Thank you for helping generate most GBIF records for most species since 2020

GBIF (the Global Biodiversity Information Facility) is an online archive that brings together global biodiversity data from over 81,000 datasets by thousands of publishers. These publishers include natural history museums like the Smithsonian museum and also citizen science sites like iNaturalist. GBIF provides essential monitoring capacity to inform biodiversity conservation. Over time, iNaturalist has become an increasingly significant contributor to GBIF. In fact, since 2020, most GBIF records for most species have come from iNaturalist.

Only about 8% of the GBIF records since 2020 are from iNaturalist, so how can this be true? The answer is that most GBIF records represent a relatively small number of species (records of birds from popular bird-focused citizen science publishers such as eBird and records of common European species from popular European-focused citizen science publishers such as Artportalen). iNaturalist is unique as a GBIF publisher because it has generated tens of millions of records distributed across hundreds of thousands of species. This is important from a conservation perspective because unless we rely on a relatively small number of species such as birds (~11k species globally) as proxies for global biodiversity, reducing extinction rates is going to require insights across a large portion of the ~2M named species.

Selecting 2020 as a cutoff is arbitrary, but biodiversity records that can be generated and mobilized on short time scales (e.g. years rather than decades) are particularly important from a conservation perspective because they give scientists a real-time picture of the status of biodiversity in response to rapidly changing in climate and land use.

The graph below shows all GBIF records from 2020 pooled by species and arranged in descending order by number of species. On the left hand side, American Robin (Turdus migratorius) has more GBIF records since 2020 than any other species. While the GBIF records since 2020 span 325,558 species, the distribution is very unbalanced. Even plotting the y-axis on a log scale, the number of observations per species drops off rapidly. Most GBIF observations are from a relatively small number of species. The moth Pidorus atratus which is positioned near 30,000th place for most records, has just 292 records. Near the 100,000th place, the plant Stigmaphyllon ellipticum has just 16 records on GBIF. This is out of 2M named species and likely many more waiting to be discovered.

The next graph is nearly identical except that the GBIF records are split into two groups: those published by iNaturalist in green and those from all other publishers in black. While both lines are unbalanced, notice that the green iNaturalist line is less unbalanced and crosses the black line around the 22,700th place for most records. These first ~20k species with the most records are mostly birds and common European species with many tens of thousands of records from publishers to GBIF such as eBird (a birds-only, global citizen science site) and Artportalen (an all-taxa, Sweden-only citizen science site). But for most species, most of the data on GBIF is from iNaturalist.

The next graph displays this in another way by plotting these GBIF records since 2020 as the percent of species versus the percent of records published by iNaturalist. 42% of the species have 100% of their records from iNaturalist. 58% of species have at least 58% of their records from iNaturalist. In other words, since 2020, most GBIF records for most species have come from iNaturalist.

There is likely error in these graphs from taxonomic and identification issues among the various GBIF datasets, but its difficult to know how this error would influence these statistics. The bias towards a relatively small number of well-studied species like birds in GBIF is striking. Birds represent about 0.6% of named species, but represent about 83% of all GBIF records since 2020. The iNaturalist GBIF subset since 2020 is still skewed towards birds (21%), but also has a large proportion of records from the most speciose groups: insects (22%), plants (39%) and other species (10%). Excluding iNaturalist, over 88% of GBIF records since 2020 are of birds.

I mentioned that only 8% of the GBIF records since 2020 are from iNaturalist. This is mostly because of the tremendous number of bird records coming to GBIF through platforms such as eBird. The graphs below show the relatively small bird contributions of iNaturalist to GBIF since 2020 relative to other publishers (including eBird). The data from the top 2 graphs below are identical to those above for all GBIF records since 2020 except the graphs in the left columns have log scale x-axis. These log-log graphs are more confusing to understand but they reveal differences in the lines more clearly. The bottom two graphs show just birds. Note that only 10% of bird species have at least 10% of records from iNaturalist (due to the data volume of other publishers like eBird) compared to the equivalent 58% statistic across all species.

The graphs for other terrestrial vertebrates are very different from birds. Across all GBIF records since 2020, 63% of mammal species have at least 63% of records from iNaturalist; 81% of reptile species have at least 81% of records from iNaturalist, and 73% of amphibian species have at least 73% of records from iNaturalist.
.

