Do ecologists have enough raw data for accurate studies?

[Jack Zalesskiy]

Hello, everyone! I would be very grateful if you could help me find an answer to my question.

Since in nature, everything is connected to everything I can imagine that one requires tremendous amounts of data for the research to see how much individual events may affect the changes in populations, habitats or biomes.

I know there is ILTER, who connected several institutions into a network that is generating statistics across multiple countries but they are present only in 40 countries if I am not mistaken. And even if there is enough data, then there might be problems related to storage or searching for the right information.

Just where these limitations lie in modern ecology if there are any and what is the scale of this problem?

[Sensei]

Quote

Do ecologists have enough raw data for accurate studies?

I would say, nobody have enough data. That's why new data are gathered, all the time, to fill gap. Some trends are visible only after many years of observations.

e.g. satellites are taking photos of globe, e.g. Arctic and Antarctic, and scientists are merging them in timelapse video (what is hard to overthrow argument). Comparison of them in timeline reveals changes, regardless of local government attitude to ecology. If they cut forest, or other way damaged natural environment, it can be detected by analyze of photos. And you constantly need more data, because planet is changing.

 

Edited March 20, 2018 by Sensei

[Jack Zalesskiy]

7 minutes ago, Sensei said:

I would say, nobody have enough data. That's why new data are gathered, all the time, to fill gap. Some trends are visible only after many years of observations.

e.g. satellites are taking photos of globe, e.g. Arctic and Antarctic, and scientists are merging them in timelapse video (what is hard to overthrow argument). Comparison of them in timeline reveals changes, regardless of local government attitude to ecology. If they cut forest, or other way damage natural environment, it can be detected by analyze of photos. And your constantly need more data, because planet is changing.

True. But what I am trying to get at is what limitations do ecologysts meet, or even, what limitations does ecology as a whole meet as a consequence of the lack of data. Are there any fundamental flaws with the current infrastructure when it comes to gathering, storing and presenting the data? Is a network like ILTER enough or is it still in an embryon stage in terms of scale?

[SOS2000]

I would agree there is not enough raw data from your opinions and this forum has helped me broaden my knowledge on the topic.

[Sensei]

Hostile governments are the main source of problems for ecologists. Cutting funds for ecologists is the smallest tool of hostile government against ecology..

 

[Jack Zalesskiy]

1 hour ago, Sensei said:

Hostile governments are the main source of problems for ecologists. Cutting funds for ecologists is the smallest tool of hostile government against ecology..

 

Thank you for your answers Sensei ) 

[Area54]

On 20/03/2018 at 8:59 AM, Jack Zalesskiy said:

Since in nature, everything is connected to everything I can imagine that one requires tremendous amounts of data for the research to see how much individual events may affect the changes in populations, habitats or biomes.

While what you say is generally true, it is my understanding that many ecologists focus on a very narrow topic, at least initially. Thus, by tackling the problem piecemeal, it becomes manageable. I agree that challenges must exist when one attempts to synthesise data to create a larger picture, but huge strides have been made in dealing with Big Data.

Early days - Exciting Times.

[pavelcherepan]

On 3/20/2018 at 4:59 PM, Jack Zalesskiy said:

Since in nature, everything is connected to everything I can imagine that one requires tremendous amounts of data for the research to see how much individual events may affect the changes in populations, habitats or biomes.

As Area54 pointed, science theories usually look into a small subset of data, that's why we have a very old physics joke about "spherical chickens in vacuum".  Also for many complex systems (or even relatively simple like 3-body problem) there are no analytical solutions, but only numerical models, and the more conditions you take to describe the system, the higher the computing power required to model such system, which grows exponentially with each new condition. So I'd say it's not that we don't have enough data, it's finding a way to model ecosystems with all the data we currently have available. Hopefully, developments in quantum computing will help.

[John Cuthber]

In general, science does not show that anything is true- it just fails to show that it is false.

Ecology is no different.

[Jack Zalesskiy]

16 hours ago, Area54 said:

While what you say is generally true, it is my understanding that many ecologists focus on a very narrow topic, at least initially. Thus, by tackling the problem piecemeal, it becomes manageable. I agree that challenges must exist when one attempts to synthesise data to create a larger picture, but huge strides have been made in dealing with Big Data.

Early days - Exciting Times.

1

Area54, thank you for the detailed reply, it was very interesting to read it. Agree, this is exciting times indeed! 

9 hours ago, pavelcherepan said:

As Area54 pointed, science theories usually look into a small subset of data, that's why we have a very old physics joke about "spherical chickens in vacuum".  Also for many complex systems (or even relatively simple like 3-body problem) there are no analytical solutions, but only numerical models, and the more conditions you take to describe the system, the higher the computing power required to model such system, which grows exponentially with each new condition. So I'd say it's not that we don't have enough data, it's finding a way to model ecosystems with all the data we currently have available. Hopefully, developments in quantum computing will help.

3

Hello Pavelcherepan! Very interesting... Do you think blockchain can help us with complex modulating? I read that decentralized databases allow distributing the load among all connected units, so, in theory, we could distribute the calculation between thousands of computers?

2 hours ago, John Cuthber said:

In general, science does not show that anything is true- it just fails to show that it is false.

Ecology is no different.

Still, you need the data and the ability to process it to arrive at any conclusions. 

[MathGeek]

It depends on the specific question at hand.  Relating two ecological variables is not too hard in a given system, though some variables are easier to measure than others.  Once one has a body of measurements, then one can compute correlations and p-values to test a hypothesis about the relationship.  One study I'm involved with relates body condition to water temperature in fish.  Computing the correlations and even the slopes (how much fish condition decreases for a given increase in temperature) is straight forward.  Weighing 500-1000 fish of a given species over a 7 year span is usually sufficient for decent correlations and p-values as well as slopes that are significantly different from zero.  

Moving beyond correlations to establish underlying causality is harder.