Echinostelium arboreum, Licea succulenticola or Physarum licheniforme (Lado & al., 2011(Lado & al., , 2013)), and species hitherto known only from South American deserts such as Licea eremophila and Physarum atacamense (Wrigley de Basanta & al., 2010a;2012) (Lado & al., 2013), that were not present in the coastal desert of Peru.The Patagonian steppe and southern Chile are areas with almost 80% of species distinct from the coastal desert of Peru, reflecting the geographical distance and marked ecological differences between the two regions.It is perhaps important to point out that the moderately high and high complementarity between the coastal desert of Peru and other South American arid areas reinforces once again the idea that each biogeographical area has its own characteristic myxobiota, as has already been pointed out in other American studies (Estrada-Torres & al., 2009;Wrigley de Basanta & al., 2010b;Lado & al., 2011Lado & al., , 2013)).These patterns of distribution may also result from the geographic barrier formed by the Andes and the corridors of the interandean valleys that permit interchange of some but not all biota.
In summary, the present paper represents the first systematic study of the myxomycetes from this arid biogeographical area, previously virtually unexplored for these microorganisms.The results obtained support the fact that  the coastal desert of Peru is surprisingly rich in myxomycetes, from each taxonomic order of the group, and that even within South America, each arid area has a unique assemblage of the organisms, although there are species common to all.The results also indicate that the myxomycetes are a normal component of the Peruvian desert flora, with an ecological role that is yet to be fully determined, since they have been isolated from the typical vegetation of the area including endemic plants.Although cacti, as is customary in drylands of the Americas, were important substrates with a high yield of myxomycete collections, the two plant genera that proved to be the most productive were the genus Capparis and the genus Tillandsia.These are important components of the lomas formations, essential for stabilising the sand dunes, and appear to be a natural habitat for myxomycetes.These were even more abundant and varied in areas under the influence of coastal fogs.In spite of the vast and varied distances between the lomas formations that make up much of the vegetation in this study, and the extreme aridity of some places, myxomycetes were found in 87% of the locations sampled.This raises questions as to the methods of dispersal and distribution of the organisms over such disjunct geographically isolated formations, designated unique delimited ecosystems by Dillon & al., (2011).These data not only refute the paradigm that the myxomycetes are associated mainly with humid temperate and tropical ecosystems but also suggest that these microorganisms are important elements of biodiversity and should certainly be considered in establishing conservation priorities in these fragile ecosystems.

Fig. 70 .Fig. 71 .
Fig. 70.Number of collections and species of myxomycete at each different substrate pH.
. The striking difference between the species found in Central Chile and the study area, where 87% of the species are different, could be due to the clearly Mediterranean vegetation of the former favouring the development of foliicolous species such as Didymium chilense Estrada, Lado & D. Wrigley, Didymium comatum (Lister) Nann.-Bremek., Didymium eximium Peck, Didymium laxifilum G. Lister & J. Ross and Physarum newtonii T. Macbr.

Table 2 .
Comparison of sampling methods for collections.

Table 3 .
Comparison of sampling methods for species.

Table 4 .
Percentage complementarity of the assemblage of myxomycetes from the Peruvian coastal desert with other arid areas in South America.