The graphs for fish and the two most speciose groups (insects and plants) are also similar to terrestrial vertebrates. Across all GBIF records since 2020, 64% of fish species have at least 64% of records from iNaturalist; 64% of insect species have at least 64% of records from iNaturalist, and 62% of plant species have at least 62% of records from iNaturalist.

The unique diversity of species recorded on iNaturalist reflects the unique diversity of the iNaturalist community. The iNaturalist community brings together people posting observations from all parts of the globe and people with identification expertise on all parts of the tree of life. Thank you for all your efforts to create and sustain this truly unique snapshot of the Earth’s biodiversity.

GBIF will only receive your observations if your observation license has a CC0, CC-By, or CC-BY-NC license
You can set your observation license separately from your photo license if you wanted to, for example, allow GBIF to harvest your observations but not your photos or sounds. From your Account settings -> Content & Display -> Licensing, separately set observation license, photo license, and sound license to your preferences.

Currently on iNat there are 58.5M research grade observations with observation licenses that are CC0, CC-BY, or CC-BY-NC. This is the subset that goes into the archive that iNat makes available for GBIF to harvest. There are currently 145.4M observations total (casual + needs id + research grade) on iNaturalist. Of these, 81.4M are research grade and 99.6M have CC0, CC-BY, or CC-BY-NC observation licenses.

Publicado el sábado, 18 de marzo de 2023 a las 03:04 PM por loarie loarie

Comentarios

Thanks for this insight, @loarie -- super interesting!

Anotado por sambiology hace cerca de un año

This is a great way to answer the question: 'why iNaturalist?'. I'm going to share it all over my platforms.

Anotado por stevewoodhall hace cerca de un año

Wow, so most of the reptile observations on GBIF are from iNaturalist? That's impressive!

Anotado por zygy hace cerca de un año

Nice analysis! Thanks, @loarie.

Anotado por janetwright hace cerca de un año

Interesting graphs.

The pattern of the numbers seems odd, why are the % species (x-axis) always identical to the % observations (y-axis)?

63% of mammal species have at least 63% of records from iNaturalist
81% of reptile species have at least 81% of records from iNaturalist
73% of amphibian species have at least 73% of records from iNaturalist.
64% of fish species have at least 64% of records from iNaturalist
64% of insect species have at least 64% of records from iNaturalist
62% of plant species have at least 62% of records from iNaturalist

Anotado por ipat hace cerca de un año

Thanks for posting this!

Anotado por sedgequeen hace cerca de un año

@ipat - I was inspired by the h-index which is a way academics measure impact of publications. The idea is that if you arrange papers in descending order by number of citations, the h-index is where the number of papers equals the number of citations (e.g. h-index of X means you have X papers with at least X number of citations).

Its a nice statistic for summarizing 2-dimensional data. For example, if you had 100 papers with just 1 citation, thats not ideal and the low h-index of 1 would capture that. Similarly if you had just 1 paper with 100 citations, thats not ideal and the low h-index of 1 would also capture that. But if you had 50 papers with at least 50 citations that would be super impressive and the high h-index of 50 would capture that.

In the same way, iI'm arguing that the biodiversity datasets we need to build to meet our monitoring/conservation needs are similar in that we need alot of observations spread across of alot of species. For example having 1M observations of just 1 species isn't super useful and having just 1 observation of 1M species isn't super useful. The iNat dataset in GBIF (since 2020) is unique in that it contributes the majority of records for the majority of species. The h-index like statistic I'm using here captures that as the % of species in GBIF (since 2020) has at least that % of its records from iNat. For mammals, that h-index is 63%, i.e. across all GBIF records since 2020 63% of recorded mammal species have at least 63% of their records from iNaturalist.

I made these graphs for a portion of a talk I gave recently arguing this point. Not sure if its any more or less articulate in presentation form than what I've written here.

Anotado por loarie hace cerca de un año

@loarie Thanks for the explanation, makes sense.

Anotado por ipat hace cerca de un año

Thanks for sharing this information, @loarie. It really shows the importance of iNaturalist and citizen science to the knowledge of the biodiversity worldwide.

Anotado por rsecolin hace cerca de un año

I'm not seeing fungi?

Anotado por sigridjakob hace cerca de un año

I want to nominate the entire iN staff for a Nobel prize!

