Saturday, May 30, 2015

Social Network Analysis in Cattle Breeding


Cattle raising is one of the most important agricultural activities in Colombia, it occurs in almost every region of the national territory with Antioquia, Casanare and Córdoba leading the ranking.

Cattle raising in Colombia is mostly rural and done by individuals. Currently it contributes to 3.6% of the national GDP and among Latin American countries, Colombia ranks fifth in cattle raising.

In 2014 the national distribution of cattle was the following:

Female
12’516,695
Male
6’155,541
Calves
3’921,046
Total
22’593,282

Due to the benefits of weather conditions in Colombia, cattle are bred for dairy, beef and a dual purpose, which is not possible in every country.

According to Fedegán, (the Colombian federation for cattle breeders) there are 23 types of cattle in the country, however here I will explain the most emblematic types.
                                                                               
                                            
-Brahman: Famous for producing meat.
-Gyr: Produces between 30 and 35 liters of milk daily.
-Holstein: Produces between 20 and 22 liters of milk per day and has become more important because of its milk quality.
-Jersey: Produces between 15 and 20 liters of milk daily.
-Normando: Dual purpose breed.


Basically there are two ways of buying or selling cattle in Colombia, through a livestock auction or direct agreement. A great disadvantage of livestock auctions is that they are located in big cities or towns, what makes it a difficult option for peasants to travel long distances with their animals without being sure they are going to sell them at a good price. Also, it is widely known that before the auction starts the best animals have already been sold backstage and some animals are overpriced.

Regarding a more personal selling, it mostly depends of every individual personal network, what makes it hard to know breeders from other regions.

These reasons are what led me to believe that if we had endless resources and unlimited access to information and time, a Social Network Analysis could make a huge improvement in the relationships of breeders in Colombia and mostly in the cattle selling process.

Additionally, it could help share important information about vaccines, pastures, new artificial insemination techniques, milk production levels, birth rates, among other things.

For the purpose of this analysis, I have created a network with the most relevant regions in Colombia where cattle breeders are clustered and some individual farmers are connected to them due to geographic location.




For example, let’s say farmer “d” breeds Holstein cattle, he is only connected to another farmer which is “a” and the auction of “Casanare”, however in Casanare they mostly sell Brahman and Gyr so his cattle is not as valued in that region and he is forced to sell it underpriced.

But if there was the possibility of understanding the cattle network of the country with all the specific details of what animals do they breed and under what conditions; farmer “d” could easily understand the way to contact someone in particular in the region of “Cundinamarca” who specializes in Holstein.

This network can also be valuable to spot in which region there is going to be an abundance of sacrifices to sell meat where maybe in other regions there is a shortage and it would help fighting the price hike.

Further more, this analysis can be applied for Latin America or to be more ambitious the rest of the world, in 2013 Brazil exported 668,000 heads of cattle and Colombia exported 300,000. Creating this type of network would allow breeders to find the connections in those countries we already know are demanding cattle produce such as South Korea, Japan and China.

If we were able to build this network, there would be amazing opportunities and possibilities for cattle breeders.


Cristina Osorio

1 comment:

Christopher Tunnard said...

This is a really interesting and well-conceived idea. There are 2-mode SNAs of markets and products, or markets and traders, but this is the first I've seen that involves cattle markets and the potential to improve market efficiency by matching farmers, breeds, and markets. I don't think you'd need endless resources, either, as the information must be relatively available. Nice work.