I'm delighted to be a small part of these efforts and hope that people continue to post good photos along with serious attempts to identify their findings. That! is what really makes the difference to the value of the postings.

Anotado por pam-piombino hace cerca de un año

This paragraph is really striking, thanks for sharing loarie: "Birds represent about 0.6% of named species, but represent about 83% of all GBIF records since 2020." ! The other thing that stands out is how much single countries or regions contribute to the gbif dataset (go Sweden!). Curious what your charts look like by continent. The contributions of naturalists outside North America and Europe would be expected to be even more impressive from that perspective!

Anotado por muir hace cerca de un año

So there are a lot of specialized dataset datasets in GBIF and thanks to iNaturalist the Intormation Facility becomes really Global and Biodivers

Anotado por rudolphous hace cerca de un año

Makes me feel like my contributions actually help in some way.
I still don't understand those last graphs.
How come INaturalist No. of observations are almost identical to other?
So INaturalist contributed ~100k insect records and all other's contributed ~100k insect records.
Since those are logarithmic scales, I suppose when you look at birds and fish, that's where the difference is at the top end but it seems weird that in the other categories, INaturalist is contributing almost exactly 50% of records.

Anotado por ipomopsis hace cerca de un año

Good point @rudolphous. iNat helps put the G in GBIF.

Anotado por muir hace cerca de un año

Great analysis and a strong case for the importance of iNat!

Anotado por cthawley hace cerca de un año

@sigridjakob, I had lumped Fungi into 'other' but if you pull them out, since 2020, fungi were 0.8% of all GBIF records and 2.8% of iNat GBIF records and 0.06% of other GBIF records. Based on the IUCN 2010 data I was using, fungi are 1.8% of named species.

iNat contributed more fungi records to GBIF since 2020 (1,159,832) than any other single datasets, but only 1,159,832/4,397,680 = 26.4% of the all GBIF fungi records since 2020. These are the top 5 datasets from that search:

iNaturalist Research-grade Observations: 1,159,832
Observation.org, Nature data from around the World: 960,355
Artportalen (Swedish Species Observation System): 750,275
Global soil organisms: 602,631
Danish Mycological Society, fungal records database: 233,008

Of the GBIF fungi records since 2020, 31% of species had at least 31% of their records from iNat. This is lower than any of the other groups I graphed above other than birds which is interesting. I suspect left side of the solid 'other' line in the left graph (small subset of species with many observations like Amanita muscaria) is coming from datasets like Artportalen. But I wonder if the high right side of the solid 'other' line on the left graph is driven by lots of species only coming to GBIF from datasets focused on barcoding (molds etc) that are not getting observed and identified by naturalists via citizen-science datasets. Thats what I would also expect with groups like bacteria and viruses. It would be neat to confirm that.

Anotado por loarie hace cerca de un año

Awesome! Thanks for sharing this sort of analytical data.

Anotado por loopy30 hace cerca de un año

@loarie Thank you so much for the fungi analysis! Impressive contributions by Scandinavia community science projects

Anotado por sigridjakob hace cerca de un año

This is wonderful to hear, and many thanks for so clearly and thoughtfully presenting this data.

It looks like only Research Grade are contributing. Is that correct? This does prompt me to once again ask if it's worth contributing taxa like fungi that rarely get identified or confirmed if you can identify them. It's great that iNat is contributing to this important work, but I'm wondering if there's a better place to contribute fungus observations.

Anotado por nancylightfoot hace cerca de un año

I would hope that trying to recruit more fungus identifiers to iNaturalist would be a better option than posting the fungi observations to another site. That said, I think https://mushroomobserver.org/ is a popular iNat-like site focused on mushrooms, but to my knowledge they don't share data with GBIF.

Anotado por loarie hace cerca de un año

This should be published in Science!!! Neat Job!! Congrat community!! We are doing big!!!

Anotado por diegoalmendras hace cerca de un año

So basically, what this is saying is IF the birders put down their binoculars and started recording some of the other small non-feathered stuff, the iNat system would have the entirety of biodiversity within a few weeks?

Anotado por sjl197 hace cerca de un año

@sjl197 no, because they'd never reach research grade! I'm a bird who has tried. I generally don't post anymore in several taxa because they're never ID'd.

Anotado por nancylightfoot hace cerca de un año

@sjl197 birder not bird. :)

Anotado por nancylightfoot hace cerca de un año

Amazing Work again INat’ers
👍👍

Anotado por ck2az hace cerca de un año

I generally don't post anymore in several taxa because they're never ID'd.

@nancylightfoot in my experience birds are among the most quickly identified organisms on iNaturalist, because there are so many bird experts here. So the ID wait for other groups can be a let-down by comparison, for sure. I hope you'll keep posting those other groups anyway. Many will eventually be identified, even if it takes several years, and they will be super valuable when that happens.

Thanks for the analysis @loarie and great work everyone!

Anotado por jdmore hace cerca de un año

Fascinating! Thank you! It occurred to me to compare how my Research Grade observations are distributed among the major taxa, to see how close I come to your pie chart of Named Species, IUCN, 2010. For example, of my 26,626 Research Grade observations, 1.6% are of Birds (almost 40% of those are Pileated Woodpecker holes), 70.4% are of Plants, and 19.4% are of Insects. Clearly, I need to stop photographing the few birds that come within range of my camera, and instead focus on Insects, Insects, and more Insects.

Anotado por lynnharper hace cerca de un año

Keeping with this latest thread . . . I post anything I find if I can get a decent photo for others to see. I hope for RG but don’t worry if I don’t succeed at first. The record is at least there & many times I have gotten ID corroboration much later. This might be from new folks with contributing expertise. Could be from some patient & persistent iNatter doing taxon clean-up. I have learned that, truly, I don’t know what I might find. That’s the fun of discovering.

Anotado por vernal3 hace cerca de un año

👍

Anotado por huttonia hace cerca de un año

"...truly unique snapshot of the Earth’s biodiversity", this is how we try to do!!!! Amazing

Anotado por jeancmf hace cerca de un año

@sigridjakob @loarie The way to analyze the fungi is to split them into macrofungi (likely to be on iNAT) and microfungi (not likley). Here is how we split them (based on: Mueller, G. M., J. P. Schmit, P. R. Leacock, B. Buyck, J. Cifuentes, D. E. Desjardin, R. E. Halling, K. Hjortstam, T. Iturriaga, and K.-H. Larsson. 2007. Global diversity and distribution of macrofungi. Biodiversity and Conservation 16:37-48.
and Thiers, B. M., and R. E. Halling. 2018. The Macrofungi collection consortium. Applications in Plant Sciences 6:e1021 http://www.wileyonlinelibrary.com/journal/AppsPlantSci. with additions by Matt Smith and some exclusions by Bitty Roy
Macrofungal Orders and one subclass:
Agaricales
Amylocorticiales
Aphyllophorales
Atheliales
Auriculariales
Boletales
Cantharellales, but excluding Ceratobasidiaceae? Jean agrees to exclusion of Ceratobasidiaceae in the USA
Corticiales
Dacrymycetales
Endogonales
Geastrales
Gloeophyllales
Gomphales
Helotiales, but excluding these microfungal families: Drepanopezizaceae, Hamatocanthoscyphaceae, Heterosphaeriaceae, Hydrocinaceae, Neolauriomycetaceae, Pleuroascaceae, Sclerotiniaceae
Hymenochaetales
Hysterangiales
Lepidostromatales
Pezizales
Phallomycetidae
Polyporales
Russulales
Sebacinales
Thelephorales
Trechisporales
Tremellales, but excluding the yeast like or filamentous families: Bulleraceae, Bulleribasidiaceae, Carcinomycetaceae, Cuniculitremaceae, Cryptococcaceae, Rhynchogastremataceae, Trimorphomycetaceae. Including only these: Tremellaceae, Nameateliaceae, Phaeotremellaceae and Sirobasidiaceae
Xylariales, but excluding these microfungal families: Amphisphaeriaceae, Barrmaeliaceae, Castanediellaceae, Clypeosphaeriaceae, Hyponectriaceae, Microdochiaceae, Myelospermataceae, Polystigmataceae, Requienellaceae, Zygosporiaceae. Including only these: Diatrypaceae, Graphostromataceae, Hypoxylaceae, Lopadostomataceae, Xylariaceae - Jean agrees to exclusions except some species of Rosellinia that have a subiculum (which are included anyway as in Xylariaceae)
Orbiliales
Neolectiomycetes

Anotado por bit hace cerca de un año

Très beau travail d'analyse, félicitations!
Je suis un ornithologue converti en entomologiste nocturne (observation des papillons de nuit) depuis quelques années. Cette activité m’a permis de participer à la formation, en 2019, d’un groupe de personnes intéressées à cette activité (incluant plusieurs des ornithologues expérimentés). La formation de ce groupe, supporté en partie par l'insectarium de Montréal, a été encouragée à publier leurs observations sur iNat dans le projet d'atlas des papillons de nuit du Québec ( https://www.inaturalist.org/projects/atlas-des-papillons-de-nuit-du-quebec ). Voici quelques statistiques tirées de l'atlas avant et après la formation de ce groupe d’observateurs en 2019:

en 2017, il y a eu 1 809 observations et 1 observateur avec plus de 100 observations;
en 2018, il y a eu 3 596 observations et 6 observateurs avec plus de 100 observations;
en 2019, il y a eu 15 158 observations et 17 observateurs avec plus de 100 observations;
en 2020, il y a eu 36 069observations et 31 observateurs avec plus de 100 observations;
en 2021, il y a eu 46 771 observations et 32 observateurs avec plus de 100 observations;
en 2022. il y a eu 47 329 observations et 35 observateurs avec plus de 100 observations;

Et finalement pour revenir au commentaire publié par @sjl197, notre projet commun de participer à l’atlas des papillons de nuit du Québec à partir de 2019 appuie son commentaire:
«So basically, what this is saying is IF the birders put down their binoculars and started recording some of the other small non-feathered stuff, the iNat system would have the entirety of biodiversity within a few weeks?

Google translate:
Very nice analysis, congratulations!
I have been an ornithologist converted to a nocturnal entomologist (watching moths) for a few years. This activity allowed me to participate in the training, in 2019, of a group of people interested in this activity (including several experienced ornithologists). The formation of this group, supported in part by the Montreal Insectarium, was encouraged to publish their observations on iNat in the Quebec Moth Atlas Project (https://www.inaturalist.org/projects/ atlas-of-the-night-butterflies-of-quebec ). Here are some statistics taken from the atlas before and after the formation of this group of observers in 2019:

in 2017, there were 1,809 sightings and 1 observer with more than 100 sightings;
in 2018, there were 3,596 sightings and 6 observers with more than 100 sightings;
in 2019, there were 15,158 sightings and 17 observers with more than 100 sightings;
in 2020, there were 36,069 sightings and 31 observers with more than 100 sightings;
in 2021, there were 46,771 sightings and 32 observers with more than 100 sightings;
in 2022. there were 47,329 sightings and 35 observers with more than 100 sightings;

And finally to come back to the comment posted by @sjl197, our common project to participate in the Quebec moth atlas from 2019 supports his comment:
“So basically, what this is saying is IF the birders put down their binoculars and started recording some of the other small non-feathered stuff, the iNat system would have the entirety of biodiversity within a few weeks?

Anotado por michelarrivee hace cerca de un año

81M observations on inaturalist are researchgrade but according to https://www.gbif.org/dataset/50c9509d-22c7-4a22-a47d-8c48425ef4a7 GBIF received 58M. Where are the other 23M observations? Is this a license thing?

Anotado por rudolphous hace cerca de un año

Probably, and a lot of iNat observations are copyright. I have seen some obs. with copyright state "This observation is Research Grade! However, it is not licensed for re-use and will not be shared with data repositories that respect license choices." Therefore, GBIF cannot feature the obs. on their website.

Anotado por huttonia hace cerca de un año

Thats right. Currently on iNat there are 58.5M research grade observations with observation licenses (confusingly distinct from photo licenses) that are CC0, CC-BY, or CC-BY-NC. This is the subset that goes into the archive that we make available for GBIF to harvest.

There are currently 145.4M observations total (casual + needs id + research grade). Of these, 81.4M are research grade and 99.6M have CC0, CC-BY, or CC-BY-NC observation licenses.

From your Account settings -> Content & Display -> Licensing you can set your observation license, photo license, and sound license separately, if you wanted to, for example, allow GBIF to harvest your observations but not your photos or sounds.

Anotado por loarie hace cerca de un año

Thanks for your clear explanation. Just checked I have mine observations CC-BY so they are shared. Also nice to see that all data that could be shared with GBIF is already received by them. iNat team does a great job.

Anotado por rudolphous hace cerca de un año

Hi all. I gave a short comment before, but the essence i put before was very flippant, against that i also absolutely laud/applaud @ loarie and team for making what is here now clearly shown as epic contribution to biodiversity assessment, plus of course let's be mindful of the vast number of contributors. For me as others, this analysis is awesome towards the second useful comment @stevewoodhall about "why iNat", now being shown as a massive contribution towards assessment that is open to whoever is able to contribute. I guess all i hoped for from my flippant comment before if summarised was my hope that 'birders could also look at other biodiversity", it's really part of an older lament to recorders in general in hope to become less 'vertebrate-centric' and try for more future observation efforts that better balance the ratios of biodiversity. On that, i'm concerned by the comparisons against "named species IUCN" as last i saw the taxonomy by IUCN was pretty dire - for vertebrates i expect fine, but for the other stuff i feel no-one cares about then, pfff.
o
@ancylightfoot - it's worrying the discordance with expectations, for me with inverts if something can be got to family or genus with a photo thats really good, but for people who come from realm of wll studied vertebrates (birds, reptiles etc) that's dire!
@michelarrivee much appreciated sir!

Anotado por sjl197 hace cerca de un año

Nice

Anotado por marinolinic hace cerca de un año

very interesting ! This shows the power of iNaturalist in its diversity.

Anotado por louis_aureglia hace cerca de un año

@loarie thanks a bunch for this awesome analysis! As there are still loads of users who don't know how the license settings affect data sharing, I suggest to add a para on that at the end of your text - many users might not read through the comments including yours on licensing & GBIF. Cheers!

Anotado por jakob hace cerca de un año

wow, very neat!

Anotado por charlie hace cerca de un año

Fascinating and outstanding analysis, @loarie. Thanks for sharing these in graphical form, and also your answer about the h-index, which confused me as well. And another tidbit in here - I've felt a tad guilty about having a copyright license for my photos, as I knew (or thought I knew) that they would not be accepted by GBIF. Your discussion about being able to license observations separately from photos was something I had missed before. I immediately changed my licensing to indicate CC for observations and sounds. And hopefully it will go back and include the 8000+ RG species observations and send them over to GBIF. (I checked that box but not sure whether it took.) I suspect there are quite a few others in that category who would change their licensing options if they understood it was possible to separate licensing of observations from photos.

Anotado por stevendaniel hace cerca de un año

Wow-very interesting.

Anotado por yayemaster hace cerca de un año

👏👏

Anotado por poupouksios hace cerca de un año

jakob - I added a paragraph on the end as you suggest

Anotado por loarie hace cerca de un año

On the topic of things not being identified right away, a plant once thought extinct was recently found and verified by a group of researchers. When they later checked iNat, they saw that students had posted photos of that species several years earlier. You never know what that unidentified plant or fungus might turn out to be. https://www.cnn.com/2022/04/15/world/extinct-flower-ecuador-scn/index.html

Anotado por tiwane hace cerca de un año

@loarie & @tiwane...Is it possible to search for our respective GBIF contributions---either from our observations/species categories or from our year-end stats?

Anotado por iranah hace cerca de un año

Outstanding! I'm impressed and surprised that "42% of the species have 100% of their records from iNaturalist." That really underscores the important role for iNaturalist in filling in the gaps for the many species that no other institutions are recording.

Anotado por jon_sullivan hace cerca de un año

I wonder - could a message about GBIF and copyright choices be sent to users as a 1-time explanation? Seems likely you could get a few million additional entries just from people understanding specifically what their copyright choice is keeping their data out of.

Anotado por jonhakim hace cerca de un año

Thank you to the whole inaturalist team. It is a very nice job.

Anotado por khellaf-rebbas hace cerca de un año

not discussed are the geographic distribution of iNat data which are US and Canada centric, in other words the h-indices for US, Canada, and Texas are high. I suspect the h-index for Texas observations is higher than that of nearly every country outside of North America.

Anotado por entomike hace cerca de un año

As @entomike suggests, the h-indices likely vary between regions. I would be very interested in the proportions of total species diversity observed by country or at least broken down by major regions/continents. Perhaps that could be the subject of a future iNat blog post?

Anotado por loopy30 hace cerca de un año

Añade un comentario

Entra o Regístrate para añadir comentarios